Podcasts about AB

On this week's EcoNews Report, Assemblymember Chris Rogers joins the program to discuss this year's legislative session. Asm. Rogers has emerged as an enviro legislative darling, with bills like year's AB 263, which established minimum instream flow protections for the Shasta and Scott Rivers. Asm. Rogers joins the show to preview three new and exciting bills: AB 1984 would redefine corporate powers under state law to remove corporation's ability to spend money on elections. (Asm. Rogers recommends this article to learn more.)AB 1699 would remove operational hurdles to prescribed fire and address liability issues with the goal of expanding "good fire."AB 2494 would reimagine state-owned demonstration forests, changing their management goals from "maximum sustained production" of timber to managing for climate, clean water, wildlife, and more.Support the show

Assemblymember Chris Ward (D–San Diego) held a press conference Tuesday at the State Capitol to announce the introduction of AB 1542, new legislation to strengthen protections for sensitive personal data; continued efforts to advance AB 322, a two-year bill to ban the sale of geolocation data; and renewed momentum for AB 1337, a two-year bill currently pending in the Senate Judiciary Committee to modernize public-sector privacy protections. The press conference brought together consumer advocates, civil rights organizations, and privacy experts to underscore the urgency of protecting Californians' personal information from misuse, exploitation, and sale without consent. “Californians should not have to worry that their sensitive personal information is being sold to the highest bidder,” said Assemblymember Chris Ward. “From precise location data to deeply personal information, these bills work together to stop the sale of geolocation data, strengthen protections for sensitive information, and ensure government agencies are held to modern privacy standards. California led the nation on privacy once before, and we must continue to lead as technology evolves.” Justin Brookman, Director of Tech Policy at Consumer Reports, warned that data-driven pricing and monetization practices are outpacing existing protections. “People should not have to worry that their sensitive personal information is going to be sold to the highest bidder,” Brookman said. “The California Consumer Privacy Act was groundbreaking, but it needs to be updated to address the realities of the modern data ecosystem. Companies should use personal information like geolocation to deliver the services we ask for—not to secretly monetize it through data brokers.” Advocates emphasized the heightened risks these practices pose to vulnerable communities. “When businesses sell and trade sensitive personal information like precise location or immigration status, they open the door to surveillance, targeting, and exploitation. Those harms fall the hardest on the most vulnerable in our community, including immigrants, LGBTQ+ individuals, and survivors of domestic violence and human trafficking," said Lan Le, Policy Advocate at Asian Americans Advancing Justice Southern California (AJSOCAL). “These data privacy bills send a clear message: dignity and safety are rights, not commodities.” Supporters also highlighted the need to modernize how public agencies handle personal data. “In an era of increasing digital surveillance and data collection, it's crucial that our privacy laws evolve,” said Rindala “Rin” Alajaji, Associate Director of State Affairs at the Electronic Frontier Foundation. “AB 1337 is a much-needed update to ensure local governments are held accountable for how they handle personal data.” Tracy Rosenberg, Executive Director of Oakland Privacy, underscored how the measures work together. “The bill duo of AB 1337 and AB 322 attacks our current dystopia in two vital ways,” Rosenberg said. “They modernize privacy protections, add transparency and limits around precise location data, and curb invasive practices that expose Californians to government and industry overreach.” John Bennett, Initiative Director at CITED, emphasized the broader democratic stakes. “Privacy and freedom of movement are cornerstones of a healthy democracy,” Bennett said. “It's time to strengthen our data privacy laws and fulfill the promise of California's constitutional right to privacy—so people can move, assemble, and participate in civic life without fear of surveillance.” Ward's legislative package builds on California's landmark privacy framework to protect sensitive personal data, prohibit the sale of geolocation information, and ensure privacy rights keep pace with modern technology.

We find out how "weird" we are, and we discover AB's NEW friend group.

How to Fix Your Underperforming B2B SaaS Funnel for Quick Revenue Wins In the fast-paced world of B2B SaaS, the ability to go to market, iterate on feedback, and close deals rapidly is the ultimate competitive advantage. Unfortunately, many sales and marketing teams find themselves stalled by underperforming funnels that drain resources without delivering measurable results. When growth plateaus, the challenge lies in transforming these stagnant pipelines into high-velocity growth engines without requiring massive capital or long timelines. So, how can B2B SaaS teams identify the hidden leaks in their customer journey and unlock quick-win revenue through a strategic, data-driven approach? That's why we're talking to April Syed (CEO of Aperture Codex), who shares her expertise on fixing an underperforming B2B SaaS funnel for quick revenue wins. During our conversation, April discussed the importance of leveraging data to pinpoint “quick wins,” such as streamlining sales processes and eliminating high-friction points in user onboarding. She explained how to fix “conversion killers” like messaging misalignment and highlighted the necessity of aligning marketing and sales efforts to ensure a seamless experience. April also advocated for a culture of continuous testing, using small, incremental experiments to de-risk major strategic shifts. She emphasized the value of regular customer journey mapping to maintain a predictable, sustainable, and highly efficient path to profitable growth. https://youtu.be/VeeFMznhCfw Topics discussed in episode: [07:24] Why your Ideal Customer Profile (ICP) must be a “living, breathing” document reviewed quarterly, not a static file sitting in a deck. [11:24] The critical mistake of treating marketing as a cost center rather than a revenue driver, and how it leads to “vanity metrics” over actual sales. [13:53] Why you should focus on small, incremental tests to “de-risk” big spends before committing to expensive strategies like rebrands. [18:05] The 5-Point Conversion Diagnostic: A framework to analyze time-to-value, messaging alignment, behavioral triggers, follow-up timing, and pricing friction. [23:07] A real-world example of how “pricing friction” (forcing an annual upgrade) caused a loyal promoter to churn to a competitor. [27:24] How to audit your funnel for “Quick Win” revenue opportunities in under 30 days by analyzing where deals stall in the CRM. [35:27] Why no marketing asset is ever “final”, and why high-traffic landing pages should be in a state of constant A/B testing. Companies and links mentioned: Apryl Syed on LinkedIn  Aperture Codex  Superhuman Notion  Motion Transcript Christian Klepp, Apryl Syed Apryl Syed  00:00 Brand for instance, doesn’t work itself into any metric, but it makes every metric better across the board. Sometimes we’re chasing these metrics and like the attribution of where a particular deal came from, or how did they find out about us, and we’re not thinking about all of the things that are outside in the flywheel that are, you know, causing that person to, yes, eventually convert. But were there seven or eight other things that kind of they interacted with. Christian Klepp  00:26 In the world of B2B SaaS speed is the name of the game. Get to market, quickly collect feedback, quickly iterate quickly and close deals quickly. But what happens if your sales and marketing teams get stuck with underperforming funnels that don’t generate the results you need? How can teams turn these funnels into growth machines without massive spend or long timelines? Welcome to this episode of the B2B Marketers on a Mission podcast, and I’m your host, Christian Klepp, today, I’ll be talking with Apryl Syed, who will be answering this question. She’s the CEO of ApertureCodex who gives founders the strategy and the psychology needed to jump into fast revenue gains. Let’s dive in. Okay, and away we go. Apryl Syed, welcome to the show. Apryl Syed  01:12 Thank you so much, Christian. I’m so excited to be here. Christian Klepp  01:15 Glad to have you on the show. I think we had such a great pre interview conversation. I kept telling myself I should have hit record, and I talked to you the first time, right? But, you know, two times is a charm or three times. But anyways, this is the second time we’re talking. So I’m really looking forward to this conversation Apryl, because we’re going to touch on a topic today that I think is not just relevant to sales teams. It’s really important to marketing teams as well. So I’m going to keep the audience in suspense just a little while longer while I set up this first question. Right? So you’re on a mission to help B2B SaaS teams turn underperforming funnels into growth machines without massive spend or lengthy timelines, and for people that didn’t hear that the first time, I think everybody wants something like that, right, quick results without spending massively, right? So for this conversation, I’d like to focus on the following topic and just unpack it from there, right? So how can SaaS teams leverage a quick win revenue approach for better and more predictable growth. And I mean, come on Apryl, who the heck doesn’t want that, right? Who doesn’t want predictable growth, right? So I want to kick off this conversation with two questions, and I’m happy to repeat them. So first one is, where do you see many SaaS teams struggle with revenue growth? And the second question is, what are some of the key causes of this? Apryl Syed  02:44 It’s really great, by the way. As a side note, I got turned down for a podcast this week because they said I talked too much about quick wins, and they felt that it conflicted with their policy. I won’t mention the name, they’re an agency out there, but they were all about big spend, and they felt that I conflicted with that. And this exactly ties in. This is probably why the subject that I talk about so. Christian Klepp  03:13 Well, I’m sorry for them. Apryl Syed  03:15 Yeah, that’s okay. That’s okay. We don’t, we don’t match. You know, I’m not for everyone. Well, I think that, like SaaS teams don’t realize that they’ve got data. And within their data really, really lies some of the tweaks, opportunities and things like that that can make them extra revenue that they might not be looking at today. And I think, you know, perhaps it’s in tweaking their sales process. Maybe they don’t have a sales process misalignment between sales and marketing. Marketing is talking about one thing, sales is selling another thing, or could be marketing is marketing to one type of industry and user, and sales is saying that’s not the right user. It’s something completely different, that misalignment in itself causes revenue conflict, revenue opportunities. And you know, sometimes it’s spending on expensive tools before you’ve actually broken down some of those points in the funnel. Or could be tools that you’re getting a lot of data from, or they’re not doing anything with the data on a regular basis. So I think, you know, those are where I see some of those, like, struggle with revenue because of some of those issues and and then I think your second question was kind of like, well, how to, how do they kind of avoid some of those scenarios? Right? Christian Klepp  04:40 It was more about the the key causes, but you but, but you did talk about that already, right?   Apryl Syed  04:44 So, right, right? That definitely is there. Well, I think, you know, it’s also could be, you know, where they’re chasing certain metrics and focused in, and we had this conversation earlier. It’s like brand, for instance, doesn’t work at. Yourself into any metric, but it makes every metric better across the board. So sometimes we’re chasing these metrics and like the attribution of where a particular deal came from, or how did they find out about us, and we’re not thinking about all of the things that are outside in the flywheel that are, you know, causing that person to, yes, eventually convert. But were there seven or eight other things that kind of they interacted with before they got to that point? And we had to get them ready? So, you know, can definitely be about just chasing those metrics too much, which means you avoid doing things that don’t give you that instant metric. And I think that is a big challenge and pitfall that that teams can can certainly fall into. I think also the the challenge of treating marketing as a cost center and not letting them be in charge of all of those metrics down to the sale that happen. And that might sound weird to some folks, but I’ve certainly been in enough teams and enough experiences across you know my background that I’ve seen that sometimes you can make a change in marketing. It produces a lot of leads, but those leads aren’t qualifying and they’re not turning into revenue, and yet, if the metric is producing leads, well then marketing can walk away the end of the day and meet their metrics and jobs, but if the metric is revenue, then they’ve got to go all the way to that end cycle and see that it’s a qualified opportunity. That, of course, goes back to my original point that if sales and marketing aren’t in lock sync with each other, and they don’t have a good relationship and dynamic, then it ends up in finger pointing when things aren’t going wrong, instead of both teams coming together, being on the same page and figuring out what’s going to work. And that’s that’s really the key. Christian Klepp  07:03 Absolutely, absolutely. And I think you might have brought it up, and maybe I didn’t catch it, and if not, I apologize. But like, one of the things that I didn’t notice, too, is, like, this misalignment of who, who the who the ICP (Ideal Customer Profile) is, like the assumptions that both sides have and then somehow they just cannot meet in the middle. Apryl Syed  07:24 Well, I kind of brought it up just slight when I said that marketing might be marketing to one person, and sales is selling to another, but if we just want to double click, you know, on on that, that agreement around the ICP, the reason why it’s so important, and I think it’s hard for some SaaS companies, because there’s, there could be a lot of ICPs. And I kind of have this philosophy that with an ICP, people usually maybe do these personas, as I call them, one time, maybe at a, you know, a planning session or whatever, where they’re kicking off, you know, and kind of like planning who those are, and then they leave them. They sit in a deck somewhere. They’re never looked at again. They’re never revised. I like a more fluid method with personas. I like personas to kind of be active, living and breathing in something that’s reviewed on a quarterly basis, I think is a better cadence. And the reason being is, like, we want to see how many deals we’ve closed in that particular area, how many so we should be looking at the metrics right by persona. We should also look at the messaging by persona to see how that’s working. And we should, you know, look at our team and how that flow has gone through into the sales process by persona. And kind of looking at this lens, we may figure out that one persona is working really, really well, or two or three might be working really well. And maybe there’s two or three that aren’t working really well. We might want to flush those out or put them in, what I would say is like a vault or a holding pattern. They might come back later if something’s happened, and we might want to add different ones. And the reason why quarterly is important is because, if you are selling business to business, for instance, in that business environment, there are different things that might be happening in the world, you know, geographically, politically, that might be impacting a certain persona. And it’s important to also look at that lens on a quarterly basis and say, Okay, what’s the mindset of this particular persona? What are they dealing with? What are some of their issues? What are their pressures? What is their emotional state, and then how do we want to message into that emotional state during this time? How do we want to change and revise our messaging for what’s going on in their world right now, this quarter, right you can’t keep you can’t keep messaging the same and messaging constant needs to be looked at. I would say, on a regular basis, one to check and make sure it’s working. If it’s working, keep it working at some time. At some point, though, it might stop working, and it’s important to catch that as you see those numbers trailing off, as you see that change, and not wait until too long has passed and just double down on the same persona for the sake of really work, working with it, because it was the original plan. Christian Klepp  10:27 Yeah, absolutely, absolutely these, um, these personas are, and I believe that too, they it’s not something that that’s written in stone, and then you, you to use that archaic expression, just keep it on the shelf, and then it collects dust, right? Apryl Syed  10:40 Yeah Christian Klepp  10:41 It’s something that should be monitored, as you said, because certain certain companies are working in industries where, for example, government regulation impacts them. Apryl Syed  10:51 Yes. Christian Klepp  10:52 If government regulation changes, then that perhaps also influences the way they make decisions, or decide to work with external vendors and partners and so forth, right? Apryl Syed  11:05 Absolutely. Christian Klepp  11:07 You brought you brought up a few already in the past couple of minutes. I’m just, I just want to go back to pitfall. So one of them, I think, was chasing this, chasing metrics. Right? This, this habit of constantly chasing metrics. What are some of these other pitfalls that you’d say marketing teams should avoid them. What should they be doing instead? Apryl Syed  11:24 Well, I think, you know, another pitfall that I’ve seen is kind of launching a big rebrand and expecting, you know, or that could also be a plot, a platform overhaul, software overhaul, and expecting that that’s going to move the needle faster when you could test that type of messaging out in really small ways before you go and do that big rebrand. And I’m a big fan of those, like small tests, verify and then go big. Like I’m not I’m not saying don’t ever go big. What I’m saying is like, test and measure before you go into a big cut, a big, fresh rebrand, because it’s expensive, and you want those big, expensive expenditures to be a little bit more of a sure thing than a risky thing. So de risk the big spends, riskier moves. Do small, incremental tests and say, how could we test this out on a small scale. How could we test or rebrand out? How could we test a platform change out before we do that in a small way? So I think that’s another one. I talked about a cost center. Treating marketing as a cost center is another one. So I think those are, like my big, my big three, I would say, in terms of pitfalls. Christian Klepp  12:41 Yeah, fantastic, fantastic. You, you hit on something there with your with your third point. And I want to go to that, because that’s a topic that, um, that as a marketer, personally, it riles me up a little bit, but, like, you know, but, but we have to look at this as professionals too, and say, okay, you know what? In the world of B2B, that type of pushback is almost expected, right? Because I’m not sure what your experience has been. But I also work with a lot of companies that have done either little or no marketing before, so it’s, it’s to a certain extent, it’s like Terra Australis incognita. It’s uncharted territory. They are not sure what to expect. So it’s only, it’s only normal that they, that they view it with some kind of, I wouldn’t go so far as to say, suspicion, but yeah. Like, how do you know it’s gonna work, right? So over to you. Like, what’s your experience been? How do you deal with companies that view marketing with that kind of suspicion or or have these doubts, like, Is this even going to work for us? Right? How do you deal with that? Apryl Syed  13:53 Well, I mean, from my perspective, I think again, I go back to the small tests, small wins in those beginning, like, let’s get our sea legs before we go and launch some big strategy. And I think that’s, you know, a big divide between, you know, maybe myself and yourself and some other you know, marketing agencies and firms out there is, I would rather get small, incremental wins to start. I’m not against big strategies and big spends. I think they’re both needed, but when you’re kind of coming into a team that’s either had little to no success with marketing, because maybe they’ve had some bad experiences with agencies that haven’t delivered, or they’ve tried ads, or they’ve tried this thing and they kind of have that bad taste in their mouth, right? Or they just have not done anything at all, and perhaps they’ve, they’ve grown despite that. So they’re kind of like, Hey, I’ve seen success without doing this. So why? Why do I need this? So I think an educational approach is important, kind of giving the here’s the industry benchmarks, here’s what we should. See, here’s how we are going to test. Here’s a recommended way that we do small, incremental tests. And then I also think a really, really important piece is, if it’s a company that’s been around long enough is to dive into that data I have. I have a customer that I would say sits in this category. They’ve grown tremendously. They’ve had a very successful business, and they’ve never marketed before. And if I were to come in there with some big rebrand strategy, big moves, look at me like you’re crazy. We don’t need that. I mean, in all honesty, what are they looking for? They’re looking for incremental revenue gains. So how am I going to produce incremental revenue gains? I’m going to look at their data and see where there’s holes in gaps today, where, yes, marketing, but marketing is a very, very broad term. Marketing can be brands, marketing could be emails, marketing can be social media. Marketing can be customer advocacy, customer emails churn, you know, upgrading customers into other models. So when I say I look at data, I look at what their customers are doing, and what I get from that is, where is my ideal customer, because it’s going to show me in their base. So who might I want to go after and experiment with? First, those are going to be my biggest areas for opportunity of wins, where, with their existing customer base, can I sell something more or different for them to increase revenue in that way? I think that’s another big and then I look at where there may be failures across the process in their data. If it’s a SaaS company, let’s look at their free the trial, trial, you know, to paid, paid to churn, and look at those numbers and say, are they hitting industry standard for their industry? Can I improve any of these metrics? Let me go look at all of the various different things that are going to change these metrics. Where can I start to experiment to get incremental change? That’s how you give success to a team. And they start feeling like, Okay, we should invest more here. We should do more here, because it’s working. Now, let’s double down. Let’s triple down. Let’s do more, then you can go after those bigger strategies. Christian Klepp  17:26 Yep, yep, no, absolutely, absolutely, no. I’m glad, I’m glad you brought those up, because that’s a great segue into the next question, which I think you’re all too familiar with, right? So I think when we first talked, right in our previous conversation you were talking, you mentioned something called a five point conversion diagnostic, which uncovers, I think you refer to them as conversion killers, right? You can cover these conversion killers without expensive tools or massive product like changes or revamps, right? So if you could please walk us through this five point approach and how teams can leverage that. Apryl Syed  18:05 Now this is particularly for SaaS, that trial to onboarding experience and the time that I the thing that I look for the most in there is time to value. How long does it take for the customer to experience value is going to be indicative of how long their trial has to be with that onboarding experience, and are they legitimately going to get into the point of buying early, even because they can’t wait to utilize this tool or buying, of course, the moment that the trial, the trial the trial ends. That is all about time to value. The second is about messaging alignment. So does the promise that we give, if it’s a landing page, whatever that experience is that someone comes through to then get to that product, does the promise of what we’re giving them match what the experience is going to be in the software, and how long does it take again, from that time to value, for them to get to that matched experience of what we promised that will also be a predictor of so if we were, you know, on a scale from zero to 10, 10 being like matched, it perfectly, zero being not matching at all, we’d want to rate our company on that scale, and kind of see for the time to value and for the misalignment, where are we? Then I would kind of go after like behavioral triggers, and I would try to figure out what actions correlate with conversion. So I would look at everybody that’s converted, and I would say, what parts of the software did they touch right? Are they looking at, are they experiencing, which then would predict, like, if people do these five things and the solution, then we know that they’re going to convert. And you can use either, like a Pender or you know, products like that that give you some of that analysis and data. Or maybe it’s, you know, sitting in your CRM, but that would tell you and inform you about your messaging as well. Like, what should we be messaging about? These are the key things that people want out of this solution, and that’s going to inform your next piece, which is, I would look at the follow up timing, the sequencing. How frequently do we talk? I often, I’m a big superhuman fan, and I talk about superhumans onboarding experience, which I think is awesome. And of course, they get a little bit of a leg up because they are an email solution, so they see when you’re in the tool. But I have found that, like the timely messages and the trickling of features that they give you right when you’re ready to use that feature has been so well thought out. And if you have, if you have not experienced it, and you’re a SaaS product owner, Founder, CEO, I highly encourage you to go through their onboarding experience, because that, to me, is like the pinnacle, or one of the pinnacles of what you should want your users to experience, like these just great aha moments right when they’re ready to receive them as part of that trial period before conversion. That make sure that we’re just touching them at the right moments. And then the last piece that I look at is pricing and packaging friction. And here’s, this is, you know, this is something that’s changing an awful lot right now. SaaS is under pressure to maybe look at not seeds, but maybe it’s volume, but then volume is not great, because people can’t predict it, and certainly can’t budget appropriately for it. So there is all kinds of pricing friction happening right now that needs to be figured out, but understanding where people are dropping off and where in that you know, how many clicks do they need to do before they buy? What is that whole buying process like? What is the upgrading process like? Put it through the pressure test. See how many steps it is. Challenge yourself. If you can reduce the steps, make it easier. I’ll give you an example. I was a big, big user of the motion app for a really long time. I probably sold, let’s say, 10 to 20 of these to other people, because I was such a promoter and such a fan of motion, they changed something in their solution related to how many credits, and what happened is it stopped recording my meetings for me automatically, which meant didn’t go into my notes anymore. Didn’t automatically create my tasks for me. That’s a pretty big feature, and obviously I so I went to upgrade, and the upgrade didn’t allow for me to choose a monthly it only allowed me to upgrade to choose an annual. Christian Klepp  23:06 Why? Apryl Syed  23:07 Yeah, which did what to me as the user. I then went into the shopping mode, essentially, and I said, Now I’m going to go shop and look at, well, what other tools are out there that can do the same functionality. Because now, if I have to commit to an annual plan, so much changing in AI this year, I’m not sure if I can commit to an annual plan. It had nothing to do with the amount of dollar spent. It had everything to do with commitment. And here I was a promoter of their solution. I ended up canceling and I went with notion, because I realized that notion had added a significant number of AI features at a much lower price, which I know a lot of people complain about notion being expensive, and it isn’t as good of a user experience now that I’m using motion and yet notion. Yet, I’m still on notion, and I left motion app, which is probably better, because they put me through this experience. And I say that as an example not to and I don’t know if they fix that, but we make these decisions all the time, sitting from our lens, looking at what we want the outcome to be, and we don’t think through what that user experience is going to be, and we’re killing conversions, in some cases, by these little levers and moves that we make, and sometimes we don’t even realize that. So I really encourage, encourage founders, encourage, you know, everyone at the company go back through and look at these tiny little things that each one of them on the loan alone could be costing you revenue, costing you conversions along the pathway. Christian Klepp  24:53 Absolutely, absolutely. And we’re working with a client that’s that’s an that’s in tech right now, and the thing that we keep. Talking about is you gotta, you know, yes, of course you’re excited if you start developing more features and what have you right? But look at this through the lens of the user, right? I mean, I can totally relate to your to your situation. I mean, even things like for example, and this is probably like oversimplifying it. But the last update that Instagram did is driving me absolutely crazy. Like, why would you update something your interface that has already been working for the users, and now? Why do you update it so and completely change where the buttons are on the layout so people have to waste time looking for worse, the send button. I mean, you know, it’s just beyond me, right? Apryl Syed  25:45 Yeah, and it’s funny, and they actually, Instagram, for a long while, did a lot of user testing before they would roll out features, and did these limited, I didn’t see any of that necessarily. With this last rollout. Christian Klepp  25:58 No. Apryl Syed  25:59 Apple did a very similar, like their latest update introduced many phone changes in terms of prioritization of, you know, messaging and all that sort of stuff. And it’s like a common we’re finding commonality saying, like, Oh man, I hate this latest I don’t know how many people have said I hate this latest update, and it’s because it’s created too much friction in the process. We need enough friction, but not too much friction. And that balance, in itself, unfortunately, is like the most difficult thing to figure out. And if you’re not talking to your customers, if you’re not talking to people, you will never figure it out, because you’ll be making an assumption. Christian Klepp  26:38 Exactly, exactly. Okay, so we talked about this at the beginning of the conversation, but you mentioned something called a quick win revenue framework. And I know from what you were telling me that that was a little bit controversial to somebody else you spoke to. Apryl Syed  26:55 Yeah. Christian Klepp  26:56 But you know what we are, we are all embracing in the show. You know. Apryl Syed  27:00 Thank you. Christian Klepp  27:00 Not not judgmental. But in fact, the focus here is to help B2B Marketers. In your case, B2B SaaS Marketers to become better and to improve. So if we’re going to focus on this quick win revenue framework, where would you identify low hanging revenue opportunities in under 30 days. So talk to us about that. Apryl Syed  27:24 Yes, well, it sits at this crossroads between marketing and sales, right? And that’s why you’ve got to have such a tight friendship relationship with you know, your sales leaders and your customer success leaders. I think it has to be like such a great ecosystem. So first thing I would do is pull CRM data. I would look at where deals are stalling, you know, I would map the current funnel with actual numbers of where you have people. I would overlay that with like the industry and kind of like the marketing messaging that is created those those types of deals. And kind of look at that from the lens of, okay, here’s what we’re creating, and here’s what sales is able to close easily. Here’s what’s really lagging and taking a long time in the funnel. And it’s not to say that, like, longer is better than shorter, because, like, an enterprise deal takes longer to close than a SMB (Small and Medium-sized Business) deal. So the answer isn’t always that the SMB deal is better, but looking at that and saying, Is there anything here that is that is giving me an indicator of something I can improve on? Can improve on. So that would be, you know, number one, go through that audit, take a look at the data, see what you’ve been producing from a marketing standpoint so far, and then say, is there anything that we should be testing to do differently better? You know, what are your hypotheses that you want to go out and you want to prove with some AB testing, two look at conversion killers, right? That’s either messaging, follow up, timing or onboarding friction, some sort of friction in the process. Friction could be a form fill too it could be, you know, too heavy, too long of landing page, I would look at every single detail and way that people are coming in through the funnel and say, are we doing anything to kill conversion and sometimes, and I’ve experienced this with one brand that I’m working with, and we have an agency that’s also in there that’s doing some ad performance, and they’re getting industry well above industry standard rates. And I asked the agency, because I’m sitting in kind of like my fractional executive role, and I said, Tell me out of your entire client, raw. Stair. Where does this client sit? And they said, Oh, at the top, best performing client we have, you know what that signaled to me? They’re comfortable. They’re getting great results. They’re not trying to improve anything. They’re just trying to hold the fort down and just keep getting these great results because they think that’s a place of safety. Christian Klepp  30:23 Stop rocking the boat Apryl. Apryl Syed  30:26 I know, I know, but I look at that and say, You’re not trying hard enough. You’re not examining right and going through the funnel and looking for all the tweaks and looking for. Christian Klepp  30:36 What can it improve? Apryl Syed  30:37 Can it be improved? You’re not trying to do any of that. And in fact, I’m adding that to you. I’m adding those things. I’m asking for those things, just because I come from that space and saying, like, Hey, we should be pushing here. We should be pushing here. We should be they don’t want to push. And they’re slow, slow, slow to react. And what’s going to happen is it’s going to earn them a change out in agency, right? Because they’re not pushing. Now, unfortunately, what I think is, if that was happening, obviously was happening before I was involved this customer, they thought they’re getting, they’re getting, like, six to one on their spend. That’s fantastic. We should be happy, right? And I’m like, no, no, no, I’ve pushed, I have pushed that envelope before. I’ve seen, you know, 14% conversion on landing pages. I’ve seen 49% conversion on landing pages. When you get it really right, you should always be pushing and pushing and pushing that envelope. So really diagnose and look, are there friction killers in those processes, and where can you be improved? And it is not like, I’m getting results good enough, so let me stop. It’s not stop because that might be one of your levers to really, really get quick wins, because you could tweak something and then even tip the scale further. And who doesn’t want a big win like that? The other thing is, like, I think there’s I look at I look at email sequences and messaging. I look at every single message that we’re sending a customer through the process, through their buying journey. You know, for one client, I basically call it a customer journey map, which a lot of people don’t do anymore, but my journey map is from the moment that they hear about you, all the way through buying, how do we touch them? What do we touch? And then from buying through that sales cycle, what is that like? And the reason why I map that out is because when you do and you put the different sections, you can kind of say, well, this is the process today. What would we like that process to be? And you will find in every single one of these customer journey maps that I’ve done, five to 10 areas where you’re like, instantly know, you instantly know the experience you could be providing better. I did this for one client, and we uncovered, like, the review process for their terms and conditions. On average took like, 10 days with an average back and forth between their lawyers and our lawyers, maybe 15 times that is that a desired customer experience? No, that’s a friction creator, which could be a deal killer, could be a deal staller. So what does that desired experience look like? What should we aim to get to? How are we going to do that? What should we test first? That’s just an example of one that might be in there. So look at everything. Then it becomes, you know, build exactly what you think you’re going to test, go and launch and measure those tests. And you don’t need this to be six months, right? Depending on how much data you’re getting through, it might only take you two weeks of data. It might take you a week of data on these experiments and levers that you’re going through so figure out how long you need to run the experiment for. Run that experiment, measure those changes, and then either permanently implement the change or make changes right and refresh and do another test. Christian Klepp  34:24 Wow, that was quite the list. And I’m sure you’ve, you’ve had, like, as you, as you’ve mentioned, you’ve had pushback for, you know, some of this, for this process, because it’s it. It makes teams uncomfortable, right? But I think the point is, you know, everybody says, right, change is uncomfortable. Improvement is uncomfortable. Uncovering ways to make things better should make you feel uncomfortable, right? Apryl Syed  34:53 So true, so true. And I always, I always think like, if you’re uncomfortable and you’re feeling like. A maybe, I don’t know all the answers here. It’s a really good place to be, and that’s where real growth happens. That’s where real change happens. Christian Klepp  35:06 Yeah. So I did have one follow up question for you, Apryl, like, you know, based on this framework that you’ve just proposed, like, How often would you recommend? And I know it depends, but how often would you recommend teams to continuously monitor some of these, some of these attributes and these factors that you’ve that you’ve brought up in the past couple of minutes. Apryl Syed  35:27 Gosh, I think it is very dependent on the data that’s coming through. If you were experiencing problem in an area, deep dive in there and uncover it. Kind of do that audit and analysis and create some tests that you could run to improve it. But as a measure, the customer journey map, for example, for existence, I think that’s a living, breathing document. I think we should look at it quarterly. We should update it with the experiments and the learnings and the new things that we’ve implemented permanently so that we can track how that experience is going and make sure that it’s our desired experience that we’re putting out there. Because I think a lot of times stuff just happens and it’s not our desired experience, but we kind of think like, oh, well, this is the process, the way it has to be, or, you know, so and so said that it has to be three days. So it’s three days, and it’s like giving you a moment to step back and be like, Why could we do it different? Could we do it better? Could we do it in two days? I don’t know. Could we do it in one and, you know, so I think as often as that customer journey, when updates happen, put those updates in their document. It, look at it, say, like, what’s next on the list should always be improving. When you get to the point where you don’t have any more insights in there, and you think it’s oiled up in the best that you could possibly do it, bring some customers in, bring some customers in to look at it and get their opinion. Ask them about it. It’s a great point to now be in survey mode and ask some questions about where you might have conflicts internally, or where you just aren’t sure where to go. So I think that when it comes to like email sequences, and remind you know like those provide provides, messaging, emails, one thing landing pages, like, I think your landing page just should be in a constant AB turnaround. Every time you have five to 10,000 people hitting a landing page, you should be trying to tweak that message to see if you can make it better. Message, layout, colors, all of the kind of industry standards there, you should be constantly trying to tweak that. If you’re not using landing pages and you’re sending stuff to a page, you should try landing pages so it’s just the constant improvement of those email sequences kind of, kind of, I feel, I feel they should be similar. I feel like you’ve got to examine those on a pretty regular basis, maybe it’s monthly, and kind of determine which messages are you going to trade out. I’m doing a pretty big switch out right now for, you know, an SMB app that’s, you know, selling to other businesses. So it’s a B2B, SaaS company, and we are revising all of their messaging, going through every single one, but trying to create, like a very purposeful journey now where there hasn’t been necessarily one before. And what I just said to one of the leaders yesterday is like, this is version one of what will be probably 10 before we’re done with this iteration. Because every single time we see the data and see how people are moving through the flow, we’re going to we’re going to see that those things that we didn’t consider, there’s going to be broken pieces. Like, don’t be in a position of thinking that any of your marketing is final ever. That’s a good position to be in. It’s never final. I think about this for websites as well. Like people like, oh, we go through our big website refresh, we get the website done, and then now we don’t have to touch the website. Oh, you should be, like, touching the website all the time. Experiment with the messaging on the homepage. Like to think that you got the messaging right the first time. I wish, I wish, and I’ve been in this industry for more than 25 years, I wish, and I’m considered, considering, considered a messaging, you know, wizard. Sometimes, it sometimes takes five or six tries before you get that like, nailed one, and that’s because persona, you know, it’s like how the person is feeling. It’s the emotional draw, and it’s the features, the problem of the pain and all of that coming into one like, I wish, I wish there was an AI tool that could get that right. But it’s not, they’re not. Christian Klepp  40:00 I haven’t found one yet. Apryl Syed  40:01 Yeah. You know, it’s only through really, really overworking that message and seeing the data come in that you kind of like, finally get to maybe a place that’s good, and then guess what? Your persona changes or something happens to so. So don’t ever think of it as, oh, to set it and forget it, it. It should be like it. And there’s also, like, Don’t tweak it too fast that you don’t have enough data coming through. Like, that’s also, I can, I can see that being a message, but have enough data, review that data on a regular basis, make some changes, test it. It’s like little incremental tests and learn. So that’s going to be kind of like it’s either in that category, which is like, test and learn, test and learn, test and learn constantly tweaking, or a quarterly or an annual kind of review. Christian Klepp  40:54 Fantastic, fantastic. Apryl. This was such a great conversation. Thank you so much for your time and for sharing your expertise and experience with the listeners. Um, please. Quick introduction to yourself and how folks out there can get in touch with you. Apryl Syed  41:07 Well, my company is Apeture Codex. Best way to get in touch with me is just Apryl Syed at LinkedIn. That’s where I’m most active, is on LinkedIn, and you can book an appointment with me right off of my LinkedIn. And so that’s like the best, best way to find me out there. Christian Klepp  41:27 Fantastic, fantastic. And we’ll be sure to drop those links in the show notes once the episode goes live. So Apryl, once again, thanks so much for your time. Take care, stay safe and talk to you soon. Apryl Syed  41:38 All right. Thank you so much, Christian. Christian Klepp  41:40 Okay, Bye, for now. Apryl Syed  41:41 Bye.

