Podcasts about u s

The USDA's World Agricultural Supply and Demand Estimates report shows the 2025-26 U.S. corn outlook is for greater exports and lower ending stocks, and the USDA is facing doubts about the reliability of its data from farmers, grain traders, and economists.

Enjoy this special feed drop of our sister show 'In This Economy?!'Canadian restaurant operators could be in for another rough year – thanks to uncertainty around trade with the U-S and higher food costs – along with a few other factors.And the industry, which has been struggling for years, is important to the country's economy. But the numbers are not trending in the right direction.Hope is not lost.Host Kris McCusker takes a closer look at a report from Restaurants Canada with the President and CEO, Kelly Higginson. We love feedback at The Big Story, as well as suggestions for future episodes. You can find us:Through email at hello@thebigstorypodcast.ca Or @thebigstory.bsky.social on Bluesky

X: @GarrettInExile @americasrt1776 @ileaderssummit @NatashaSrdoc @JoelAnandUSA @supertalk Join America's Roundtable radio co-hosts Natasha Srdoc and Joel Anand Samy with the Honorable Thomas Garrett, Jr., member of the Commonwealth of Virginia's House of Delegates and former US Congressman. The conversation will focus on the state of America's economy, Trump's economic reforms, US-Iran Talks, America's ties with Israel, Virginia's radical changes with major tax hikes and sweeping gun control legislation and a new redistricting initiative which may leave Virginia's Congressional delegation with a 10-1 in favor of Democrats rather than the current 6-5 Republican edge. This could pose challenges in the mid-term elections.a leading attorney, currently serving as a legislator in the Commonwealth of Virginia, former Assistant Attorney General and former U.S. Congressman. The Washington Post's Editorial Board: "Brass-knuckled hypocrisy in Virginia" Quote: _The self-styled democracy party isn't behaving democratically. Democrats in Richmond are trying to effectively disenfranchise millions of Virginians by redrawing congressional maps to give themselves 10 of the commonwealth's 11 House seats — giving Democrats control of 91 percent of House seats in a state where Republicans lost the last presidential election by just six points. Most know better, including the governor. Abigail Spanberger was among the two-thirds of Virginians who voted in 2020 to transfer once-a-decade redistricting from the legislature to a bipartisan commission. “Gerrymandering is detrimental to our democracy,” she said back then. On Friday, Spanberger signed a bill to schedule an April 21 referendum that would move it back. The governor said it was necessary “to let voters respond to extreme measures taken by other states.”_ Bio | Tom Garrett Virginia Delegate Tom Garrett earned his Bachelor's degree from the University of Richmond. After the University of Richmond, Tom Garrett became an artillery officer in the United States Army. Achieving the rank of Captain, Tom led soldiers overseas—most notably while deployed in Bosnia. Upon returning to the States, Tom earned his J.D. from the University of Richmond and quickly became an Assistant Attorney General for Virginia. In 2016, Tom was elected to represent Virginia's 5th Congressional District in the United States House of Representatives. Tom won that election with the most votes ever in the 5th Congressional District. While in Congress, Tom served on the Foreign Affairs Committee, Homeland Security Committee, Education and Workforce Committee and was a member of the Freedom Caucus. An expert on Iran and the Middle East, Tom Garrett's analysis and insights are enlightening as America's foreign policy and national security concerns are focused on a strategic region adversely impacted by Iran, a state sponsor of terrorism. In the years since, Tom Garrett has dedicated his life to fighting for the oppressed and forgotten not only here in America, but around the world. Tom has been working on a global docu-series project, Exile, which tells the untold stories of those who are persecuted based on their faith or ethnicity. In addition to continuing his work as a defense attorney, Tom has served as a consultant and most recently, cofounder for firms working in global energy development. americasrt.com https://summitleadersusa.com/ | https://jerusalemleaderssummit.com/ America's Roundtable on Apple Podcasts: https://podcasts.apple.com/us/podcast/americas-roundtable/id1518878472 X: @GarrettInExile @americasrt1776 @ileaderssummit @NatashaSrdoc @JoelAnandUSA @supertalk America's Roundtable is co-hosted by Natasha Srdoc and Joel Anand Samy, co-founders of International Leaders Summit and the Jerusalem Leaders Summit. America's Roundtable radio program focuses on America's economy, healthcare reform, rule of law, security and trade, and its strategic partnership with rule of law nations around the world. The radio program features high-ranking US administration officials, cabinet members, members of Congress, state government officials, distinguished diplomats, business and media leaders and influential thinkers from around the world. Tune into America's Roundtable Radio program from Washington, DC via live streaming on Saturday mornings via 68 radio stations at 7:30 A.M. (ET) on Lanser Broadcasting Corporation covering the Michigan and the Midwest market, and at 7:30 A.M. (CT) on SuperTalk Mississippi — SuperTalk.FM reaching listeners in every county within the State of Mississippi, and neighboring states in the South including Alabama, Arkansas, Louisiana and Tennessee. Tune into WTON in Central Virginia on Sunday mornings at 6:00 A.M. (ET). Listen to America's Roundtable on digital platforms including Apple Podcasts, Spotify, Amazon, Google and other key online platforms. Listen live, Saturdays at 7:30 A.M. (CT) on SuperTalk | https://www.supertalk.fm

AI disruption fears hit equity markets once again, sending the S&P 500 and Nasdaq lower with transport, commercial real estate and software stocks all under pressure. Meanwhile, global leaders gather for the Munich Security Conference as U.S. allies in Europe look to chart a more independent course. And in corporate news, L'Oreal Q4 sales miss expectations as strong numbers out of the U.S. and Europe fail to offset weakness in China, sending U.S.-listed shares sharply lower, but CEO Nicolas Hieronimus tells CNBC he's confident the French beauty giant will bounce back.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

John Mesko, CEO at the Potato Sustainability Alliance, says they help improve the economic, environmental and social aspects of potato production in the U.S. and Canada.

Canadian restaurant operators could be in for another rough year – thanks to uncertainty around trade with the U-S and higher food costs – along with a few other factors.And the industry, which has been struggling for years, is important to the country's economy. But the numbers are not trending in the right direction.Hope is not lost.Host Kris McCusker takes a closer look at a report from Restaurants Canada with the President and CEO, Kelly Higginson. Do you have a topic that's confounding you in this economy? We'll be happy to dig into it for you and get you the answers you need. Email us at: rogerspodcastnetwork@rci.rogers.com. Thank you for listening!

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]:

Pakistan's Khawaja Asif said,'US used & dumped us like toilet paper'. He was speaking about Pakistan's terror problem and the country's involvement in Afghanistan conflict. #CutTheClutter Episode 1795 looks at Pakistani Defence Minister's comments in detail, and the country's Afghanistan dilemma, in the backdrop of escalating tensions. ThePrint Editor-In-Chief Shekhar Gupta also takes you back to 2011, when he had explained why India should stay away from Afghanistan after US withdrawal, and leave it to Pakistan.