Dzień dobry – guten Tag! Maike und Ingo kommen an ihre Grenze – und ein Stück darüber hinaus. Die beiden reisen mit der Stadttore-Linie RE4 ins polnische Stettin. Wie es mit der Verständigung über Sprachbarrieren hinweg klappt und welche freundliche Geste der Fahrgäste sie überrascht hat, verrät Kundenbetreuerin Silke Schulz. Am Heumarkt in Stettin fühlt sich Maike gleich wie zu Hause, bevor das Podcast-Team den roten Faden findet und ihm zu den Geheimnissen der geschichtsträchtigen Stadt folgt. Am Schloss der Pommerschen Herzöge bleibt Ingo glatt der Mund offenstehen. Da gibt es einen alten Fluch, und dann blüht ihnen auch noch die Höchststrafe. Ab in den Kerker! Nach diesem Schreck winkt eine Stärkung mit den herzhaften Stettiner Pasteten. Und wohin geht`s danach? Zu einem „Eisberg“ mit besonderem Klang! Hört, wer von beiden besser polnisch spricht, was Maike und Ingo im „Stadttor mit Geschmack“ erleben und wohin sie ihre Zeitreise führt.Links zu den Partnern:https://visitszczecin.eu/de https://filharmonia.szczecin.pl/de https://zamek.szczecin.pl/de/ https://wedelpijalnie.pl/en

Mason Cosby connects with Putney Cloos, the Chief Marketing Officer at Bombora, to clarify exactly what intent data is and how revenue teams should use it. While many marketers view intent data strictly as a tool for sales prioritization, Putney argues that the use cases go much deeper. She explains the difference between first-party data collected from owned properties and third-party data gathered from across the web.ㅤA major focus of the conversation is data provenance. Putney highlights a critical but often overlooked question: Does your data provider actually have the right to use the data they sell? She details how Bombora uses a proprietary tag within a data co-op to ensure consent is explicitly granted for intent tracking.ㅤMason and Putney also explore how to operationalize this data without buying a massive, all-in-one platform. They discuss the "unbundled" ABM stack, where data is ported directly into the tools teams already use, such as CRMs, CDPs, or advertising platforms like The Trade Desk and LinkedIn.ㅤGuest BioPutney Cloos is the Chief Marketing Officer at Bombora, where she leads marketing and communications to expand the company's "Data Co-op" and intent data solutions. She previously served as CMO at Cision, where she built the company's first account-level data strategy. Putney also spent nearly a decade at American Express, rising to Vice President of Commercial Demand Generation and creating Amex's first commercial demand-generation and ABM capabilities. She holds an AB from Harvard College and an MBA from the Kellogg School of Management.ㅤWhat We CoverDefining intent data: Putney clarifies that intent is behavioral data showing when a company is actively researching a solution, distinct from static firmographic data.First-party vs. third-party signals: The difference between tracking known visitors on your own site versus capturing research behaviors across the broader B2B internet.The importance of data rights: Why marketers must ask if their provider has explicit consent to use data for intent purposes, rather than relying on repurposed bidstream data.Reducing customer churn: How customer success teams use intent signals to spot when current clients start researching competitors before the renewal conversation happens.Informing the product roadmap: Using search behavior to identify unmet market needs or features that prospects are actively seeking.The unbundled tech stack: How to deploy intent data directly into existing platforms like Salesforce, HubSpot, or programmatic ad exchanges without needing a standalone ABM platform.Start simple: Putney's advice to pick one specific use case—like sales prioritization—and master it before adding...

*

Parallel zu den Jacksons verfolgt Michael seine Solokarriere – und dann kommt „Thriller“. Ein Album, das alles verändert. Michael wird zum Weltstar. Und das bedeutet: Ab jetzt gibt es keine Privatsphäre mehr. Alles wird kommentiert. Und alles wird zum Gegenstand von Spekulationen und Gerüchten. Schläft er in einer Druckkammer, oder kauft er die Knochen des Elefantenmenschen…? Hier findet ihr den Saily Code: https://saily.com/menschExecutive Producer: Ruben Schulze-FröhlichRedaktion: Heiko Behr, Mira Dönges, Tamara AllinHost: Mira Dönges, Heiko BehrSounddesign: Felix StäbleinProduktionsleitung: Josephine AleytBei „Mensch!“ erzählen Mira und Heiko die spannendsten, bewegendsten und überraschendsten Geschichten aus dem echten Leben unserer Lieblingspromis – authentisch, nahbar und voller Emotionen. Von Taylor Swift und Kanye West über Hape Kerkeling und Dieter Bohlen bis hin zu Heidi Klum und Madonna. Hosted on Acast. See acast.com/privacy for more information.

Cal has disappeared so it is up to AB and Chip to run this pod

本集「閱讀經典」單元，同樣邀請第一季導讀人朱宥勳老師對談，時序從 50 年代跨入 60至70 年代，導讀最「做自己」、也最飽受爭議的現代主義大師——七等生。 朱宥勳老師選讀了三篇經典，分別是〈我愛黑眼珠〉、〈AB 夫婦〉與〈跳遠選手退休了〉。節目將討論七等生的生命軌跡與創作動機；同時一窺在那個當時的時空背景下，台灣文壇兩大主義（現代主義與鄉土文學）是如何交鋒、跌宕與循環。 . 來賓｜朱宥勳（作家） 主持｜Linda（誠品職人） . ▍ 邊聽邊讀 七等生全集 https://esliteme.pse.is/8mkd2w 七等生《我愛黑眼珠》https://esliteme.pse.is/8mkd5l 七等生《僵局》https://esliteme.pse.is/8mkd99 ▍ 延伸閱讀 王禎和 https://esliteme.pse.is/8nwxwy 施明正 https://esliteme.pse.is/8nwy5q . ⭓ 誠品聯名卡︱天天賺回饋 活動詳情

AB sits down with Gordon Thomson, Cisco's SVP EMEA Sales and EMEA President, for a great talk about The EMEA Moment, sovereign critical infrastructure, and how AI is a tool that is helping unlock human potential.

Nach der Schule geht's in den Hort. Ab dem Schuljahr 2026/27 soll das in ganz Deutschland schrittweise möglich sein. Denn dann gilt der Rechtsanspruch auf Nachmittagsbetreuung an Grundschulen. Doch gut gemeint ist noch nicht gut gemacht. Neubig, Magdalena; Schaar, Jörn www.deutschlandfunk.de, Hintergrund

Episode 102: Slap Shot With Special Guest: Megan Craig (Associate Professor of Philosophy and Art, Stony Brook University) The gloves are off this week on Philosophy in Film as the gang faces off with George Roy Hill's bruising classic, Slap Shot! Set in the fading mill town of Charlestown, the film follows player-coach Reggie Dunlop, played by Paul Newman, as he tries to save a failing hockey team by leaning into spectacle, violence, and the sudden popularity of the Hanson Brothers. What begins as a desperate bid for ticket sales spirals into a question about integrity, entertainment, and what happens when winning becomes secondary to drawing blood. At centre ice, Craig drops the puck with Producer's Notes (), while Alain takes out some teeth with the Beauclair Synopsis (). In Philosopher's Corner, Chris digs into the film's storied history and stitches connections to our hometown. The gang heads to the penalty box for the Round Table () to consider the ethics of aggression, masculinity on ice, and the simmering class tensions bubbling beneath the boards. Reviews sound the final buzzer as we tally the hits, the heart, and whether Slap Shot earns its place in the hall. As always, we explore the philosophical and non-philosophical aspects of the film, because when the crowd wants a fight, someone still has to decide what the game is really about.

AB'nin savaşın dördüncü yılında hayata geçirmek istediği Rusya'ya karşı yaptırım paketi, Macaristan'ın vetosuyla karşılaştı. CHP kurultay davasının Aziz İhsan Aktaş davasıyla birleştirilmesi talep edildi.Bu bölüm Sneaks Up hakkında reklam içermektedir. Türkiye Basketbol Federasyonu ile imzalanan anlaşmayla Sneaks Up, Basketbol Milli Takımlarının resmi sponsorları arasına katıldı. Ayrıntılı bilgiye buradan ulaşabilirsiniz.

Çavuşesku'nun Termometresi'nde Ekin Keleş moderatörlüğünde Prof. Dr. Burak Bilgehan Özpek ve İlkan Dalkuç ile Savunma ve Uluslararası Ticaret Hukuku Uzmanı Av. Şafak Herdem "Özel Askerî Operasyon"un 4. senesinde Rusya'nın Ukrayna'yı işgalini ve Millî Dayanışma, Kardeşlik ve Demokrasi Komisyonu'nun raporunu ve raporun ardından Bahçeli'nin Öcalan ve KCK açıklamalarını değerlendiriyor.Av. Şafak Herdem'in Linkedin Sayfası:https://www.linkedin.com/in/safak-herdem00:00 Giriş00:34 Daktilo1984 işgalin 4. yılında da Ukrayna'nın yanında04:20 Rusya'ya yaptırımların (2014'ten bugüne) ne kadar etkisi oldu?08:50 Yaptırımlar hangi ülkede hangi alanlarda ne ölçüde etkili olabilir?10:20 Rusya demokratik bir toplum olmadığı için yaptırımlar ülkeyi değil halkı etkiledi denilebilir mi?12:50 Biden gitti, Trump geldi; Rusya'ya ABD yaptırımlarında farklılık var mı?15:30 AB, yaptırımların dolanılmasını önlemek için proaktif yaklaşımlar sergiliyor16:05 Rusya'ya yaptırımların dolaylı hedefi: (ava giderken avlanayazan) Türkiye19:10 Türkiye'nin Rusya'dan savunma sanayisinde giderek uzaklaşmasının sonuçlarına dair21:40 S-400'ü aldık ama bi' sor, niye aldık? (sadece yanlış cevaplar)24:50 Yaptırımlar Rusya'nın savaşı sona erdirmesinde nasıl etkili olabilir?27:30 Millî Dayanışma, Kardeşlik ve Demokrasi Komisyonu'nun raporuna değil hapse bak, sokağa bak32:10 Bahçeli, herkesin farklı anlayabileceği sözleri niye söylüyor?34:50 Çözüm Süreci=Muhalefeti toptan, topyekün, tamamen susturma süreci37:32 Olaylar hep Kürtler üzerinden gerçekleşiyor ama Kürtlerden bahseden yok39:02 İsrail ile PKK arasında bir yakınlık, çok yakınlıktan bahsedenler vardı...40:20 İktidar ve muhalefetin mevzi kazanma mücadelesinde Kürt kamuoyuna "verilecekler"41:52 Bu süreçte insanların sağduyusu beni memnun etti (ulusalcı, milliyetçi kesimlerin BİLE)42:54 Süreç'in saha realitesiyle ilgisi yokken neden Bahçeli sürekli jeopolitik diyor?45:40 Münih Güvenlik Konferansı'nın dayattığı gerçeklik Türkiye'ninki değil ABD'ninki48:20 Öcalan'ın paradigması: 10 yıldır muhalefetle yürüdük, bir de iktidarla yürüyelim49:00 Çözüm Süreci, Immoral Tales'in (1973) sanat filmi olarak değerlendirilmesi gibidir (manası çok derin)55:20 Benim bir şey söylememe gerek yok: Türkiye, Gazze Barış Kurulu'nda İsraill'le aynı masada01:03:20 İmralı'ya CHP gitmeyince hiç kimse gitmemiş oluyor01:07:50 Demokratikleşme herkes için iyi bir şeyse niye..? Ayrıcalıklardan yararlanmak için bu kanala katılın:https://www.youtube.com/channel/UCWyDy24AfZX8ZoHFjm6sJkg/joinBizi Patreon'dan Destekleyin

Jeden Mittwoch beantworten Hannes und Niclas eine Frage aus der PAPAS-Community.Hier geht es um kleine Unsicherheiten, mal um große Sorgen und manchmal um ganz alltägliche Situationen, bei denen man sich fragt: Geht das nur mir so? Spoiler: tut es nicht. Wir sitzen nämlich alle im selben Boot.Kleine große Fragen ist ein bisschen wie Domian für Eltern. Ab jetzt jeden Mittwoch.Schickt uns doch eure Fragen gerne hier unter der Folge, per Mail oder auf Instagram.Wir freuen uns von euch zu hören < 3 Hosted on Acast. See acast.com/privacy for more information.

Send a textAre you an Amazon seller feeling like you're just spinning your wheels? This video introduces the 4 Pillars of Amazon Growth Masterclass for Amazon sellers, designed to help you scale your business beyond basic optimization tactics. We'll explore effective strategies for how to sell on amazon, including advanced amazon seo techniques and amazon fba insights, ensuring your amazon listing stands out. Learn how to achieve real sales growth and move past common PPC pitfalls.Learn how AB testing improves click-through rate, conversion rate, and overall Amazon sales performance. This video breaks down how to test main images, A+ content, pricing strategy, and product detail pages using real Amazon data. Understand how SEO, PPC, catalog merchandising, and design work together to increase ranking and drive consistent growth.Amazon Dynamic Pricing Strategy: https://myamazonguy.com/amazon-product-launch/amazon-dynamic-pricing-strategy/Amazon FBA Product Launch Pricing Strategy: myamazonguy.com/price--------------------------------------------------------------------------Want free resources? Dowload our Free Amazon guides here:Amazon Proft Margin Defense 2026: https://hubs.ly/Q042trRH0Amazon PPC Guide 2026 is here!: https://bit.ly/4lF0OYXAmazon SEO Toolkit 2026: https://bit.ly/4oC2ClTAmazon Seller Strategy Report 2026: https://bit.ly/3YN1RME2026 Ecommerce Website & SEO Readiness Checklist: https://hubs.ly/Q040Jg0M0Amazon Crisis Kit: https://bit.ly/4maWHn0Timestamps:00:00 – Why Sellers Feel Stuck on Amazon01:02 – The Four Pillars of Amazon Growth04:00 – Why Click-Through Rate Drives 18% of Sales05:43 – Dynamic Pricing Strategy Explained09:05 – Coupons vs Promotions vs Deals15:55 – Lightning Deals and Seasonal Strategy22:42 – Design Mistakes That Kill Conversion23:01 – Main Image CTR Tactics26:51 – Infographics That Prevent Bad Reviews32:35 – A+ Content That Increases Conversion35:19 – How SEO, PPC, Design, and Pricing Work Together38:18 – Rufus and AI Impact on Amazon Search46:31 – Using Amazon AB Testing for Conversion Data54:27 – Common Seller Mistakes with Growth Strategy56:36 – Next Design Changes After Main Image58:39 – Fundamentals Every Amazon Seller Must Learn________________________________Follow us:LinkedIn: https://www.linkedin.com/company/28605816/Instagram: https://www.instagram.com/stevenpopemag/Pinterest: https://www.pinterest.com/myamazonguys/Twitter: https://twitter.com/myamazonguySubscribe to the My Amazon Guy podcast:My Amazon Guy podcast: https://podcast.myamazonguy.comApple Podcast: https://podcasts.apple.com/us/podcast/my-amazon-guy/id1501974229Spotify: https://open.spotify.com/show/4A5ASHGGfr6s4wWNQIqyVwSupport the show

We talk about skater Alysa Liu and how someone local did her unique hair, and we congratulate AB on filming her first wedding over the weekend.

AB sits down with Zeus Kerravala, founder of ZK Research and a leading technology industry analyst, for a great conversation on the role analysts play in the world of tech, trends and innovations related to AI infrastructure, why the network is critical to AI workloads, and more.

AB sits down with Gordon Thomson, Cisco's SVP EMEA Sales and EMEA President, for a great talk about The EMEA Moment, sovereign critical infrastructure, and how AI is a tool that is helping unlock human potential.

Der After-Match-Talk & Turmfunk- Highlights zum Spiel SSV Jahn Regensburg – SV Wehen Wiesbaden (2:1) in der 3. Liga-Saison 2025/26. _ Ps. Komm in unseren Discord-Channel zum Diskutieren: https://discord.gg/b5SzkBcX6b oder klassisch im https://jahnground.de Forum Pps. Unterstütze das Podcast-Projekt finanziell: https://1889fm.de/unterstuetzen/ (Ab 5€ Unterstützung gibt es ein kleines Stickerpaket zugeschickt) Alle Nachrichten zu unserem geliebten Jahn erhältst du ab sofort auch in unserem WhatsApp-Kanal: https://whatsapp.com/channel/0029VbAu0GA65yDDLd3Fj701

Tom Petrini, Aaron Blackerby, and Zach Montana recap the Spurs' two decisive Austin wins, highlighting the team's pace, transition scoring, and physical play. They discuss key performances including Stephon Castle's defense on Devin Booker, De'Aaron Fox's steady production, Wemby's rim attacks and playmaking, and Julian Champagnie's impact as a starter. The hosts share Austin Week fan and community moments, praise the Spurs organization's engagement efforts in Austin, and break down how the Spurs' backcourt of Fox, Dylan Harper, and Castle are finishing efficiently at the rim and fitting together. They make the case for Victor Wembanyama in end-of-season awards conversations, debating narrative, voter behavior, and the balance of offense vs. defense. The episode closes with a preview of the upcoming Rodeo Road Trip games vs. the Pistons, Raptors, and Nets, including matchups, style contrasts, and weekly record predictions.00:00 Welcome Back to the Silver & Black Coffee Hour (Austin Edition)00:39 Bold Prediction Victory Lap: Spurs Hit 40 Wins in Austin01:54 Austin Week Vibes: Fans, Family, and the I-35 Series Energy05:17 Game 1 Breakdown: Spurs Run the Suns Off the Floor09:03 Game 2 vs Kings: Wemby's ‘Just Okay' Dominance + Defensive Focus12:33 Legends, Media Row, and Community Moments Around Austin Week17:48 High Praise Segment Begins: Spurs' Austin Engagement & Marketing Blueprint20:38 Zach's Numbers Corner: Fox/Harper/Castle Backcourt Fit Is Real25:40 AB's High Praise: Julian Champagnie's I-35 Series Impact29:34 Questions of the Week: Making the National Case for Victor Wembanyama31:45 Wemby's MVP Case: Value on Both Ends + Award Trifecta Talk32:37 Narrative vs Stats: Late-Season MVP Push, TV Spotlight & Voter Fatigue37:14 Advanced Numbers Breakdown: On/Off, Offensive vs Defensive Impact, True Shooting39:05 “Prevo Fatigue” Debate: Why Voters Should Judge the Present (and Wemby's Strength)43:04 Rodeo Road Trip Preview: What to Learn in the Final 26 Games45:02 Pace as Identity: Playing Faster Without the Turnovers50:40 Game 1 Preview — Pistons: Physical Dogfight, Paint Battle & Cade Matchups55:53 Game 2 Preview — Raptors: Trap-Game Vibes, Length on Defense & Schedule Edge58:59 Game 3 Preview — Nets: Stack Wins, Back-to-Back Concerns & Take Care of Business01:00:55 Weekly Record Predictions + Closing Thanks, Fanbase Expectations & Enjoy the Ride

Mit dem Ende des Kalten Krieges sanken die Verteidigungsausgaben in den westlichen Staaten deutlich, in Europa bis 2015 im Schnitt um rund 25 Prozent gegenüber 1990. Diese „Friedensdividende“ wurde jedoch nicht genutzt, um Staat und Abgabenlast zu reduzieren, sondern um den Wohlfahrtsstaat auszubauen. Angesichts der geopolitischen Zeitenwende liegt es auf der Hand, dass wir diese Entwicklung umkehren müssen. Die Politik müsste Sozialausgaben zugunsten von Verteidigung und Investitionen in die wirtschaftliche Basis kürzen. Doch damit tut sie sich offenkundig schwer. Stattdessen setzt man darauf, das staatliche Budget auszuweiten, um beides zu haben: den Sozialstaat und die höheren Verteidigungsausgaben. Das passt zur historischen Erfahrung: Guns-and-Butter-Rüstung obendrauf, während der Sozialstaat bleibt. Die Folge sind mehr Schulden und mehr Steuern – und das dauerhaft.Europa diskutiert über gemeinsame Schulden und neue „Safe Assets“, um Aufrüstung, Transformation und geopolitische Herausforderungen zu finanzieren. Doch wie sicher sind Staatsanleihen wirklich, wenn Staaten unter Druck geraten? Die aktuelle Forschung zeigt, dass ausgerechnet in Kriegen und in Krisenzeiten Staatsanleihen historisch häufig massive Realverluste für Gläubiger und Sparer gebracht haben.Im Expertengespräch erklärt Dr. Christoph Trebesch, Professor für Makroökonomie an der Universität Kiel und Direktor der Kiel Initiative in Geopolitics and Economics am IfW Kiel, die historischen Muster von Aufrüstung, die langfristigen fiskalischen Folgen und die Frage, ob Europa sich eine Guns-and-Butter-Politik überhaupt leisten kann.HörerserviceAnalyse Are Government Bonds Safe in Times of War and Pandemic?: https://is.gd/0qrHzV Studie Guns and Butter: The Fiscal Consequences of Rearmament and War: https://is.gd/NrTOgr beyond the obviousNeue Analysen, Kommentare und Einschätzungen zur Wirtschafts- und Finanzlage finden Sie unter think-bto.com.NewsletterDen monatlichen bto-Newsletter abonnieren Sie hier.RedaktionskontaktWir freuen uns über Ihre Meinungen, Anregungen und Kritik unter podcast@think-bto.com.Handelsblatt – 2026 beginnt rasant. Umso wichtiger ist fundiertes Wissen. Wenn Sie das ganze Jahr über gut informiert sein wollen, haben wir ein besonderes Angebot für Sie: 40 Prozent Rabatt auf ein Handelsblatt-Jahresabo – gedruckt oder digital. Ab 4,79 € pro Woche erhalten Sie klare Fakten, exklusive Hintergründe, starke Meinungen und wertvolle Impulse – damit Sie wirtschaftliche Entwicklungen noch besser einordnen können.Sichern Sie sich den Rabatt bis zum 23.02.2026 unter ⁠handelsblatt.com/wissen2026.WerbepartnerInformationen zu den Angeboten unserer aktuellen Werbepartner finden Sie hier. Hosted on Acast. See acast.com/privacy for more information.