Good afternoon, I'm _____ with today's episode of EZ News. Taipei signs NT$12.2 billion Nvidia deal The Taipei City Government has signed a 12.2-billion N-T agreement with Nvidia for the T-17 and T-18 plots of land at the Beitou-Shilin Technology Park. Mayor Chiang Wan-an says construction on the U-S technology company's new Taiwan headquarters could now begin as early as June. The deal grants Nvidia 50-year land surface rights, extendable by up to 20 years and Chiang is describing the agreement as a major milestone and is also pledging (保證) full municipal support. Under Nvidia's investment plan, more than 40-billion N-T will be spent during construction, and operations are expected to create over 10,000 jobs. Nvidia C-E-O Jensen Huang is scheduled to visit Taiwan from June 2 to 5 for Computex Taipei. TSMC shares surge almost 69% in 'Year of Snake' Shares in Taiwan Semiconductor Manufacturing soared almost 69-per cent in the Year of the Snake. T-S-M-C shares closed at a record high of 1,915 N-T on Wednesday - the final trading day of the Year of the Snake - despite coming off off a historic intraday high of 1,925 N-T early in the session. The chipmaker's share price jumped 780 N-T during the Year of the Snake and contributed about 6,262 points to the Tai-Ex's overall rise for the year. The strong showing raised T-S-M-C's market capitalization by more than 20-trillion N-T during the year. Taipei Zoo to renovate its tropical rainforest area And, The Taipei City Zoo has announced plans to renovate its tropical rainforest area. According to the zoo,the renovations will improve animal welfare, update some of its aging facilities and are in line with (與…一致,符合,按照) international trends in professional zoo development. The Asian tropical rainforest area opened in 1997. The zoo will be relocating its Asian elephants Yu-Hsin and Yu-Kai to the African animal area later this year during the renovation period. The zoo will also be transfering other animals, such as the Malayan tiger, sun bear, and leopard, following the elephants' move. Republicans join Democrats to rebuke Trump's Tariffs on Canada US lawmakers have voted to terminate President Donald Trump's tariffs on Canada - with six Republicans joining Democrats in the House of Representatives. The resolution passed 219 to 211 - sending the bill to the Senate where a similar (相似的) measure passed last year. Toni Waterman has more. Sci Initiative Finds HumanCaused Climate Change Drove Wildfires A team of researchers say that human-caused climate change had an important impact on the recent wildfires that engulfed parts of Chile and Argentina's Patagonia region. That's according to a report released Wednesday by World Weather Attribution, a scientific initiative that investigates extreme weather events. The report says the hot, dry and gusty weather that fed last month's deadly wildfires in central and southern Chile was made around 200% more likely by human-made greenhouse gas emissions. Ande the high-fire-risk conditions that fueled (激起，助長) the blazes still racing through southern Argentina were made 150% more likely. World Weather Attribution says “These trends are projected to continue in the future as long as we continue to burn fossil fuels.” That was the I.C.R.T. EZ News, I'm _____. －－－－以下為 SoundOn 動態廣告－－－－ 新感覺夾心土司 多種口味隨心挑選 讓你隨時隨地都有好心情 甜蜜口感草莓夾心、顆粒層次花生夾心、濃郁滑順可可夾心 主廚監製鮪魚沙拉、精選原料金黃蛋沙拉 輕巧美味帶著走，迎接多變的每一天 7-Eleven多種口味販售中 https://sofm.pse.is/8qdyge -- Hosting provided by SoundOn

WAC President, Michael Schadler says a major focus for them this year will be the U.S.-Mexico-Canada Agreement that is scheduled for an update in the coming months.

In episode 2005, Jack and Miles are joined by host of There Are No Girls on the Internet, Bridget Todd, to discuss… Difference Between Europe and The US, The Best We Can Do In The US Is A MAGA Senator Saying NOW I SEE WHAT THE BIG DEAL IS, “Penisgate” Is The Latest Olympics Cheating Scandal and more! The Epstein scandal is taking down Europe’s political class. In the US, they’re getting a pass. The Best We Can Do In The US Is A MAGA Senator Saying NOW I SEE WHAT THE BIG DEAL IS What US ski jumpers think about ‘wild’ penis-gate scandal at 2026 Winter Olympics Who is Anthony Ammirati? Meet the French pole vaulter whose ‘bulge’ cost him a medal Ski jumpers sceptical of penis injection reports Skisprung-Verband reagiert auf Penis-Wirbel Rumors Fly Claiming Olympic Ski Jumpers Are Injecting Their Genitals LISTEN: Slipping Into Darkness by The FunkeesSee omnystudio.com/listener for privacy information.

On this Salcedo Storm Podcast:JOE DIGENOVA IS A FORMER UNITED STATES ATTORNEY FOR WASHINGTON, D.C.  AND VICTORIA TOENSING IS THE FORMER CHIEF COUNSEL FOR THE SENATE INTELLIGENCE COMMITTEE.