Éric de Kermel est romancier. Il a vécu des années importantes de sa vie au Maroc, dont une bonne partie de son enfance. Après une première participation l'an dernier à Zeteo, où il nous avait parlé de son roman L'archipel de Claire, il est venu présenter son nouveau livre À la découverte de l'Islam et des musulmans, qu'il a co-écrit avec son amie Reem Yasmina Laghrari.Reem Yasmina Laghrari est docteure en pharmacie. Née aux États-Unis, elle vit et pratique la médecine au Maroc, le pays de ses racines familiales. Passionnée par les questions culturelles et religieuses, elle a déjà écrit Les Prophètes à la lumière du Coran et de la Bible, livre publié en 2024 et préfacé par Éric de Kermel.Éric de Kermel et Reem Yasmina Laghrari partagent le même rêve : établir des passerelles entre les spiritualités, et rappeler les liens très forts qui rapprochent les grandes religions. C'est le cas pour les trois religions monothéistes, le Judaïsme, le Christianisme, et l'Islam, et particulièrement de ces deux dernières.D'où le projet de ce livre, né d'une discussion amicale lors d'une de leurs rencontres au Maroc : Écrire tout ce que nous partageons, tout ce dont nous héritons les uns des autres, souvent sans le savoir.Éric de Kermel et Reem Yasmina Laghrari ne sont ni théologiens, ni historiens. Leur idée, c'est plutôt celle de révéler le visage d'un Islam lumineux, joyeux et tolérant, et de démêler ce qui repose sur les coutumes ancestrales, les enseignements religieux tirés du Coran et la Tradition prophétique.Avec profondeur, gravité parfois, mais aussi humour et légèreté, ils nous donnent un avant-goût d'un livre qu'ils ont écrit sous le format original d'un Abécédaire riche d'une centaine de thèmes. Ils abordent des sujets aussi variés que ceux, étonnants, de l'abeille ou de l'écologie. Ils évoquent la symbolique de la langue arabe, le lyrisme des arts et la mystique du soufisme. Ils parlent de la femme, de la notion de Dieu, de l'importance du désert. Ils évoquent aussi les questions du voile, de la charia et du djihad. Ils nous disent encore l'importance, dans l'Islam, de la figure de Marie de Nazareth.Avec Éric de Kermel et Reem Yasmina Laghrari, nous découvrons une nouvelle fois à quel point les grandes sagesses et les spiritualités se rejoignent. Leur regard contemporain est à la fois réaliste et tendre. Rempli d'espérance, il est porté par un parfum oriental à la fois mystérieux, poétique et lumineux.Au moment où nous mettons cet épisode en ligne, nous apprenons d'Éric de Kermel le décès de son amie Leila Shahid, qu'il cite au cours de cet épisode. En quelques lignes, voici ce qu'Éric a tenu à nous dire : « Leïla Shahid était mon amie. Elle était très sensible à la démarche que nous portions avec Reem et avait salué la parution de notre livre. Je lui dédie ma participation à cet épisode. »Pour lire À la découverte de l'Islam et des musulmans, le livre co-écrit par Éric de Kermel et Reem Yasmina Laghrari, cliquer ici.CE QUI NOUS RAPPROCHE est tellement plus fort que ce qui nous sépareChers amis, chères amies,Chers auditeurs, chères auditrices,C'est ce qui me frappe le plus, à l'occasion de ce nouvel épisode qui met en lumière tant de rapprochements entre le Christianisme et l'Islam, et qui déborde largement la dimension religieuse et même spirituelle de ce sujet, c'est que ce qui nous rapproche est tellement plus fort et plus profond que ce qui nous sépare...Nos regards et nos cœurs sont souvent troublés par le diabolos, le séparateur. Ce qui nous fait croire à ce qui n'est pas, qui créé la dualité, et qui veut séparer ce qui est intrinsèquement et essentiellement uni.Un mystique orthodoxe m'apprenait, il y a quelques mois, à quel point il ne fallait pas tomber dans cette illusion que nous partageons tous. Ne donnons pas de réalité au Royaume du mal et du néant. Seul existe le Royaume Céleste et du réel. Celui du Dieu qui est. C'est à ce Royaume, et à aucun autre, que nous appartenons.L'humanité est composée de peuples différents, répartis sur des zones géographiques différentes qui ont exercent toutes une influence profonde sur nos cultures, nos spiritualités, nos modes de vie, nos rapports au vivant. Et pourtant, nous provenons tous de la même essence, celle d'un seul créateur génial et divin. Nous formons tous les parties d'un même corps.Et cela s'applique non seulement à l'humanité, mais à tout le vivant. Je suis de plus en plus frappé, et émerveillé, de découvrir toutes les connexions et les interactions qui nous relient tous dans un seul corps, une seule création. Les scènes si attendrissantes qui révèlent les émotions des animaux avec nous ou entre eux, que nous aimons tellement voir sur internet, en sont une illustration. Les liens mystérieux qui agissent entre les minéraux, les végétaux et nous en sont une autre. Il y a une infinité de liens cosmiques.Je pense en écrivant ces lignes au Milieu Divin de Teilhard de Chardin. Et je pense aussi beaucoup à la poésie cosmique du Petit Prince de Saint-Éxupéry… Des sujets magnifiques qui feraient certainement de magnifiques épisodes sur Zeteo.Ce corps que forme l'humanité, qui faisait dire à Annick de Souzenelle qu'elle est l'épouse divine de Dieu, nous en avons un avant-goût intense et vibrant dans toutes les communautés qui nous rassemblent. Comme, modestement mais sûrement, pour celle que nous formons ensemble autour de ce podcast, de ses invités et invitées, de ses auditeurs et auditrices, de ses donateurs et donatrices.Il fait vraiment gris dehors, avec ces tempêtes et ces pluies qui se succèdent inlassablement depuis le début de l'année. En même temps, il fait tellement beau entre nous, avec tout l'amour et toutes les beautés qui nous rapprochent !Tout près les uns des autres,Guillaume Devoud--------------    Pour soutenir l'effort de Zeteo, podcast sans publicité et d'accès entièrement gratuit, vous pouvez faire un don. Il suffit pour cela de cliquer sur l'un des deux boutons ci-dessous, pour le paiement de dons en ligne au profit de l'association Telio qui gère Zeteo.Cliquer ici pour aller sur notre compte de paiement de dons en ligne sécurisé par HelloAsso.Ou cliquer ici pour aller sur notre compte Paypal.Vos dons sont défiscalisables à hauteur de 66% : par exemple, un don de 50€ ne coûte en réalité que 17€. Le reçu fiscal est généré automatiquement et immédiatement à tous ceux qui passent par la plateforme de paiement sécurisé en ligne de HelloAssoNous délivrons directement un reçu fiscal à tous ceux qui effectuent un paiement autrement (Paypal, chèque à l'association Telio, 76 rue de la Pompe, 75016 Paris – virement : nous écrire à info@zeteo.fr ).  Pour lire d'autres messages de nos auditeurs : cliquer ici.Pour en savoir plus au sujet de Zeteo, cliquer ici.Pour lire les messages de nos auditeurs, cliquer ici.Nous contacter : contact@zeteo.frProposer votre témoignage ou celui d'un proche : temoignage@zeteo.fr

Katy Faust sees a not-so-hidden thread that connects divorce, gay marriage, IVF, surrogacy, child trafficking, and more: the historically-recent pivot of putting adult desires before children's needs and well-being. Today, we're discussing the history and data behind this disturbing trend and how we can fight back for the most vulnerable.Them Before Us: https://thembeforeus.com/ NEW: Check out our Merch store! https://shop.lilaroseshow.com/Join our new Patreon community! https://patreon.com/lilaroseshow - We'll have BTS footage, ad-free episodes, and early access to our upcoming guests.A big thanks to our partner, EWTN, the world's leading Catholic network! Discover news, entertainment and more at https://www.ewtn.com/ Check out our Sponsors: -Cozy Earth: Better Sleep, Brighter Days - Get the highest quality sleep essentials for 20% OFF at https://cozyearth.com/lila!-Seven Weeks Coffee: https://www.sevenweekscoffee.com Buy your pro-life coffee and Save up to 25% with promo code 'LILA' & get a free gift: http://www.sevenweekscoffee.com-EveryLife: https://www.everylife.com Buy diapers from an amazing pro-life diaper company and use code LILA to get 10% off!-Presidio Healthcare: Healthcare and doctors who share your values. If you're in TEXAS visit: https://www.presidiocare.com/ If you're NOT in Texas, visit: https://www.prolifeproviders.com/00:00:00 - Intro00:02:58 - Katy's background00:08:03 - Katy's “two moms”00:21:26 - The harms of lacking a mother/father00:35:49 - Disturbing LA mansion news00:46:33 - What is Surrogacy?00:55:16 - Do you need a mom and dad?01:06:45 - Any large studies on same-sex households?01:15:05 - Another study01:17:24 - No Fault Divorce01:30:07 - Ab*se data 01:34:22 - Obergefell and Same Sex Marriage 01:45:51 - Epstein

BONUS: From Individual AI Wins to Team-Wide Transformation What happens when the leaders we trust to guide transformation become the bottleneck slowing it down? In this episode, Monica Marquez—with 25+ years in people transformation at Goldman Sachs, Google, and beyond—reveals why the old equation of effort equals success is breaking down, and what leaders must unlearn to thrive in the age of AI. The Leadership Crisis Nobody Trained You For "No one ever really teaches you what it really takes to be a leader. You know what you do really well, but how do you help other people do that too? That's when I realized it comes down to becoming a really good leader."   Monica's origin story captures a universal struggle: being promoted for technical excellence, then discovering that leading people requires completely different skills. She spent her career at organizations like Goldman Sachs, Bank of America, Ernst & Young, and Google realizing that systems weren't built for everyone—and that the real work of leadership is redesigning those systems to unlock human potential. Today, through her company Flipwork, she helps leaders and teams become what she calls "agentic humans"—people who leverage AI to get ahead rather than getting left behind. The Command and Control Trap "Most leadership development still rewards the command and control archetype. The person who has all the answers, the decisive hero. But AI moves so fast that when you think you've fixed something, it changes the next day. Leaders are starting to become bottlenecks."   The research shows the problem clearly: middle management is where AI adoption stalls. These leaders cling to command and control because relinquishing it feels like losing their value. Worse, they have an unspoken fear of managing AI agents—they don't want to be liable for outputs they don't fully control. Monica reframes this: treat your AI tools like an artificial intern, not artificial intelligence. You wouldn't take an intern's first draft and hand it to leadership. You train them, provide context, and finesse the output. The same discipline applies to LLMs. Rewriting the Success Equation "Effort = success is the old equation. That's pre-AI. The new equation is impact equals success. Output equals success, and impact equals worth."   This might be the most important shift leaders need to make. When tasks that took 4 hours now take 30 minutes, deeply conditioned beliefs about work ethic get threatened. Monica sees leaders questioning their worth because they're producing faster. "I was always taught I have to work twice as hard to get half as far," she shares. "Now what used to take me 10 hours, I can get done in 4. Am I not worthy anymore of being a high performer?" The answer is to measure impact, not effort—and that requires rewiring beliefs that may be decades old. Why Individual AI Adoption Doesn't Scale "Teams are using AI as individual contributors, but they aren't using AI in their actual workflows and the handoffs. That's why leaders are scratching their heads, like, why aren't we seeing the ROI bubble up into the team?"   Here's the gap most organizations miss: individuals save an hour or two per day using AI for personal productivity, but the team never sees compounding benefits. The handoffs between team members remain manual. The friction points persist. Monica's solution is "flip labs"—90-day sprints where teams take one critical workflow, dissect it, and rebuild it with AI. Where can AI handle the $10 tasks so humans can focus on $10,000 decisions? Where should humans remain in the loop? IKEA did this with customer service, retraining displaced workers into design roles. Revenue increased without adding headcount. Leading Through Uncertainty "We're humans wired for certainty, but Agile is a system designed for uncertainty. That's where the behavioral psychology comes in—how do you help people move forward despite the uncertainty?"   The fundamental challenge is biological: our brains seek certainty, but the only certain thing now is that change will come faster than we can adapt. Monica works with teams to create psychologically safe spaces for experimentation—AB testing old workflows against AI-augmented ones, measuring outputs, and learning from failures. "Sometimes we learn more from the failures than we do the successes," she notes. The leaders who create permission for testing and learning will pull ahead; those who demand control will become the bottleneck that slows their entire organization.   About Monica Marquez Monica Marquez is a leadership and workplace AI advisor with 25+ years in people transformation. She coined the "returnship" at Goldman Sachs, helped found Google's Product Inclusion Council, and now guides leaders and teams to adopt AI, agile, and inclusion practices that drive results through her company Flipwork, Inc.   You can connect with Monica Marquez on LinkedIn and subscribe to her Ay, Ay, Ay! AI newsletter at themonicamarquez.com.

Professional Series: https://www.patreon.com/luisdelvillardj/shop Follow the Ibiza Sensations playlist on Spotify: www.spoti.fi/3Z6pDkI Hi my friends !! You can still listen for free to the first 250 Ibiza Sensations episodes. From 251 the sets became Premium and you can only a lower quality file for free. The Premium Series offers full qaulity listening plus 2 extra mixes every month and some exclusive Live Streamings. If you join the Premium Series now, you can get more than 150 hours of new mixes, and for only  2 euro monthly or 24 a year. Die ersten 250 Folgen von Ibiza Sensations könnt ihr noch kostenlos hören. Ab 251 Folgen sind die Sets Premium, und ihr könnt nur noch eine Datei in niedrigerer Qualität kostenlos hören. Die Premium-Serie bietet volle Qualität plus zwei zusätzliche Mixe pro Monat und exklusive Live-Streams. Wenn ihr jetzt der Premium-Serie beitretet, erhaltet ihr über 150 Stunden neue Mixe für nur 2 Euro monatlich oder 24 Euro jährlich. Você ainda pode ouvir gratuitamente os primeiros 250 episódios de Ibiza Sensations. A partir de 251, os episódios passaram a ser Premium e você só pode ouvir um arquivo de qualidade inferior gratuitamente. A Série Premium oferece audição em alta qualidade, além de 2 mixagens extras por mês e algumas transmissões ao vivo exclusivas. Se você assinar a Série Premium agora, poderá obter mais de 150 horas de novas mixagens por apenas 2 euros por mês ou 24 por ano. Je kunt de eerste 250 afleveringen van Ibiza Sensations nog steeds gratis beluisteren. Vanaf aflevering 251 zijn de afleveringen Premium geworden en kun je alleen nog gratis naar een bestand met een lagere kwaliteit luisteren. De Premium Series biedt luisterplezier in hoge kwaliteit, plus 2 extra mixen per maand en een aantal exclusieve livestreams. Als je je nu abonneert op de Premium Series, krijg je meer dan 150 uur aan nieuwe mixen voor slechts 2 euro per maand of 24 euro per jaar. Todavía puedes escuchar los primeros 250 episodios de Ibiza Sensations gratis. A partir del episodio 251, los episodios se convirtieron en Premium y solo puedes escuchar gratis un archivo de menor calidad. La Serie Premium ofrece alta calidad de escucha, además de 2 mezclas extra al mes y transmisiones en vivo exclusivas. Si te suscribes a la Serie Premium ahora, puedes obtener más de 150 horas de nuevas mezclas por solo 2 euros al mes o 24 euros al año. Вы по-прежнему можете бесплатно послушать первые 250 эпизодов Ibiza Sensations. Начиная с эпизода 251, эпизоды стали Premium, и вы можете бесплатно послушать только файл более низкого качества. Premium Series предлагает высококачественное прослушивание, а также два бонусных микса в месяц и эксклюзивные прямые трансляции. Если вы подпишетесь на Premium Series сейчас, вы можете получить более 150 часов новых миксов всего за €2 в месяц или €24 в год. لا يزال بإمكانك الاستماع إلى أول 250 حلقة من Ibiza Sensations مجانًا. بدءًا من الحلقة 251، أصبحت الحلقات متاحة للخدمة المميزة، حيث يمكنك الاستماع فقط إلى ملفات بجودة أقل مجانًا. تقدم الخدمة المميزة استماعًا عالي الجودة، بالإضافة إلى مجموعتين إضافيتين شهريًا وبثًا مباشرًا حصريًا. باشتراكك في الخدمة المميزة الآن، يمكنك الحصول على أكثر من 150 ساعة من المجموعات الجديدة مقابل 2 يورو فقط شهريًا أو 24 يورو سنويًا. Join !! Apúntate !!   https://www.patreon.com/luisdelvillardj   You know how important is to be connected so it's time to join me on Social Media! Facebook, Twitter, and Instagram! WEBSITE: http://www.luisdelvillar.com Instagram: https://instagram.com/luisdelvillardj/ Facebook: https://www.facebook.com/LuisdelVillardj Twitter: https://twitter.com/LuisdelVillardj SHOP ONLINE : https://shop.spreadshirt.net/luisdelvillardj/ Itunes: https://podcasts.apple.com/podcast/ibiza-sensations/id521062568 Soundcloud: http://soundcloud.com/luis-del-villar Mixcloud: http://www.mixcloud.com/LuisdelVillar/ Hearthis.at: https://hearthis.at/L6BkT28Z/ Podbean: https://luisdelvillardj.podbean.com/ YouTube: http://www.youtube.com/@IbizaSensationsbyLuisdelVillar Google Podcast: http://bit.ly/2RCu3MZ Overcast: https://overcast.fm/itunes521062568/ibiza-sensations-by-luis-del-villar

Enjoy as Hannah and AB show off their latest hobbies.

California's High-Speed Rail Authority is under fire as a new bill, AB 1608, seeks to allow the agency to withhold investigative records from the public. Authored by Assemblywoman Lori Wilson, the bill claims it's to protect the state, but critics argue it shields vital information about the project's ballooning costs, now estimated at over $100 billion. This move raises transparency concerns, especially given the project's history of delays and financial mismanagement. The inspector general, created to monitor the project, would gain the power to keep internal discussions and personal correspondence private, potentially limiting public oversight of this controversial infrastructure project. With costs skyrocketing and completion dates pushed back, this bill adds fuel to the fire, raising questions about accountability and the future of California's high-speed rail.

In this episode, Nathan with Family Protection Ministries gives an update from the California State Legislature as the bill introduction deadline hits and hundreds of proposals are being filed.He begins with AB 1631, the mandatory kindergarten bill, explaining how it mirrors past efforts and why it raises concerns about expanding state control over families and early education.Nathan also walks through several additional bills that could affect parents and homeschoolers, including:AB 1598AB 1628AB 18AB 1884AB 1914AB 8For each, he outlines what the bill proposes, how it could impact families, and what red flags we should be watching for as the legislative session unfolds.FPM School Choice Articlehttps://fpmca.org/school-choice-week/FPM Website:fpmca.org

Nach der Sommerfrische, dem Prinzen Heinrich und der Rheinsberger Kultur, geht es heute literarisch zu, auf den Spuren von Kurt Tucholsky im Bilderbuch für Verliebte. “Das ganze Glück ihrer großen Liebe” (Kurt Tucholsky) Der subjektive Einstieg Manchmal wird man von Büchern eingeholt und wirft einen voll aus der Realität des Lebens. Ein kleines Büchlein von Kurt Tucholsky hatte das vor etwa zwei Jahren geschafft, als ich die Geschichte mal wieder in die Hand nahm. Ein wenig glücksschwebend angesichts des Wiederlesens mit Claire und Wölfchen beschloss ich. Da wo die glücklich waren, da will auch ich mal hin. Philosophie auch in der Musikkunst Rheinsberg – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Gefühl, Freiheit, Liebe – das alles soll Rheinsberg sein? Besuche ich etwa einen Glücksort und schwelge in Leichtigkeit? Gedacht und getan: Im letzten Herbst. Dass daraus im November ein Spätsommer werden würde, war noch nicht klar. Das Wetter jedenfalls tat sein Bestes, um Städtchen und Literatur ins beste Licht zu setzen. Der Literat Kurt Tucholsky betrat 1912 mit Rheinsberg nicht nur die literarische Bühne, er veröffentlichte damit auch seinen ersten Bestseller. Die reale Geschichte? – Else Weil, auf einer Infotafel am Ratskeller in Rheinsberg – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Ein Jahr davor (oder war es schon 1910?) verbrachte er mit seiner damaligen Verlobten Else Weil ein ähnliches Wochenende – in Rheinsberg. Oder war es genau dieses Wochenende von Wölfchen alias Kurt und Else alias Claire? Else Weil jedenfalls wurde Kurt Tucholskys erste Frau. Die Story Romantik a la Rheinsberg: Obelisk, Postsäule am Triangelplatz – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Die Figuren verschwimmen mit der Realität. Kurt Tucholsky und Else Weil waren um 1910 da. Ihre Alter-Egos Wolfgang und Claire traten erst mit der Veröffentlichung des Buchs im Jahr 1912 auf den Plan. Diese Vermischung hört man des Öfteren auch in den Originaltönen aus Rheinsberg. Die Charaktere geraten immer öfter durcheinander. Aber ohne Kurt kein Wölfchen und Clairchen und ohne Else auch keine Reise mit Kurt. Die beiden Geschichten könnten identisch sein und spiegeln großes Glück. Glück des Moments. Glück des Lebens? Der touristische Influencer? Das wollte ich auch herausfinden. Einige Facts deuten darauf hin. Die Sommerfrische hatte um die Wende zum 20. Jahrhundert dazu geführt, dass immer mehr Berliner begannen das Umland zu entdecken. Tucholsky Porträt im Museum – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Er hatte es zwar nicht beabsichtigt, aber das Büchlein sorgte im Jahr nach dem Erscheinen für einen regelrechten Rheinsberg-Boom. Es soll Sonderzüge gegeben haben, die bis zu 6.000 Menschen an einem Wochenende nach Rheinsberg und an die brandenburgischen Seen brachten. Das schafft heute allenfalls „Lonely Planet“. War Rheinsberg der „Overtourism-Sündenfall“? Wenn ja, hat es das damals 2.000 Einwohner zählende Städtchen wohl verkraftet. Merkpunkt: Die Bahn konnte und könnte sehr viel bewirken in Sachen Tourismus. Damals wurden auch kleine Orte an die Schiene angeschlossen, Mobilität ermöglicht. Das ist heute weitgehend Historie. Glücklicherweise gibt es den Bahnhof Rheinsberg und Verbindungen nach Berlin bis heute. Nur: Wie gehen Touristen heute mit Rheinsberg, Tucholsky und der kleinen wie großen Geschichte um? Den Kernsatz dazu hört Ihr im Podcast von Stadtführerin Jeanette Lehmann. Das Museum Kurt Tucholsky Museum: Dauerausstellung – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Kurt Tucholsky: Den Journalisten, Satiriker, Autoren und Menschen besser kennenlernen. Wer das möchte ist hier an der richtigen Stelle. Es ist ein kleines, aber sehr feines Museum, das sich zudem (auch im Tucholsky Sinne) um schreibende und bildende Künstler:innen mit diversen Veranstaltungen kümmert. Ich bin mit Peter Graf, dem literaturwissenschaftlich-künstlerischen Projektmanager des Museums, durch die Dauerausstellung gegangen, habe das kleine Büchlein, die Erstausgabe von „Rheinsberg“ gesehen, Tucholskys Schaffen kennengelernt und bin dem Menschen Kurt T. und seinem Schicksal begegnet. Das rührt an. Die Erstausgabe: Rheinsberg – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Es ist sicher kein „Claire und Wölfchen Museum“; so wie Tucholsky auch nicht auf den Autoren des Bilderbuchs für Verliebte zu reduzieren ist. Wer nach Rheinsberg kommt, sollte das Kurt Tucholsky Museum im Schloss aber zur Pflichtstation machen. Es lohnt sich. Es gibt ein Kombiticket. Damit kann man die Schlossführung mitmachen und danach auch das Museum besuchen. Machen! Das Schild    Die öffentliche Liebeserklärung an das „Bilderbuch für Verliebte“ findet man in der Straße und Anlage „Am Markt“, gleich gegenüber von Ratskeller und Triangelplatz. Liebeserklärung an eine Stadt – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Tucholsky hat den Text nicht verfasst. Niemand weiß so genau, wie das erste Schild dort hingekommen ist, sagen die Rheinsberger. Fest steht aber, dass das restaurierte Schild vor einigen Jahren vom Verein Stadtgeschichte angebracht wurde. Inzwischen war klar, dass es sich hervorragend als „Insta-Location“ eignet. Ich schließe mich da gerne an: Ist toll für ein verliebtes Selfie. Ach ja, ich war ja alleine dort. Das Café Claire Café Claire – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Da darf man auch alleine hin. Es gibt wundervolle Kuchen, leckere Kaffee- und Teespezialitäten und mittags auch kleine Snacks. Über den Birnen-Schmand-Kuchen hab ich im letzten Podcast schon geschwärmt. In der ersten Novemberwoche 2025 bei 15 Grad draußen in der Sonne zu sitzen, machte mein Glück perfekt. Die Kurt Tucholsky Buchhandlung Kurt Tucholsky Buchhandlung in Rheinsberg – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Er befindet sich in unmittelbarer Nähe des Café Claire in der Schlossstraße. Hier kann man alte Buchkultur schnuppern, moderne Kinderliteratur erleben und natürlich den einen oder anderen Tucholsky-Band mitnehmen. Geht auch als Geschenk. Ich finde der Laden gehört einfach zum Rheinsberg-Erlebnis dazu und das schon seit Jahrzehnten. Die Filme Es gibt zwei „Rheinsberg“ – Verfilmungen. Rheinsberg 1 Der erste stammt aus dem Jahr 1967. Kurt Hoffmann hat ihn in der BRD gedreht. Da ist von der guten alten Zeit die Rede. Eine werbende und wertende Aussage, die Kurt Tucholsky sicher verneint hätte. Trotzdem ist es – so man ein wenig Feingefühl für Romantik hat – eine tolle Verfilmung, deren Tonausschnitte ich zur Illustration der Podcast-Akteure benutzt habe. Ein Manko gibt es trotzdem. Der Film konnte damals nicht direkt in Rheinsberg gedreht werden. ES gab zwar eine Anfrage. Die zuständige DEFA lehnte das Ansinnen ab, weil sich die Örtlichkeit nicht in einem filmenswerten Zustand befände. Damit hatte sie sicher recht. Man musste ausweichen, zum Beispiel nach Gut und Schloss Panker in Schleswig-Holstein, das sich in diesem Film als Rheinsberg präsentieren musste. Und gerade auch deshalb wollte ich hin – nach Rheinsberg, um zu sehen, wie es da tatsächlich ausschaut. Ganz unter uns: Schöner als im Film. – Das war mir ganz schnell klar. Da genügte schon ein kleiner Rundgang durch den Schlosspark und auch durchs Städtchen. Heute wäre die Schlossanlage und das Schlosstheater auch wieder in „filmenswertem Zustand“. Der Restaurierung ab 1992 sei Dank. Rheinsberg 2 Hier handelt es sich um eine DEFA DDR-Produktion aus dem Jahr 1987. Zwanzig Jahre später setzt man Schloss und Umgebung geschickt in Szene. Dass die Inhalte sehr viel freizügiger „rüberkommen“ ist sicher auch der Zeit geschuldet. Man kann ihn kostenfrei bei YouTube streamen. Meine Einschätzung: Kann man auch so machen. Was fehlt ist, das auch von Tucholsky angedeutete, Berliner Idiom. Die 1967er Claire (Cornelia Froboess) bleibt hier unübertroffen. Deshalb musste ihre Stimme auch zwingend in den Podcast. Specht oder Schleiereule “Ab ins Schilf”: Grienericksee / Rheinsberg – Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Da gibt es in Buch wie Film 1 einen Disput zwischen Wölfchen und Claire. Er behauptet am See einen Specht zu hören. Sie besteht darauf, es sei eine Schleiereule. Aber die Beiden streiten ja auch darüber, ob der Baum, auf den sie blicken eine Akazie oder “ne Magnolie is”. Auf der Suche nach Claires Schleiereule laufe ich zum See. Setze mich auf eine Bank, schau direkt in den Schilfgürtel und stelle fest, dass auch ich ein wenig verliebt bin. In Rheinsberg, in Claire, in Tucholsky, ins Lesen und Träumen, in die Sehnsucht und in die Natur, in die ich gerade schaue. Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD Mehr zum Thema im Reiseradio Rheinsberger Sommerfrische Rheinsberg – Von Preußen, Prinzen und Paradiesen Information & Links Kurt Tucholsky – Museum Rheinsberg Stadtgeschichte Rheinsberg Kurt Tucholsky Buchhandlung, Rheinsberg Tourismus Information Rheinsberg Brandenburgische Seenplatte Ruppiner Seenland Reiseland Brandenburg Hinweise Die Recherche für diesen Podcast wurde unterstützt von Reiseland Brandenburg und seinen Partnern vor Ort. Meine Meinung wurde nicht beeinflusst! Foto: Rüdiger Edelmann / ttb-media TON-TEXT-BILD The post D-RR307 Rheinsberg & das Bilderbuch für Verliebte first appeared on Deutsches Reiseradio (German Travelradio).

Ab 1:29 Min. - Lisa Ridzén: Wenn die Kraniche nach Süden ziehen | Gelesen von Walther Kreye | 8 Std. 29 Min. | Der Hörverlag || Ab 10:15 Min. - Stefanie Sargnagel: Opernball | Gelesen von der Autorin | 1 Std. 39 Min.| Argon || Ab 18:01 Min. - Wolfgang Schorlau: Der unaufhaltsame Aufstieg des Ministers Karsten Richter - eine Satire | Gelesen von Bjarne Mädel | 3 Std. 4 Min. | Argon Verlag || Ab 26:15 Min. - Moa Backe Åstot: Himmelsfeuer | Gelesen von Jannik Schümann | 5 Std. 48 Min. | Ab 14 Jahren | Argon Verlag

Klar, ab und zu kommen hier und da mal neue Leute dazu, und wie bei einem Geschäft mit weit geöffneter Eingangstür verirrt sich sicher auch mal jemand in den rhetorischen Krimskrams-Laden Baywatch Berlin, der eigentlich ganz woanders hinwollte und deswegen auch recht schnell mit gerümpfter Nase wieder geht. Kann alles sein. Im Prinzip kommen aber immer die gleichen Leute, und ihre Interessen altern einfach mit den drei Podcastern. Deswegen ist mittlerweile auch ein fester Bestandteil der Themenliste: Was muss man wann im Leben alles entschieden und geregelt haben? Die meisten Dinge wurden schon besprochen. Wann Darmspiegelung? Wie genau Patientenverfügung? Ab wann nicht mehr anziehen wie Ski-Aggu? Welche Hobbys gehen noch? Wo ab jetzt besaufen, wenn nicht mehr im Club? Und wie hält man sich den Rest der Welt vom Leib? Alles nachzuhören in bereits erschienenen Episoden von Baywatch Berlin. Die Beantwortung dieser Fragenliste geht in der aktuellen Folge nun weiter und darf gerne als Anregung fürs eigene Leben verstanden werden. Die Fragen sind: Wie stellt man sicher, dass bei der eigenen Beerdigung ein sexy Bild von einem auf dem Sargdeckel steht? Und was muss man jetzt schon verfügen, damit einen später, wenn das eigene Leben mal verfilmt wird, nicht ein wesentlich hübscherer Schauspieler spielt und die ganze Welt enttäuscht ist, wenn ein echtes Porträt von einem im Abspann zu sehen ist? Weitere offene Baywatch-Punkte: Wie ging es weiter in Schmitts Cringe-Holidays? Warum kann man aus Jakobs Knien schon bald eine geile Brosche machen? Und warum wartet Klaas bis heute auf die Gammelzahnfee – und was erhofft er sich von so einer eigentlich? Sie merken schon: Nicht nur Xavier Naidoo kümmert sich um die elementaren Fragen des Lebens. Ordentlich Scheisse labern können wir auch ganz alleine.

In dieser Folge spreche ich mit Christine, 58, über den Suizid ihres Sohnes vor 14 Monaten. Das neue Buch "Jugendsuizid - Das Leben danach" könnt ihr ab sofort hier bestellen: https://www.waldschnecke-verlag.de/product-page/das-leben-danach-jugendsuizid HIER KÖNNT IHR DAS BUCH "SUIZID - DAS LEBEN DANACH" von Elisa Roth bestellen: https://www.waldschnecke-verlag.de/product-page/suizid-das-leben-danach Hier findet ihr alles zu den Huggis: www.huggi.shop www.selbstwort.com (hier findet ihr den Spendenlink) „Trauer nach Suizid –Hilfe für Betroffene“ Der Podcast für Suizidbetroffene und alle,die mehr zum Thema wissen möchten. Ein Podcast von AGUS e.V. – Angehörige um Suizid. Ab 10. September wird wöchentlich eine Folge auf Spotify und auf weiteren Podcast-Plattformen veröffentlicht. Robert-Enke-Stiftung: 05105-77-5555-33 www.agus-selbsthilfe.de/ (Für Suizid-Trauernde) www.frnd.de (Freunde fürs Leben)

Berlin und Brandenburg wollen zur Deeptech-Schmiede Europas werden. Dafür ist eine neue Startup-Fabrik namens "Juni" offiziell gestartet, kurz für: "Just unite to innovate". Ab 2030 soll jeden Tag ein neues Startup entstehen. Von Efthymis Angeloudis

Do you know how to twizzle?! Neither do we! Alice and Elise settle in for another chat, with AB sporting a numb lip, and Elise coming off a filthy cold and a truly rude crook neck. Neither have been doing anything dangerous on the snow or ice, but there's a heap of Winter Olympics chat – the injuries, the drama, the sizzle, the Aussie pride. It's a very surface-level summary focusing on memes, ill-placed declarations of love, violent crashes and why on earth someone created a sport that involves skis AND guns. AB and EB compare the injuries on the slopes to the injuries we face on the roads/trails/track. There is a lot of chat about the advent of bone stress injuries and the online commentary surrounding them. Rest, sleep, nutrition and load management are all your best friends here, and the social media comparison trap is NOT IT. There's also hype building as we come into marathon season, with loads of Australians lining up at Osaka, Tokyo, Boston and London! Truly love this time of year! Make sure you come along to our free International Women's Day event with Love the Run next Sunday 1 March at Collingwood Athletics Track from 3:30pm. There'll be a run, some pilates and a panel discussion with Australian race walker and Olympic medallist Jemima Montag, four-time World Championships marathoner, Sarah Klein, and the one and only Alice Baquie, with Elise Beacom on the mic. Can't wait to see you there! Register here: https://www.lovetherun.com.au/iwdrunning -- Subscribe to Run With It wherever you get your podcasts, so you don't miss a thing! -- Follow us on Instagram: @runwithit.pod @alicebaquie @elisebeacom -- Intro/outro music by Dan Beacom Graphic design by Kate Scheer

On today's Bible Answer Man broadcast (02/18/26), Hank notes that today is Ash Wednesday, which is the beginning of the forty days of preparation for celebrating the death, burial, and resurrection of our Lord and Savior Jesus Christ. During this time, Christians remember our sinfulness, we repent, we ask for God's forgiveness, and we recognize that God's forgiveness comes at an infinite price—the death of Jesus Christ on the cross on our behalf. Hank also discusses Clean Monday, which is part of the Eastern Church's observance of the Great Lent that will be observed this coming Monday.Hank also answers the following questions:Which historian was the first to mention the Gospels? Jason - Harrisburg, PA (6:24)If Jesus died on a Friday and rose on Sunday, isn't that just two days? Jason - Harrisburg, PA (7:12)Could you clarify your statement on the doctrine of 'once saved, always saved'? Jonathan - Calgary, AB (8:57)I've had someone question me about Christ being in the grave for three days; would you mind explaining that? Kurt - Spokane, WA (16:08)What would have happened if Adam and Eve had not eaten the fruit from the tree of the knowledge of good and evil? Keith - Kalamazoo, MI (18:57)I believe the Church misunderstands the timing of the day of preparation and Jesus' crucifixion. Tracy - Hayden, ID (21:59)

2026 Live Tour: 17.10.26 Mainz, 23.10. und 24.10.26 Berlin - Tickets unter www.wahreverbrechen-podcast.deIn dieser Folge geht es um einen Fall der bis heute schwer auszuhalten ist: eine Tat innerhalb der Familie, ein Motiv, das wie „Schutz“ klingt, und eine Faktenlage, die Fragen offen lässt. Der Täter meldet sich selbst bei der Polizei, die Ermittlungen beginnen noch am selben Abend – und schnell steht nicht nur die Tat, sondern auch das „Warum“ im Raum. TW: In dieser Folge geht es um ein Tötungsdelikt und Gewalt. Bitte hört nur, wenn ihr euch damit sicher fühlt.Frage der WocheWann wird aus „Ich will dich schützen“ zu „Ich entscheide über dich“?*Enthält Werbung*Enthält Affiliate-Links*++++Unser Buch: DIE ZEUGEN - Fiktive Ich-Erzähler berichten über ihre Begegnungen mit der dunklen Seite der Täter.

Have a client that wants to switch Med Supp plans? If they live in a Med Supp birthday rule state, it might be easier than you think! Learn more about the states and their rules in this ASG Podcast episode.   Read the text version  

Merriam-Webster's Word of the Day for February 17, 2026 is: abdicate • AB-dih-kayt • verb Abdicate usually means “to renounce a position of power, such as a throne, high office, dignity, or function.” It can also mean “to fail to do what is required by (a duty or responsibility).” // I know many challenges lie ahead, but I take this role on willingly, and will not abdicate my responsibility. See the entry > Examples: “The story revolves around a plan by dark forces to kidnap the royal heirs and force the prince to abdicate his throne to an evil wizard.” — Screen Daily, 5 Jan. 2026 Did you know? Give it up for abdicate, a word powerful enough to undo a coronation. If you need a term to describe formally throwing in the towel, this one should prove—perhaps ironically—a royal success. Coming from the Latin verb abdicāre, “to resign, renounce, withdraw,” (which traces back to the verb dīcere, meaning “to speak, state”), abdicate is used primarily for those who give up sovereign power or who evade a very serious responsibility. English has dīcere to thank for a variety of other words, among them dictate, contradict, prediction, and the crown jewel of them all: dictionary.

Send a text✈️✨ Die US-Luftfahrtbehörde Federal Aviation Administration hat der neuen Allegris Business Class in der Lufthansa Boeing 787-9 endlich grünes Licht gegeben. Ab sofort sind 25 von 28 Business-Class-Sitzen buchbar – nur drei Plätze (2A, 2E, 2K) bleiben vorerst gesperrt.Nach monatelangen Verzögerungen rund um Zertifizierung, Befestigungsmodule und HIC-Vorgaben (Head Injury Criteria) kann Lufthansa damit einen großen Teil der Dreamliner-Teilflotte wirtschaftlich einsetzen.Die 787-9 mit Allegris fliegt ab Frankfurt unter anderem nach Austin, Rio de Janeiro, Kapstadt, Shanghai und Hong Kong – weitere Ziele folgen im Sommer.Ist das jetzt der Befreiungsschlag für das Prestigeprodukt? Oder bleibt es bei hohen Kosten, Aufpreismodellen und Kritik aus der Community?

Jeden Mittwoch beantworten Hannes und Niklas eine Frage aus der PAPAS-Community.Hier geht es um kleine Unsicherheiten, mal um große Sorgen und manchmal um ganz alltägliche Situationen, bei denen man sich fragt: Geht das nur mir so? Spoiler: tut es nicht. Wir sitzen nämlich alle im selben Boot.Kleine große Fragen ist ein bisschen wie Domian für Eltern. Ab jetzt jeden Mittwoch.Schickt uns doch eure Fragen gerne hier unter der Folge, per Mail oder auf Instagram.Wir freuen uns von euch zu hören < 3 Hosted on Acast. See acast.com/privacy for more information.

Irene Chen is the Co-Founder and Partner at Parker Thatch, a role she has held for over 24 years. Her top skills include Brand Development, Fashion, and Social Media. Before co-founding Parker Thatch, Irene served as the Director of Product Development for Donna Karan. She is a graduate of the University of California, Los Angeles. Matthew Grenby is the Partner and Co-Founder of Parker Thatch, a position he has held for over 24 years. His expertise lies in Strategy, Start-ups, and Entrepreneurship. Prior to Parker Thatch, he was a Vice President at Castling Group, where he led UX and design to launch online divisions for major brands, and a Data Scientist at Intel, developing novel data visualizations. He holds an MBA from Columbia Business School, an MS from the M.I.T. Media Lab , an MS in Graphic Design from ArtCenter College of Design , and an AB in English from Harvard University. In This Conversation We Discuss:[00:00] Intro[00:56] Bootstrapping growth through cash flow[03:23] Turning local talent into a luxury launchpad[07:45] Sponsor: Klaviyo [09:52] Applying corporate training to startups[12:31] Challenging traditional production paths[18:48] Sponsor: Intelligems [20:48] Standardizing core products for efficiency[24:47] Sponsor: Electric Eye[25:56] Persisting through daily business doubt[29:40] Callouts[29:50] Reinventing challenges for better outcomes[31:34] Leveraging community for business insights[32:02] Maintaining connections for future opportunities[36:03] Rebranding for clarity and customer reachResources:Subscribe to Honest Ecommerce on YoutubeLuxury products for everyday ease and elegance parkerthatch.com/Follow Irene Chen linkedin.com/in/irene-chen-16b16823/Follow Matthew Grenby linkedin.com/in/matthewgrenby/Book a demo today at intelligems.io/Schedule an intro call with one of our experts electriceye.io/connectGet your free demo https://www.klaviyo.com/honestIf you're enjoying the show, we'd love it if you left Honest Ecommerce a review on Apple Podcasts. It makes a huge impact on the success of the podcast, and we love reading every one of your reviews!