The USDA reports that during a routine inspection in Florida, a horse arriving from Argentina was found with an open wound containing New World Screwworm larvae, and President Trump says the U.S. will cut tariffs on Indian goods to 18% after India agreed to stop purchasing Russian crude oil.

California farmers suffer major setback as Del Monte announced plans to close its only remaining cannery in the state, and President Trump says the U.S. will cut tariffs on Indian goods to 18% after India agreed to stop purchasing Russian crude oil.

Episode 375 of RevolutionZ has as guest Kathy Kelly. When journalists are barred and killed, doctors are targeted, and mountainous rubble hides unexploded ordnance, a society is violated twice—physically and narratively. Our guest, Kathy Kelly, connects what headlines obscure: how U.S. weapons shipments function as political green lights, how “ceasefire” rhetoric papers over daily violations, and how displacement in the West Bank is driven by soldiers, settlers, and a structure designed to make staying impossible.Kathy brings the human scale back into focus. From a makeshift white flag walk into Jenin to evenings with families in Gaza, she shares the intimate choices people make under siege—protecting elders, scavenging firewood, teaching children to read the sky for drones. These stories resist the flattening of body counts, revealing what war does to witnesses and perpetrators alike. Kathy explores how international law erodes when powerful states flout norms, why nuclear ambitions can spread under the guise of “civilian” programs, and how those choices ricochet into U.S. life through policing exchanges, PTSD, and the quiet normalization of force.Kathy also talks strategy. She tells how student encampments and divestment campaigns pried open university endowments and hedge fund ties. How cultural voices amplify names and memories that institutions try to erase. How growing activism keeps movements alive and oriented. Kathy reflects on practical commitments—from tax resistance to hospitality—that shift resources away from violence and toward care and building a revolution of values sturdy enough to change institutions: living more simply, sharing more fairly, ending the reflex to eliminate those who resist subordination to “national interests,” and actively organizing sustainable resistance. Her message: read and remember, organize locally, join boycott and divestment efforts, and align daily choices with the future you want. Support the show

Growing speculation over a proposal to hand Australians' personal and biometric data to the United States is raising alarm among privacy and security experts. As other countries with visa-free arrangements with the U-S consider proposals to expand their data-sharing agreements, the Australian Government is yet to clarify if they're considering the move.

Lecture par Marie Constant Entretien mené par Sophie Joubert « Le 22 février 1947, par temps d'orage, il arriva en train à Brême. C'est ce qu'il crut du moins. » Pendant les grands procès visant à éradiquer le nazisme dans la vie publique allemande, Jacob Lenz, avocat et capitaine de l'US army est convoqué par l'un de ces tribunaux afin de juger une affaire hors du commun. Des oiseaux parleurs – des mainates – nichant dans une forêt des environs, ont appris à chanter des hymnes nazis et les transmettent à leur descendance. Quel sort doit-on leur réserver ? Est-il, en quelque sorte, leur avocat ? À lire – Jean-Yves Jouannais, Une forêt, Albin Michel, 2026

;By Request: Steve Stockton's LESSER-KNOWN CRYPTIDS of the U.S. Vol 1-4. Written and narrated by Steve StocktonBecome a supporter of this podcast: https://www.spreaker.com/podcast/missing-persons-mysteries--5624803/support.

Republican gubernatorial candidate Stacy Garrity says an Allegheny County party leader is her choice for a running mate. The city of Philadelphia is headed to court, demanding the National Park Service return its exhibit depicting the lives of enslaved persons under President George Washington. The exhibit was recently removed by federal employees, acting on a White House executive order calling for the removal of displays in U-S national parks that "disparage" the nation. Authorities in Lebanon County have released the identity of the man fatally shot by State Police early Wednesday while troopers were attempting to serve a warrant. A new Pennsylvania law designed to keep unregulated vapes out of the hands of kids may not actually work as planned. PennDOT is announcing more than 300 new parking spaces are now available for truckers across Pennsylvania. And a deep dive: Pittsburgh’s oldest print newspaper is set to shut down in just a few months. According to the Nieman Lab, that would make Pittsburgh the largest city in the country without a real daily newspaper. Did you know that if every one of WITF’s sustaining circle members gives as little as $12 more a month, we'd close the gap caused by federal funding cuts? Increase your gift at https://witf.org/increase or become a new sustaining member at www.witf.org/givenow, and thanks!Support WITF: https://www.witf.org/support/give-now/See omnystudio.com/listener for privacy information.

The Justice Department has released its final tranche of the Epstein files - we'll look at what they have and haven't included. And, the U.S. in a partial government shutdown again, although this is one is expected to be shorter than the record-breaking shutdown that happened during the fall. Plus, another winter storm is hitting the U-S this weekend, this time, hitting parts of the Southeast.Learn more about sponsor message choices: podcastchoices.com/adchoicesNPR Privacy Policy

In this weekend's episode, three segments from this past week's Washington Journal. First: A conversation with Jillian Snider – a former law enforcement officer and senior fellow at the R Street Institute. We talk about ICE operations in Minneapolis – and best practices for law enforcement in the wake of another fatal shooting there. Then: Amid the fallout in Minnesota, President Trump tried to pivot back to the economy and efforts on affordability. We dig into the numbers with Natalie Baker of the Center for American Progress and Brittany Madni from the Economic Policy Innovation Center. Finally: President Trump may have backed off his threat to takeover Greenland – but relations are still frayed between the U-S and Europe. That conversation with Andrew Roth of The Guardian -- and Stefanie Bolzen of the German News channel VELT. Learn more about your ad choices. Visit megaphone.fm/adchoices

In episode 1997, Jack and Miles are joined by comedian, actor, writer, and host of Worse Than You, Mo Fry Pasic, to discuss… That’s why Teenage Mutant Ninja Goebbels Is In The Hot Seat... Not Under The Bus Yet, TikTok Isn’t The Only Tech Company Backing ICE, Melania Doc Watch and more! Daily Zeitgeist: Our 2000th Episode is Here!!!... Chuck E Cheese - Most Perfect Day (Rap Song) Scoop: Blame game erupts over Trump team's false claim Alex Pretti sought "massacre" Pressure grows on Stephen Miller after Alex Pretti killing but Trump unlikely to cut ties TikTok users say they can’t upload anti-ICE videos. The company blames tech issues TikTok is investigating why some users can't write 'Epstein' in messages The Trump-approved US TikTok is off to a rough start TikTok Says It’s Not Censoring ‘Free Palestine’ Comments. Users See Something Different TikTok now specifically tracks immigration status and gender identity TikTok alternative Skylight soars to 380K+ users after TikTok US deal finalized TikTok's new terms of service spark backlash, but experts say they're an industry standard Meta Is Blocking Links to ICE List on Facebook, Instagram, and Threads How Amazon Powers ICE’s Deportation Machine Apple Took Down These ICE-Tracking Apps. The Developers Aren't Giving Up SCOOP: Apple Made ICE Agents a Protected Class “Melania” Movie Popcorn Bucket Hits ebay for $29.99 — Can Be Purchased Without Actually Seeing the Film Fuming Melania Puts Trump Aides at Risk of Ouster: Wolff Melania Invites Host of Z-List Celebs to Vanity Doc Premiere LISTEN: Danger by The Lijadu Sisters and also check out NUR-D's music here: https://nurdrocks.com/See omnystudio.com/listener for privacy information.

'End of Times' Utah mom is nabbed in Croatia for kidnapping. Her terrified tots are trapped in an overseas orphanage - while their dad wages war to bring them home to the U-S. A sick Iowa mom is busted for trying to hawk her tot for a five-figure payday. Plus, bitter sports betters target a preschooler over her pigskin picks. Jennifer Gould reports. See omnystudio.com/listener for privacy information.

From a new logo to new messaging, Pears grown here in the Northwest and around the country are getting a new, fun, fresh and modern look for consumers to enjoy.

Ukrainian President Volodymyr Zelenskyy says a US document on security guarantees is "100 per cent ready" to be signed. He also indicated some progress was made at tri