What if a company could deliver high quality products at low cost, improving the value for customers and giving it a competitive edge, all while offering higher pay and career growth opportunities for its employees and not hurting the bottom line?Zeynep Ton is a professor at MIT's Sloan School of Management, president of the Good Jobs Institute, and author of The Case for Good Jobs: How Great Companies Bring Dignity, Pay, and Meaning to Everyone's Work. Zeynep joins Greg to explain the interconnected components of the “good job strategy,” such as standardization, empowerment, cross-training, simplification, and the incorporation of slack in schedules. She emphasizes that companies should view their workforce as value drivers rather than costs to be minimized, advocating for investment in employees for better productivity and sustainable company growth.*unSILOed Podcast is produced by University FM.*Episode Quotes:The ‘good job strategy' requires systems thinking43:47: A lot of organizations operate in silos, and ‘the good job strategy' requires systems thinking, interconnected decisions, and all the decisions coming back to: how do we create value for the customer and how does this interact with other choices to deliver that type of value? And as long as we do the AB testing and requiring on, rigorous, and I do not think it is rigorous, it is, yeah, it is math, but it is not rigorous logic, it will be very difficult to adopt this.Standardization is a gift28:51: Standardization is a gift because there are so many things I do not even have to think about. So, think each of these choices is helpful to say what are the mindsets that are driving the choices, when used that way, and standardization is not just about work, [but also] standardization of management practices.Why ‘the good job strategy' creates competitive advantage13:02: I can see a lot of companies in the same industry using ‘the good job strategy' as long as they have a differentiation in the eyes of their customers and they're improving their value, continuously using the strategy. It's not good jobs that differentiates. It's the customer value that is a source of competitive advantage.Why unmet basic needs drive employee turnover17:02: You ask our students what motivates people. Everybody is gonna talk about is a sense of belonging, achievement, meaning, recognition. Of course, those things are the motivators. But so many people do not have their basic needs met. And there is tremendous lack of awareness. And those are, oftentimes, the biggest reasons for employee turnover that I have seen in many organizations that I work with.Show Links:Recommended Resources:Good Jobs Institute Toyota Production SystemJohn Paul MacDuffieCharlie MungerQueueing theory“How CEOs Manage Time” by Michael Porter and Nitin NohriaBob NardelliPete StavrosGuest Profile:Faculty Profile at MIT Sloan School of ManagementProfessional WebsiteProfessional Profile on LinkedInGuest Work:The Case for Good Jobs: How Great Companies Bring Dignity, Pay, and Meaning to Everyone's WorkThe Good Jobs Strategy: How the Smartest Companies Invest in Employees to Lower Costs and Boost Profits Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Can you trust real estate listing photos anymore?With editing software and AI tools becoming more advanced, it is easier than ever to alter images of a property before it hits the market. That is exactly why California Assembly Bill 723 (AB 723) is going into effect on January 1, 2026.In this episode of the Kern County Real Estate Review, Laurie McCarty is joined by Chris O'Donnell, owner of Selling Image and one of Bakersfield's most trusted real estate photographers, to break down what AB 723 really means for buyers, sellers, and agents.AB 723 requires a reasonably conspicuous disclosure on any listing photo that has been digitally altered or AI-generated in a way that changes physical elements of a property. It also requires access to the original, unedited images through a link, URL, or QR code. The goal is transparency and protecting buyers from misleading property images.In this episode, we discuss:• What counts as a “digitally altered” real estate photo• How AI is changing real estate photography• What must now be disclosed under AB 723• How this law impacts virtual staging• Why transparency in listing photos matters• What buyers should watch for when viewing homes online• How agents and photographers are adapting to the new requirementsIf you are buying, selling, or simply browsing homes online, this is a conversation you need to hear.Tune in to understand how AB 723 will change real estate listings in California and what it means for the future of real estate photography.

Today we welcome Clint Kellum, the new (as of November 2025) Director of the California Department of Cannabis Control.  The CDCC licenses and regulates cannabis businesses, including regulation of cannabis growers, manufacture of cannabis products, and sales, transportation and tracking of cannabis goods. Kellum took the helm just after Governor Newsom signed AB 564, a bill to reset the state's cannabis excise tax rate at 15% until 2028, giving California's struggling legal cannabis market hope after a rough few years. Kellum explains what AB 564 means for the industry and the state, and looks at challenges and opportunities ahead. And, as always, we tell you who had the Worst Week in California Politics.1:03 Bill introductions and lobbying reports3:42 The Top Two Simulator6:04 Clint Kellum6:45 AB5648:09 The taxation environment12:42 Complexities of descheduling13:30 Startups?15:02 Illegal cannabis seizures19:46 Direct to consumer sales21:54 Biggest challenges and opportunities25:49 How did you get into this?29:03 #WWCAWant to support the Capitol Weekly Podcast? Make your tax deductible donation here: capitolweekly.net/donations/Capitol Weekly Podcast theme is "Pickin' My Way" by Eddie Lang "#WorstWeekCA" Beat provided by freebeats.io Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

durée : 00:37:48 - Les Nuits de France Culture - par : Mathias Le Gargasson - Chacun s'est déjà identifié à un personnage de roman ou à un personnage historique. Comment l'histoire et la littérature façonnent-elles notre vision du couple ? Héloïse et Abélard, Tristan et Iseut, et bien d'autres révèlent les paradoxes de l'amour à travers les siècles. - réalisation : Emily Vallat - invités : Philippe Ariès Historien

Willkommen zur ultimativen Diablo-Paartherapie! In dieser Folge habe ich mir Diablo-Urgestein Jessirocks auf die Couch geholt, um seine Beziehung mit dem ARPG aufzuarbeiten. Wir blicken zurück auf die rosaroten Anfangszeiten, den brutalen Vertrauensbruch durch Diablo Immortal und das endlose Auf und Ab in Diablo 4. Wie viel Frust, fehlendes Endgame und teure Shop-Cosmetics kann eine treue Community noch ertragen? Warum fällt es uns so schwer, einfach zu Path of Exile zu wechseln? Außerdem packt Jessi seinen Aluhut aus: Wir diskutieren über einen möglichen Warlock-Shadowdrop, den anstehenden Komplettumbau aller Klassen und die Frage, ob Microsoft den Stecker zieht, wenn dieses Add-on floppt. Alle Links zum GameStar Podcast und unseren Werbepartnern: https://linktr.ee/gamestarpodcast

From rewriting Google's search stack in the early 2000s to reviving sparse trillion-parameter models and co-designing TPUs with frontier ML research, Jeff Dean has quietly shaped nearly every layer of the modern AI stack. As Chief AI Scientist at Google and a driving force behind Gemini, Jeff has lived through multiple scaling revolutions from CPUs and sharded indices to multimodal models that reason across text, video, and code.Jeff joins us to unpack what it really means to “own the Pareto frontier,” why distillation is the engine behind every Flash model breakthrough, how energy (in picojoules) not FLOPs is becoming the true bottleneck, what it was like leading the charge to unify all of Google's AI teams, and why the next leap won't come from bigger context windows alone, but from systems that give the illusion of attending to trillions of tokens.We discuss:* Jeff's early neural net thesis in 1990: parallel training before it was cool, why he believed scaling would win decades early, and the “bigger model, more data, better results” mantra that held for 15 years* The evolution of Google Search: sharding, moving the entire index into memory in 2001, softening query semantics pre-LLMs, and why retrieval pipelines already resemble modern LLM systems* Pareto frontier strategy: why you need both frontier “Pro” models and low-latency “Flash” models, and how distillation lets smaller models surpass prior generations* Distillation deep dive: ensembles → compression → logits as soft supervision, and why you need the biggest model to make the smallest one good* Latency as a first-class objective: why 10–50x lower latency changes UX entirely, and how future reasoning workloads will demand 10,000 tokens/sec* Energy-based thinking: picojoules per bit, why moving data costs 1000x more than a multiply, batching through the lens of energy, and speculative decoding as amortization* TPU co-design: predicting ML workloads 2–6 years out, speculative hardware features, precision reduction, sparsity, and the constant feedback loop between model architecture and silicon* Sparse models and “outrageously large” networks: trillions of parameters with 1–5% activation, and why sparsity was always the right abstraction* Unified vs. specialized models: abandoning symbolic systems, why general multimodal models tend to dominate vertical silos, and when vertical fine-tuning still makes sense* Long context and the illusion of scale: beyond needle-in-a-haystack benchmarks toward systems that narrow trillions of tokens to 117 relevant documents* Personalized AI: attending to your emails, photos, and documents (with permission), and why retrieval + reasoning will unlock deeply personal assistants* Coding agents: 50 AI interns, crisp specifications as a new core skill, and how ultra-low latency will reshape human–agent collaboration* Why ideas still matter: transformers, sparsity, RL, hardware, systems — scaling wasn't blind; the pieces had to multiply togetherShow Notes:* Gemma 3 Paper* Gemma 3* Gemini 2.5 Report* Jeff Dean's “Software Engineering Advice fromBuilding Large-Scale Distributed Systems” Presentation (with Back of the Envelope Calculations)* Latency Numbers Every Programmer Should Know by Jeff Dean* The Jeff Dean Facts* Jeff Dean Google Bio* Jeff Dean on “Important AI Trends” @Stanford AI Club* Jeff Dean & Noam Shazeer — 25 years at Google (Dwarkesh)—Jeff Dean* LinkedIn: https://www.linkedin.com/in/jeff-dean-8b212555* X: https://x.com/jeffdeanGoogle* https://google.com* https://deepmind.googleFull Video EpisodeTimestamps00:00:04 — Introduction: Alessio & Swyx welcome Jeff Dean, chief AI scientist at Google, to the Latent Space podcast00:00:30 — Owning the Pareto Frontier & balancing frontier vs low-latency models00:01:31 — Frontier models vs Flash models + role of distillation00:03:52 — History of distillation and its original motivation00:05:09 — Distillation's role in modern model scaling00:07:02 — Model hierarchy (Flash, Pro, Ultra) and distillation sources00:07:46 — Flash model economics & wide deployment00:08:10 — Latency importance for complex tasks00:09:19 — Saturation of some tasks and future frontier tasks00:11:26 — On benchmarks, public vs internal00:12:53 — Example long-context benchmarks & limitations00:15:01 — Long-context goals: attending to trillions of tokens00:16:26 — Realistic use cases beyond pure language00:18:04 — Multimodal reasoning and non-text modalities00:19:05 — Importance of vision & motion modalities00:20:11 — Video understanding example (extracting structured info)00:20:47 — Search ranking analogy for LLM retrieval00:23:08 — LLM representations vs keyword search00:24:06 — Early Google search evolution & in-memory index00:26:47 — Design principles for scalable systems00:28:55 — Real-time index updates & recrawl strategies00:30:06 — Classic “Latency numbers every programmer should know”00:32:09 — Cost of memory vs compute and energy emphasis00:34:33 — TPUs & hardware trade-offs for serving models00:35:57 — TPU design decisions & co-design with ML00:38:06 — Adapting model architecture to hardware00:39:50 — Alternatives: energy-based models, speculative decoding00:42:21 — Open research directions: complex workflows, RL00:44:56 — Non-verifiable RL domains & model evaluation00:46:13 — Transition away from symbolic systems toward unified LLMs00:47:59 — Unified models vs specialized ones00:50:38 — Knowledge vs reasoning & retrieval + reasoning00:52:24 — Vertical model specialization & modules00:55:21 — Token count considerations for vertical domains00:56:09 — Low resource languages & contextual learning00:59:22 — Origins: Dean's early neural network work01:10:07 — AI for coding & human–model interaction styles01:15:52 — Importance of crisp specification for coding agents01:19:23 — Prediction: personalized models & state retrieval01:22:36 — Token-per-second targets (10k+) and reasoning throughput01:23:20 — Episode conclusion and thanksTranscriptAlessio Fanelli [00:00:04]: Hey everyone, welcome to the Latent Space podcast. This is Alessio, founder of Kernel Labs, and I'm joined by Swyx, editor of Latent Space. Shawn Wang [00:00:11]: Hello, hello. We're here in the studio with Jeff Dean, chief AI scientist at Google. Welcome. Thanks for having me. It's a bit surreal to have you in the studio. I've watched so many of your talks, and obviously your career has been super legendary. So, I mean, congrats. I think the first thing must be said, congrats on owning the Pareto Frontier.Jeff Dean [00:00:30]: Thank you, thank you. Pareto Frontiers are good. It's good to be out there.Shawn Wang [00:00:34]: Yeah, I mean, I think it's a combination of both. You have to own the Pareto Frontier. You have to have like frontier capability, but also efficiency, and then offer that range of models that people like to use. And, you know, some part of this was started because of your hardware work. Some part of that is your model work, and I'm sure there's lots of secret sauce that you guys have worked on cumulatively. But, like, it's really impressive to see it all come together in, like, this slittily advanced.Jeff Dean [00:01:04]: Yeah, yeah. I mean, I think, as you say, it's not just one thing. It's like a whole bunch of things up and down the stack. And, you know, all of those really combine to help make UNOS able to make highly capable large models, as well as, you know, software techniques to get those large model capabilities into much smaller, lighter weight models that are, you know, much more cost effective and lower latency, but still, you know, quite capable for their size. Yeah.Alessio Fanelli [00:01:31]: How much pressure do you have on, like, having the lower bound of the Pareto Frontier, too? I think, like, the new labs are always trying to push the top performance frontier because they need to raise more money and all of that. And you guys have billions of users. And I think initially when you worked on the CPU, you were thinking about, you know, if everybody that used Google, we use the voice model for, like, three minutes a day, they were like, you need to double your CPU number. Like, what's that discussion today at Google? Like, how do you prioritize frontier versus, like, we have to do this? How do we actually need to deploy it if we build it?Jeff Dean [00:02:03]: Yeah, I mean, I think we always want to have models that are at the frontier or pushing the frontier because I think that's where you see what capabilities now exist that didn't exist at the sort of slightly less capable last year's version or last six months ago version. At the same time, you know, we know those are going to be really useful for a bunch of use cases, but they're going to be a bit slower and a bit more expensive than people might like for a bunch of other broader models. So I think what we want to do is always have kind of a highly capable sort of affordable model that enables a whole bunch of, you know, lower latency use cases. People can use them for agentic coding much more readily and then have the high-end, you know, frontier model that is really useful for, you know, deep reasoning, you know, solving really complicated math problems, those kinds of things. And it's not that. One or the other is useful. They're both useful. So I think we'd like to do both. And also, you know, through distillation, which is a key technique for making the smaller models more capable, you know, you have to have the frontier model in order to then distill it into your smaller model. So it's not like an either or choice. You sort of need that in order to actually get a highly capable, more modest size model. Yeah.Alessio Fanelli [00:03:24]: I mean, you and Jeffrey came up with the solution in 2014.Jeff Dean [00:03:28]: Don't forget, L'Oreal Vinyls as well. Yeah, yeah.Alessio Fanelli [00:03:30]: A long time ago. But like, I'm curious how you think about the cycle of these ideas, even like, you know, sparse models and, you know, how do you reevaluate them? How do you think about in the next generation of model, what is worth revisiting? Like, yeah, they're just kind of like, you know, you worked on so many ideas that end up being influential, but like in the moment, they might not feel that way necessarily. Yeah.Jeff Dean [00:03:52]: I mean, I think distillation was originally motivated because we were seeing that we had a very large image data set at the time, you know, 300 million images that we could train on. And we were seeing that if you create specialists for different subsets of those image categories, you know, this one's going to be really good at sort of mammals, and this one's going to be really good at sort of indoor room scenes or whatever, and you can cluster those categories and train on an enriched stream of data after you do pre-training on a much broader set of images. You get much better performance. If you then treat that whole set of maybe 50 models you've trained as a large ensemble, but that's not a very practical thing to serve, right? So distillation really came about from the idea of, okay, what if we want to actually serve that and train all these independent sort of expert models and then squish it into something that actually fits in a form factor that you can actually serve? And that's, you know, not that different from what we're doing today. You know, often today we're instead of having an ensemble of 50 models. We're having a much larger scale model that we then distill into a much smaller scale model.Shawn Wang [00:05:09]: Yeah. A part of me also wonders if distillation also has a story with the RL revolution. So let me maybe try to articulate what I mean by that, which is you can, RL basically spikes models in a certain part of the distribution. And then you have to sort of, well, you can spike models, but usually sometimes... It might be lossy in other areas and it's kind of like an uneven technique, but you can probably distill it back and you can, I think that the sort of general dream is to be able to advance capabilities without regressing on anything else. And I think like that, that whole capability merging without loss, I feel like it's like, you know, some part of that should be a distillation process, but I can't quite articulate it. I haven't seen much papers about it.Jeff Dean [00:06:01]: Yeah, I mean, I tend to think of one of the key advantages of distillation is that you can have a much smaller model and you can have a very large, you know, training data set and you can get utility out of making many passes over that data set because you're now getting the logits from the much larger model in order to sort of coax the right behavior out of the smaller model that you wouldn't otherwise get with just the hard labels. And so, you know, I think that's what we've observed. Is you can get, you know, very close to your largest model performance with distillation approaches. And that seems to be, you know, a nice sweet spot for a lot of people because it enables us to kind of, for multiple Gemini generations now, we've been able to make the sort of flash version of the next generation as good or even substantially better than the previous generations pro. And I think we're going to keep trying to do that because that seems like a good trend to follow.Shawn Wang [00:07:02]: So, Dara asked, so it was the original map was Flash Pro and Ultra. Are you just sitting on Ultra and distilling from that? Is that like the mother load?Jeff Dean [00:07:12]: I mean, we have a lot of different kinds of models. Some are internal ones that are not necessarily meant to be released or served. Some are, you know, our pro scale model and we can distill from that as well into our Flash scale model. So I think, you know, it's an important set of capabilities to have and also inference time scaling. It can also be a useful thing to improve the capabilities of the model.Shawn Wang [00:07:35]: And yeah, yeah, cool. Yeah. And obviously, I think the economy of Flash is what led to the total dominance. I think the latest number is like 50 trillion tokens. I don't know. I mean, obviously, it's changing every day.Jeff Dean [00:07:46]: Yeah, yeah. But, you know, by market share, hopefully up.Shawn Wang [00:07:50]: No, I mean, there's no I mean, there's just the economics wise, like because Flash is so economical, like you can use it for everything. Like it's in Gmail now. It's in YouTube. Like it's yeah. It's in everything.Jeff Dean [00:08:02]: We're using it more in our search products of various AI mode reviews.Shawn Wang [00:08:05]: Oh, my God. Flash past the AI mode. Oh, my God. Yeah, that's yeah, I didn't even think about that.Jeff Dean [00:08:10]: I mean, I think one of the things that is quite nice about the Flash model is not only is it more affordable, it's also a lower latency. And I think latency is actually a pretty important characteristic for these models because we're going to want models to do much more complicated things that are going to involve, you know, generating many more tokens from when you ask the model to do so. So, you know, if you're going to ask the model to do something until it actually finishes what you ask it to do, because you're going to ask now, not just write me a for loop, but like write me a whole software package to do X or Y or Z. And so having low latency systems that can do that seems really important. And Flash is one direction, one way of doing that. You know, obviously our hardware platforms enable a bunch of interesting aspects of our, you know, serving stack as well, like TPUs, the interconnect between. Chips on the TPUs is actually quite, quite high performance and quite amenable to, for example, long context kind of attention operations, you know, having sparse models with lots of experts. These kinds of things really, really matter a lot in terms of how do you make them servable at scale.Alessio Fanelli [00:09:19]: Yeah. Does it feel like there's some breaking point for like the proto Flash distillation, kind of like one generation delayed? I almost think about almost like the capability as a. In certain tasks, like the pro model today is a saturated, some sort of task. So next generation, that same task will be saturated at the Flash price point. And I think for most of the things that people use models for at some point, the Flash model in two generation will be able to do basically everything. And how do you make it economical to like keep pushing the pro frontier when a lot of the population will be okay with the Flash model? I'm curious how you think about that.Jeff Dean [00:09:59]: I mean, I think that's true. If your distribution of what people are asking people, the models to do is stationary, right? But I think what often happens is as the models become more capable, people ask them to do more, right? So, I mean, I think this happens in my own usage. Like I used to try our models a year ago for some sort of coding task, and it was okay at some simpler things, but wouldn't do work very well for more complicated things. And since then, we've improved dramatically on the more complicated coding tasks. And now I'll ask it to do much more complicated things. And I think that's true, not just of coding, but of, you know, now, you know, can you analyze all the, you know, renewable energy deployments in the world and give me a report on solar panel deployment or whatever. That's a very complicated, you know, more complicated task than people would have asked a year ago. And so you are going to want more capable models to push the frontier in the absence of what people ask the models to do. And that also then gives us. Insight into, okay, where does the, where do things break down? How can we improve the model in these, these particular areas, uh, in order to sort of, um, make the next generation even better.Alessio Fanelli [00:11:11]: Yeah. Are there any benchmarks or like test sets they use internally? Because it's almost like the same benchmarks get reported every time. And it's like, all right, it's like 99 instead of 97. Like, how do you have to keep pushing the team internally to it? Or like, this is what we're building towards. Yeah.Jeff Dean [00:11:26]: I mean, I think. Benchmarks, particularly external ones that are publicly available. Have their utility, but they often kind of have a lifespan of utility where they're introduced and maybe they're quite hard for current models. You know, I, I like to think of the best kinds of benchmarks are ones where the initial scores are like 10 to 20 or 30%, maybe, but not higher. And then you can sort of work on improving that capability for, uh, whatever it is, the benchmark is trying to assess and get it up to like 80, 90%, whatever. I, I think once it hits kind of 95% or something, you get very diminishing returns from really focusing on that benchmark, cuz it's sort of, it's either the case that you've now achieved that capability, or there's also the issue of leakage in public data or very related kind of data being, being in your training data. Um, so we have a bunch of held out internal benchmarks that we really look at where we know that wasn't represented in the training data at all. There are capabilities that we want the model to have. Um, yeah. Yeah. Um, that it doesn't have now, and then we can work on, you know, assessing, you know, how do we make the model better at these kinds of things? Is it, we need different kind of data to train on that's more specialized for this particular kind of task. Do we need, um, you know, a bunch of, uh, you know, architectural improvements or some sort of, uh, model capability improvements, you know, what would help make that better?Shawn Wang [00:12:53]: Is there, is there such an example that you, uh, a benchmark inspired in architectural improvement? Like, uh, I'm just kind of. Jumping on that because you just.Jeff Dean [00:13:02]: Uh, I mean, I think some of the long context capability of the, of the Gemini models that came, I guess, first in 1.5 really were about looking at, okay, we want to have, um, you know,Shawn Wang [00:13:15]: immediately everyone jumped to like completely green charts of like, everyone had, I was like, how did everyone crack this at the same time? Right. Yeah. Yeah.Jeff Dean [00:13:23]: I mean, I think, um, and once you're set, I mean, as you say that needed single needle and a half. Hey, stack benchmark is really saturated for at least context links up to 1, 2 and K or something. Don't actually have, you know, much larger than 1, 2 and 8 K these days or two or something. We're trying to push the frontier of 1 million or 2 million context, which is good because I think there are a lot of use cases where. Yeah. You know, putting a thousand pages of text or putting, you know, multiple hour long videos and the context and then actually being able to make use of that as useful. Try to, to explore the über graduation are fairly large. But the single needle in a haystack benchmark is sort of saturated. So you really want more complicated, sort of multi-needle or more realistic, take all this content and produce this kind of answer from a long context that sort of better assesses what it is people really want to do with long context. Which is not just, you know, can you tell me the product number for this particular thing?Shawn Wang [00:14:31]: Yeah, it's retrieval. It's retrieval within machine learning. It's interesting because I think the more meta level I'm trying to operate at here is you have a benchmark. You're like, okay, I see the architectural thing I need to do in order to go fix that. But should you do it? Because sometimes that's an inductive bias, basically. It's what Jason Wei, who used to work at Google, would say. Exactly the kind of thing. Yeah, you're going to win. Short term. Longer term, I don't know if that's going to scale. You might have to undo that.Jeff Dean [00:15:01]: I mean, I like to sort of not focus on exactly what solution we're going to derive, but what capability would you want? And I think we're very convinced that, you know, long context is useful, but it's way too short today. Right? Like, I think what you would really want is, can I attend to the internet while I answer my question? Right? But that's not going to happen. I think that's going to be solved by purely scaling the existing solutions, which are quadratic. So a million tokens kind of pushes what you can do. You're not going to do that to a trillion tokens, let alone, you know, a billion tokens, let alone a trillion. But I think if you could give the illusion that you can attend to trillions of tokens, that would be amazing. You'd find all kinds of uses for that. You would have attend to the internet. You could attend to the pixels of YouTube and the sort of deeper representations that we can find. You could attend to the form for a single video, but across many videos, you know, on a personal Gemini level, you could attend to all of your personal state with your permission. So like your emails, your photos, your docs, your plane tickets you have. I think that would be really, really useful. And the question is, how do you get algorithmic improvements and system level improvements that get you to something where you actually can attend to trillions of tokens? Right. In a meaningful way. Yeah.Shawn Wang [00:16:26]: But by the way, I think I did some math and it's like, if you spoke all day, every day for eight hours a day, you only generate a maximum of like a hundred K tokens, which like very comfortably fits.Jeff Dean [00:16:38]: Right. But if you then say, okay, I want to be able to understand everything people are putting on videos.Shawn Wang [00:16:46]: Well, also, I think that the classic example is you start going beyond language into like proteins and whatever else is extremely information dense. Yeah. Yeah.Jeff Dean [00:16:55]: I mean, I think one of the things about Gemini's multimodal aspects is we've always wanted it to be multimodal from the start. And so, you know, that sometimes to people means text and images and video sort of human-like and audio, audio, human-like modalities. But I think it's also really useful to have Gemini know about non-human modalities. Yeah. Like LIDAR sensor data from. Yes. Say, Waymo vehicles or. Like robots or, you know, various kinds of health modalities, x-rays and MRIs and imaging and genomics information. And I think there's probably hundreds of modalities of data where you'd like the model to be able to at least be exposed to the fact that this is an interesting modality and has certain meaning in the world. Where even if you haven't trained on all the LIDAR data or MRI data, you could have, because maybe that's not, you know, it doesn't make sense in terms of trade-offs of. You know, what you include in your main pre-training data mix, at least including a little bit of it is actually quite useful. Yeah. Because it sort of tempts the model that this is a thing.Shawn Wang [00:18:04]: Yeah. Do you believe, I mean, since we're on this topic and something I just get to ask you all the questions I always wanted to ask, which is fantastic. Like, are there some king modalities, like modalities that supersede all the other modalities? So a simple example was Vision can, on a pixel level, encode text. And DeepSeq had this DeepSeq CR paper that did that. Vision. And Vision has also been shown to maybe incorporate audio because you can do audio spectrograms and that's, that's also like a Vision capable thing. Like, so, so maybe Vision is just the king modality and like. Yeah.Jeff Dean [00:18:36]: I mean, Vision and Motion are quite important things, right? Motion. Well, like video as opposed to static images, because I mean, there's a reason evolution has evolved eyes like 23 independent ways, because it's such a useful capability for sensing the world around you, which is really what we want these models to be. So I think the only thing that we can be able to do is interpret the things we're seeing or the things we're paying attention to and then help us in using that information to do things. Yeah.Shawn Wang [00:19:05]: I think motion, you know, I still want to shout out, I think Gemini, still the only native video understanding model that's out there. So I use it for YouTube all the time. Nice.Jeff Dean [00:19:15]: Yeah. Yeah. I mean, it's actually, I think people kind of are not necessarily aware of what the Gemini models can actually do. Yeah. Like I have an example I've used in one of my talks. It had like, it was like a YouTube highlight video of 18 memorable sports moments across the last 20 years or something. So it has like Michael Jordan hitting some jump shot at the end of the finals and, you know, some soccer goals and things like that. And you can literally just give it the video and say, can you please make me a table of what all these different events are? What when the date is when they happened? And a short description. And so you get like now an 18 row table of that information extracted from the video, which is, you know, not something most people think of as like a turn video into sequel like table.Alessio Fanelli [00:20:11]: Has there been any discussion inside of Google of like, you mentioned tending to the whole internet, right? Google, it's almost built because a human cannot tend to the whole internet and you need some sort of ranking to find what you need. Yep. That ranking is like much different for an LLM because you can expect a person to look at maybe the first five, six links in a Google search versus for an LLM. Should you expect to have 20 links that are highly relevant? Like how do you internally figure out, you know, how do we build the AI mode that is like maybe like much broader search and span versus like the more human one? Yeah.Jeff Dean [00:20:47]: I mean, I think even pre-language model based work, you know, our ranking systems would be built to start. I mean, I think even pre-language model based work, you know, our ranking systems would be built to start. With a giant number of web pages in our index, many of them are not relevant. So you identify a subset of them that are relevant with very lightweight kinds of methods. You know, you're down to like 30,000 documents or something. And then you gradually refine that to apply more and more sophisticated algorithms and more and more sophisticated sort of signals of various kinds in order to get down to ultimately what you show, which is, you know, the final 10 results or, you know, 10 results plus. Other kinds of information. And I think an LLM based system is not going to be that dissimilar, right? You're going to attend to trillions of tokens, but you're going to want to identify, you know, what are the 30,000 ish documents that are with the, you know, maybe 30 million interesting tokens. And then how do you go from that into what are the 117 documents I really should be paying attention to in order to carry out the tasks that the user has asked? And I think, you know, you can imagine systems where you have, you know, a lot of highly parallel processing to identify those initial 30,000 candidates, maybe with very lightweight kinds of models. Then you have some system that sort of helps you narrow down from 30,000 to the 117 with maybe a little bit more sophisticated model or set of models. And then maybe the final model is the thing that looks. So the 117 things that might be your most capable model. So I think it has to, it's going to be some system like that, that is really enables you to give the illusion of attending to trillions of tokens. Sort of the way Google search gives you, you know, not the illusion, but you are searching the internet, but you're finding, you know, a very small subset of things that are, that are relevant.Shawn Wang [00:22:47]: Yeah. I often tell a lot of people that are not steeped in like Google search history that, well, you know, like Bert was. Like he was like basically immediately inside of Google search and that improves results a lot, right? Like I don't, I don't have any numbers off the top of my head, but like, I'm sure you guys, that's obviously the most important numbers to Google. Yeah.Jeff Dean [00:23:08]: I mean, I think going to an LLM based representation of text and words and so on enables you to get out of the explicit hard notion of, of particular words having to be on the page, but really getting at the notion of this topic of this page or this page. Paragraph is highly relevant to this query. Yeah.Shawn Wang [00:23:28]: I don't think people understand how much LLMs have taken over all these very high traffic system, very high traffic. Yeah. Like it's Google, it's YouTube. YouTube has this like semantics ID thing where it's just like every token or every item in the vocab is a YouTube video or something that predicts the video using a code book, which is absurd to me for YouTube size.Jeff Dean [00:23:50]: And then most recently GROK also for, for XAI, which is like, yeah. I mean, I'll call out even before LLMs were used extensively in search, we put a lot of emphasis on softening the notion of what the user actually entered into the query.Shawn Wang [00:24:06]: So do you have like a history of like, what's the progression? Oh yeah.Jeff Dean [00:24:09]: I mean, I actually gave a talk in, uh, I guess, uh, web search and data mining conference in 2009, uh, where we never actually published any papers about the origins of Google search, uh, sort of, but we went through sort of four or five or six. generations, four or five or six generations of, uh, redesigning of the search and retrieval system, uh, from about 1999 through 2004 or five. And that talk is really about that evolution. And one of the things that really happened in 2001 was we were sort of working to scale the system in multiple dimensions. So one is we wanted to make our index bigger, so we could retrieve from a larger index, which always helps your quality in general. Uh, because if you don't have the page in your index, you're going to not do well. Um, and then we also needed to scale our capacity because we were, our traffic was growing quite extensively. Um, and so we had, you know, a sharded system where you have more and more shards as the index grows, you have like 30 shards. And then if you want to double the index size, you make 60 shards so that you can bound the latency by which you respond for any particular user query. Um, and then as traffic grows, you add, you add more and more replicas of each of those. And so we eventually did the math that realized that in a data center where we had say 60 shards and, um, you know, 20 copies of each shard, we now had 1200 machines, uh, with disks. And we did the math and we're like, Hey, one copy of that index would actually fit in memory across 1200 machines. So in 2001, we introduced, uh, we put our entire index in memory and what that enabled from a quality perspective was amazing. Um, and so we had more and more replicas of each of those. Before you had to be really careful about, you know, how many different terms you looked at for a query, because every one of them would involve a disk seek on every one of the 60 shards. And so you, as you make your index bigger, that becomes even more inefficient. But once you have the whole index in memory, it's totally fine to have 50 terms you throw into the query from the user's original three or four word query, because now you can add synonyms like restaurant and restaurants and cafe and, uh, you know, things like that. Uh, bistro and all these things. And you can suddenly start, uh, sort of really, uh, getting at the meaning of the word as opposed to the exact semantic form the user typed in. And that was, you know, 2001, very much pre LLM, but really it was about softening the, the strict definition of what the user typed in order to get at the meaning.Alessio Fanelli [00:26:47]: What are like principles that you use to like design the systems, especially when you have, I mean, in 2001, the internet is like. Doubling, tripling every year in size is not like, uh, you know, and I think today you kind of see that with LLMs too, where like every year the jumps in size and like capabilities are just so big. Are there just any, you know, principles that you use to like, think about this? Yeah.Jeff Dean [00:27:08]: I mean, I think, uh, you know, first, whenever you're designing a system, you want to understand what are the sort of design parameters that are going to be most important in designing that, you know? So, you know, how many queries per second do you need to handle? How big is the internet? How big is the index you need to handle? How much data do you need to keep for every document in the index? How are you going to look at it when you retrieve things? Um, what happens if traffic were to double or triple, you know, will that system work well? And I think a good design principle is you're going to want to design a system so that the most important characteristics could scale by like factors of five or 10, but probably not beyond that because often what happens is if you design a system for X. And something suddenly becomes a hundred X, that would enable a very different point in the design space that would not make sense at X. But all of a sudden at a hundred X makes total sense. So like going from a disk space index to a in memory index makes a lot of sense once you have enough traffic, because now you have enough replicas of the sort of state on disk that those machines now actually can hold, uh, you know, a full copy of the, uh, index and memory. Yeah. And that all of a sudden enabled. A completely different design that wouldn't have been practical before. Yeah. Um, so I'm, I'm a big fan of thinking through designs in your head, just kind of playing with the design space a little before you actually do a lot of writing of code. But, you know, as you said, in the early days of Google, we were growing the index, uh, quite extensively. We were growing the update rate of the index. So the update rate actually is the parameter that changed the most. Surprising. So it used to be once a month.Shawn Wang [00:28:55]: Yeah.Jeff Dean [00:28:56]: And then we went to a system that could update any particular page in like sub one minute. Okay.Shawn Wang [00:29:02]: Yeah. Because this is a competitive advantage, right?Jeff Dean [00:29:04]: Because all of a sudden news related queries, you know, if you're, if you've got last month's news index, it's not actually that useful for.Shawn Wang [00:29:11]: News is a special beast. Was there any, like you could have split it onto a separate system.Jeff Dean [00:29:15]: Well, we did. We launched a Google news product, but you also want news related queries that people type into the main index to also be sort of updated.Shawn Wang [00:29:23]: So, yeah, it's interesting. And then you have to like classify whether the page is, you have to decide which pages should be updated and what frequency. Oh yeah.Jeff Dean [00:29:30]: There's a whole like, uh, system behind the scenes that's trying to decide update rates and importance of the pages. So even if the update rate seems low, you might still want to recrawl important pages quite often because, uh, the likelihood they change might be low, but the value of having updated is high.Shawn Wang [00:29:50]: Yeah, yeah, yeah, yeah. Uh, well, you know, yeah. This, uh, you know, mention of latency and, and saving things to this reminds me of one of your classics, which I have to bring up, which is latency numbers. Every programmer should know, uh, was there a, was it just a, just a general story behind that? Did you like just write it down?Jeff Dean [00:30:06]: I mean, this has like sort of eight or 10 different kinds of metrics that are like, how long does a cache mistake? How long does branch mispredict take? How long does a reference domain memory take? How long does it take to send, you know, a packet from the U S to the Netherlands or something? Um,Shawn Wang [00:30:21]: why Netherlands, by the way, or is it, is that because of Chrome?Jeff Dean [00:30:25]: Uh, we had a data center in the Netherlands, um, so, I mean, I think this gets to the point of being able to do the back of the envelope calculations. So these are sort of the raw ingredients of those, and you can use them to say, okay, well, if I need to design a system to do image search and thumb nailing or something of the result page, you know, how, what I do that I could pre-compute the image thumbnails. I could like. Try to thumbnail them on the fly from the larger images. What would that do? How much dis bandwidth than I need? How many des seeks would I do? Um, and you can sort of actually do thought experiments in, you know, 30 seconds or a minute with the sort of, uh, basic, uh, basic numbers at your fingertips. Uh, and then as you sort of build software using higher level libraries, you kind of want to develop the same intuitions for how long does it take to, you know, look up something in this particular kind of.Shawn Wang [00:31:21]: I'll see you next time.Shawn Wang [00:31:51]: Which is a simple byte conversion. That's nothing interesting. I wonder if you have any, if you were to update your...Jeff Dean [00:31:58]: I mean, I think it's really good to think about calculations you're doing in a model, either for training or inference.Jeff Dean [00:32:09]: Often a good way to view that is how much state will you need to bring in from memory, either like on-chip SRAM or HBM from the accelerator. Attached memory or DRAM or over the network. And then how expensive is that data motion relative to the cost of, say, an actual multiply in the matrix multiply unit? And that cost is actually really, really low, right? Because it's order, depending on your precision, I think it's like sub one picodule.Shawn Wang [00:32:50]: Oh, okay. You measure it by energy. Yeah. Yeah.Jeff Dean [00:32:52]: Yeah. I mean, it's all going to be about energy and how do you make the most energy efficient system. And then moving data from the SRAM on the other side of the chip, not even off the off chip, but on the other side of the same chip can be, you know, a thousand picodules. Oh, yeah. And so all of a sudden, this is why your accelerators require batching. Because if you move, like, say, the parameter of a model from SRAM on the, on the chip into the multiplier unit, that's going to cost you a thousand picodules. So you better make use of that, that thing that you moved many, many times with. So that's where the batch dimension comes in. Because all of a sudden, you know, if you have a batch of 256 or something, that's not so bad. But if you have a batch of one, that's really not good.Shawn Wang [00:33:40]: Yeah. Yeah. Right.Jeff Dean [00:33:41]: Because then you paid a thousand picodules in order to do your one picodule multiply.Shawn Wang [00:33:46]: I have never heard an energy-based analysis of batching.Jeff Dean [00:33:50]: Yeah. I mean, that's why people batch. Yeah. Ideally, you'd like to use batch size one because the latency would be great.Shawn Wang [00:33:56]: The best latency.Jeff Dean [00:33:56]: But the energy cost and the compute cost inefficiency that you get is quite large. So, yeah.Shawn Wang [00:34:04]: Is there a similar trick like, like, like you did with, you know, putting everything in memory? Like, you know, I think obviously NVIDIA has caused a lot of waves with betting very hard on SRAM with Grok. I wonder if, like, that's something that you already saw with, with the TPUs, right? Like that, that you had to. Uh, to serve at your scale, uh, you probably sort of saw that coming. Like what, what, what hardware, uh, innovations or insights were formed because of what you're seeing there?Jeff Dean [00:34:33]: Yeah. I mean, I think, you know, TPUs have this nice, uh, sort of regular structure of 2D or 3D meshes with a bunch of chips connected. Yeah. And each one of those has HBM attached. Um, I think for serving some kinds of models, uh, you know, you, you pay a lot higher cost. Uh, and time latency, um, bringing things in from HBM than you do bringing them in from, uh, SRAM on the chip. So if you have a small enough model, you can actually do model parallelism, spread it out over lots of chips and you actually get quite good throughput improvements and latency improvements from doing that. And so you're now sort of striping your smallish scale model over say 16 or 64 chips. Uh, but as if you do that and it all fits in. In SRAM, uh, that can be a big win. So yeah, that's not a surprise, but it is a good technique.Alessio Fanelli [00:35:27]: Yeah. What about the TPU design? Like how much do you decide where the improvements have to go? So like, this is like a good example of like, is there a way to bring the thousand picojoules down to 50? Like, is it worth designing a new chip to do that? The extreme is like when people say, oh, you should burn the model on the ASIC and that's kind of like the most extreme thing. How much of it? Is it worth doing an hardware when things change so quickly? Like what was the internal discussion? Yeah.Jeff Dean [00:35:57]: I mean, we, we have a lot of interaction between say the TPU chip design architecture team and the sort of higher level modeling, uh, experts, because you really want to take advantage of being able to co-design what should future TPUs look like based on where we think the sort of ML research puck is going, uh, in some sense, because, uh, you know, as a hardware designer for ML and in particular, you're trying to design a chip starting today and that design might take two years before it even lands in a data center. And then it has to sort of be a reasonable lifetime of the chip to take you three, four or five years. So you're trying to predict two to six years out where, what ML computations will people want to run two to six years out in a very fast changing field. And so having people with interest. Interesting ML research ideas of things we think will start to work in that timeframe or will be more important in that timeframe, uh, really enables us to then get, you know, interesting hardware features put into, you know, TPU N plus two, where TPU N is what we have today.Shawn Wang [00:37:10]: Oh, the cycle time is plus two.Jeff Dean [00:37:12]: Roughly. Wow. Because, uh, I mean, sometimes you can squeeze some changes into N plus one, but, you know, bigger changes are going to require the chip. Yeah. Design be earlier in its lifetime design process. Um, so whenever we can do that, it's generally good. And sometimes you can put in speculative features that maybe won't cost you much chip area, but if it works out, it would make something, you know, 10 times as fast. And if it doesn't work out, well, you burned a little bit of tiny amount of your chip area on that thing, but it's not that big a deal. Uh, sometimes it's a very big change and we want to be pretty sure this is going to work out. So we'll do like lots of carefulness. Uh, ML experimentation to show us, uh, this is actually the, the way we want to go. Yeah.Alessio Fanelli [00:37:58]: Is there a reverse of like, we already committed to this chip design so we can not take the model architecture that way because it doesn't quite fit?Jeff Dean [00:38:06]: Yeah. I mean, you, you definitely have things where you're going to adapt what the model architecture looks like so that they're efficient on the chips that you're going to have for both training and inference of that, of that, uh, generation of model. So I think it kind of goes both ways. Um, you know, sometimes you can take advantage of, you know, lower precision things that are coming in a future generation. So you can, might train it at that lower precision, even if the current generation doesn't quite do that. Mm.Shawn Wang [00:38:40]: Yeah. How low can we go in precision?Jeff Dean [00:38:43]: Because people are saying like ternary is like, uh, yeah, I mean, I'm a big fan of very low precision because I think that gets, that saves you a tremendous amount of time. Right. Because it's picojoules per bit that you're transferring and reducing the number of bits is a really good way to, to reduce that. Um, you know, I think people have gotten a lot of luck, uh, mileage out of having very low bit precision things, but then having scaling factors that apply to a whole bunch of, uh, those, those weights. Scaling. How does it, how does it, okay.Shawn Wang [00:39:15]: Interesting. You, so low, low precision, but scaled up weights. Yeah. Huh. Yeah. Never considered that. Yeah. Interesting. Uh, w w while we're on this topic, you know, I think there's a lot of, um, uh, this, the concept of precision at all is weird when we're sampling, you know, uh, we just, at the end of this, we're going to have all these like chips that I'll do like very good math. And then we're just going to throw a random number generator at the start. So, I mean, there's a movement towards, uh, energy based, uh, models and processors. I'm just curious if you've, obviously you've thought about it, but like, what's your commentary?Jeff Dean [00:39:50]: Yeah. I mean, I think. There's a bunch of interesting trends though. Energy based models is one, you know, diffusion based models, which don't sort of sequentially decode tokens is another, um, you know, speculative decoding is a way that you can get sort of an equivalent, very small.Shawn Wang [00:40:06]: Draft.Jeff Dean [00:40:07]: Batch factor, uh, for like you predict eight tokens out and that enables you to sort of increase the effective batch size of what you're doing by a factor of eight, even, and then you maybe accept five or six of those tokens. So you get. A five, a five X improvement in the amortization of moving weights, uh, into the multipliers to do the prediction for the, the tokens. So these are all really good techniques and I think it's really good to look at them from the lens of, uh, energy, real energy, not energy based models, um, and, and also latency and throughput, right? If you look at things from that lens, that sort of guides you to. Two solutions that are gonna be, uh, you know, better from, uh, you know, being able to serve larger models or, you know, equivalent size models more cheaply and with lower latency.Shawn Wang [00:41:03]: Yeah. Well, I think, I think I, um, it's appealing intellectually, uh, haven't seen it like really hit the mainstream, but, um, I do think that, uh, there's some poetry in the sense that, uh, you know, we don't have to do, uh, a lot of shenanigans if like we fundamentally. Design it into the hardware. Yeah, yeah.Jeff Dean [00:41:23]: I mean, I think there's still a, there's also sort of the more exotic things like analog based, uh, uh, computing substrates as opposed to digital ones. Uh, I'm, you know, I think those are super interesting cause they can be potentially low power. Uh, but I think you often end up wanting to interface that with digital systems and you end up losing a lot of the power advantages in the digital to analog and analog to digital conversions. You end up doing, uh, at the sort of boundaries. And periphery of that system. Um, I still think there's a tremendous distance we can go from where we are today in terms of energy efficiency with sort of, uh, much better and specialized hardware for the models we care about.Shawn Wang [00:42:05]: Yeah.Alessio Fanelli [00:42:06]: Um, any other interesting research ideas that you've seen, or like maybe things that you cannot pursue a Google that you would be interested in seeing researchers take a step at, I guess you have a lot of researchers. Yeah, I guess you have enough, but our, our research.Jeff Dean [00:42:21]: Our research portfolio is pretty broad. I would say, um, I mean, I think, uh, in terms of research directions, there's a whole bunch of, uh, you know, open problems and how do you make these models reliable and able to do much longer, kind of, uh, more complex tasks that have lots of subtasks. How do you orchestrate, you know, maybe one model that's using other models as tools in order to sort of build, uh, things that can accomplish, uh, you know, much more. Yeah. Significant pieces of work, uh, collectively, then you would ask a single model to do. Um, so that's super interesting. How do you get more verifiable, uh, you know, how do you get RL to work for non-verifiable domains? I think it's a pretty interesting open problem because I think that would broaden out the capabilities of the models, the improvements that you're seeing in both math and coding. Uh, if we could apply those to other less verifiable domains, because we've come up with RL techniques that actually enable us to do that. Uh, effectively, that would, that would really make the models improve quite a lot. I think.Alessio Fanelli [00:43:26]: I'm curious, like when we had Noam Brown on the podcast, he said, um, they already proved you can do it with deep research. Um, you kind of have it with AI mode in a way it's not verifiable. I'm curious if there's any thread that you think is interesting there. Like what is it? Both are like information retrieval of JSON. So I wonder if it's like the retrieval is like the verifiable part. That you can score or what are like, yeah, yeah. How, how would you model that, that problem?Jeff Dean [00:43:55]: Yeah. I mean, I think there are ways of having other models that can evaluate the results of what a first model did, maybe even retrieving. Can you have another model that says, is this things, are these things you retrieved relevant? Or can you rate these 2000 things you retrieved to assess which ones are the 50 most relevant or something? Um, I think those kinds of techniques are actually quite effective. Sometimes I can even be the same model, just prompted differently to be a, you know, a critic as opposed to a, uh, actual retrieval system. Yeah.Shawn Wang [00:44:28]: Um, I do think like there, there is that, that weird cliff where like, it feels like we've done the easy stuff and then now it's, but it always feels like that every year. It's like, oh, like we know, we know, and the next part is super hard and nobody's figured it out. And, uh, exactly with this RLVR thing where like everyone's talking about, well, okay, how do we. the next stage of the non-verifiable stuff. And everyone's like, I don't know, you know, Ellen judge.Jeff Dean [00:44:56]: I mean, I feel like the nice thing about this field is there's lots and lots of smart people thinking about creative solutions to some of the problems that we all see. Uh, because I think everyone sort of sees that the models, you know, are great at some things and they fall down around the edges of those things and, and are not as capable as we'd like in those areas. And then coming up with good techniques and trying those. And seeing which ones actually make a difference is sort of what the whole research aspect of this field is, is pushing forward. And I think that's why it's super interesting. You know, if you think about two years ago, we were struggling with GSM, eight K problems, right? Like, you know, Fred has two rabbits. He gets three more rabbits. How many rabbits does he have? That's a pretty far cry from the kinds of mathematics that the models can, and now you're doing IMO and Erdos problems in pure language. Yeah. Yeah. Pure language. So that is a really, really amazing jump in capabilities in, you know, in a year and a half or something. And I think, um, for other areas, it'd be great if we could make that kind of leap. Uh, and you know, we don't exactly see how to do it for some, some areas, but we do see it for some other areas and we're going to work hard on making that better. Yeah.Shawn Wang [00:46:13]: Yeah.Alessio Fanelli [00:46:14]: Like YouTube thumbnail generation. That would be very helpful. We need that. That would be AGI. We need that.Shawn Wang [00:46:20]: That would be. As far as content creators go.Jeff Dean [00:46:22]: I guess I'm not a YouTube creator, so I don't care that much about that problem, but I guess, uh, many people do.Shawn Wang [00:46:27]: It does. Yeah. It doesn't, it doesn't matter. People do judge books by their covers as it turns out. Um, uh, just to draw a bit on the IMO goal. Um, I'm still not over the fact that a year ago we had alpha proof and alpha geometry and all those things. And then this year we were like, screw that we'll just chuck it into Gemini. Yeah. What's your reflection? Like, I think this, this question about. Like the merger of like symbolic systems and like, and, and LMS, uh, was a very much core belief. And then somewhere along the line, people would just said, Nope, we'll just all do it in the LLM.Jeff Dean [00:47:02]: Yeah. I mean, I think it makes a lot of sense to me because, you know, humans manipulate symbols, but we probably don't have like a symbolic representation in our heads. Right. We have some distributed representation that is neural net, like in some way of lots of different neurons. And activation patterns firing when we see certain things and that enables us to reason and plan and, you know, do chains of thought and, you know, roll them back now that, that approach for solving the problem doesn't seem like it's going to work. I'm going to try this one. And, you know, in a lot of ways we're emulating what we intuitively think, uh, is happening inside real brains in neural net based models. So it never made sense to me to have like completely separate. Uh, discrete, uh, symbolic things, and then a completely different way of, of, uh, you know, thinking about those things.Shawn Wang [00:47:59]: Interesting. Yeah. Uh, I mean, it's maybe seems obvious to you, but it wasn't obvious to me a year ago. Yeah.Jeff Dean [00:48:06]: I mean, I do think like that IMO with, you know, translating to lean and using lean and then the next year and also a specialized geometry model. And then this year switching to a single unified model. That is roughly the production model with a little bit more inference budget, uh, is actually, you know, quite good because it shows you that the capabilities of that general model have improved dramatically and, and now you don't need the specialized model. This is actually sort of very similar to the 2013 to 16 era of machine learning, right? Like it used to be, people would train separate models for lots of different, each different problem, right? I have, I want to recognize street signs and something. So I train a street sign. Recognition recognition model, or I want to, you know, decode speech recognition. I have a speech model, right? I think now the era of unified models that do everything is really upon us. And the question is how well do those models generalize to new things they've never been asked to do and they're getting better and better.Shawn Wang [00:49:10]: And you don't need domain experts. Like one of my, uh, so I interviewed ETA who was on, who was on that team. Uh, and he was like, yeah, I, I don't know how they work. I don't know where the IMO competition was held. I don't know the rules of it. I just trained the models, the training models. Yeah. Yeah. And it's kind of interesting that like people with these, this like universal skill set of just like machine learning, you just give them data and give them enough compute and they can kind of tackle any task, which is the bitter lesson, I guess. I don't know. Yeah.Jeff Dean [00:49:39]: I mean, I think, uh, general models, uh, will win out over specialized ones in most cases.Shawn Wang [00:49:45]: Uh, so I want to push there a bit. I think there's one hole here, which is like, uh. There's this concept of like, uh, maybe capacity of a model, like abstractly a model can only contain the number of bits that it has. And, uh, and so it, you know, God knows like Gemini pro is like one to 10 trillion parameters. We don't know, but, uh, the Gemma models, for example, right? Like a lot of people want like the open source local models that are like that, that, that, and, and, uh, they have some knowledge, which is not necessary, right? Like they can't know everything like, like you have the. The luxury of you have the big model and big model should be able to capable of everything. But like when, when you're distilling and you're going down to the small models, you know, you're actually memorizing things that are not useful. Yeah. And so like, how do we, I guess, do we want to extract that? Can we, can we divorce knowledge from reasoning, you know?Jeff Dean [00:50:38]: Yeah. I mean, I think you do want the model to be most effective at reasoning if it can retrieve things, right? Because having the model devote precious parameter space. To remembering obscure facts that could be looked up is actually not the best use of that parameter space, right? Like you might prefer something that is more generally useful in more settings than this obscure fact that it has. Um, so I think that's always attention at the same time. You also don't want your model to be kind of completely detached from, you know, knowing stuff about the world, right? Like it's probably useful to know how long the golden gate be. Bridges just as a general sense of like how long are bridges, right? And, uh, it should have that kind of knowledge. It maybe doesn't need to know how long some teeny little bridge in some other more obscure part of the world is, but, uh, it does help it to have a fair bit of world knowledge and the bigger your model is, the more you can have. Uh, but I do think combining retrieval with sort of reasoning and making the model really good at doing multiple stages of retrieval. Yeah.Shawn Wang [00:51:49]: And reasoning through the intermediate retrieval results is going to be a, a pretty effective way of making the model seem much more capable, because if you think about, say, a personal Gemini, yeah, right?Jeff Dean [00:52:01]: Like we're not going to train Gemini on my email. Probably we'd rather have a single model that, uh, we can then use and use being able to retrieve from my email as a tool and have the model reason about it and retrieve from my photos or whatever, uh, and then make use of that and have multiple. Um, you know, uh, stages of interaction. that makes sense.Alessio Fanelli [00:52:24]: Do you think the vertical models are like, uh, interesting pursuit? Like when people are like, oh, we're building the best healthcare LLM, we're building the best law LLM, are those kind of like short-term stopgaps or?Jeff Dean [00:52:37]: No, I mean, I think, I think vertical models are interesting. Like you want them to start from a pretty good base model, but then you can sort of, uh, sort of viewing them, view them as enriching the data. Data distribution for that particular vertical domain for healthcare, say, um, we're probably not going to train or for say robotics. We're probably not going to train Gemini on all possible robotics data. We, you could train it on because we want it to have a balanced set of capabilities. Um, so we'll expose it to some robotics data, but if you're trying to build a really, really good robotics model, you're going to want to start with that and then train it on more robotics data. And then maybe that would. It's multilingual translation capability, but improve its robotics capabilities. And we're always making these kind of, uh, you know, trade-offs in the data mix that we train the base Gemini models on. You know, we'd love to include data from 200 more languages and as much data as we have for those languages, but that's going to displace some other capabilities of the model. It won't be as good at, um, you know, Pearl programming, you know, it'll still be good at Python programming. Cause we'll include it. Enough. Of that, but there's other long tail computer languages or coding capabilities that it may suffer on or multi, uh, multimodal reasoning capabilities may suffer. Cause we didn't get to expose it to as much data there, but it's really good at multilingual things. So I, I think some combination of specialized models, maybe more modular models. So it'd be nice to have the capability to have those 200 languages, plus this awesome robotics model, plus this awesome healthcare, uh, module that all can be knitted together to work in concert and called upon in different circumstances. Right? Like if I have a health related thing, then it should enable using this health module in conjunction with the main base model to be even better at those kinds of things. Yeah.Shawn Wang [00:54:36]: Installable knowledge. Yeah.Jeff Dean [00:54:37]: Right.Shawn Wang [00:54:38]: Just download as a, as a package.Jeff Dean [00:54:39]: And some of that installable stuff can come from retrieval, but some of it probably should come from preloaded training on, you know, uh, a hundred billion tokens or a trillion tokens of health data. Yeah.Shawn Wang [00:54:51]: And for listeners, I think, uh, I will highlight the Gemma three end paper where they, there was a little bit of that, I think. Yeah.Alessio Fanelli [00:54:56]: Yeah. I guess the question is like, how many billions of tokens do you need to outpace the frontier model improvements? You know, it's like, if I have to make this model better healthcare and the main. Gemini model is still improving. Do I need 50 billion tokens? Can I do it with a hundred, if I need a trillion healthcare tokens, it's like, they're probably not out there that you don't have, you know, I think that's really like the.Jeff Dean [00:55:21]: Well, I mean, I think healthcare is a particularly challenging domain, so there's a lot of healthcare data that, you know, we don't have access to appropriately, but there's a lot of, you know, uh, healthcare organizations that want to train models on their own data. That is not public healthcare data, uh, not public health. But public healthcare data. Um, so I think there are opportunities there to say, partner with a large healthcare organization and train models for their use that are going to be, you know, more bespoke, but probably, uh, might be better than a general model trained on say, public data. Yeah.Shawn Wang [00:55:58]: Yeah. I, I believe, uh, by the way, also this is like somewhat related to the language conversation. Uh, I think one of your, your favorite examples was you can put a low resource language in the context and it just learns. Yeah.Jeff Dean [00:56:09]: Oh, yeah, I think the example we used was Calamon, which is truly low resource because it's only spoken by, I think 120 people in the world and there's no written text.Shawn Wang [00:56:20]: So, yeah. So you can just do it that way. Just put it in the context. Yeah. Yeah. But I think your whole data set in the context, right.Jeff Dean [00:56:27]: If you, if you take a language like, uh, you know, Somali or something, there is a fair bit of Somali text in the world that, uh, or Ethiopian Amharic or something, um, you know, we probably. Yeah. Are not putting all the data from those languages into the Gemini based training. We put some of it, but if you put more of it, you'll improve the capabilities of those models.Shawn Wang [00:56:49]: Yeah.Jeff Dean [00:56:49]: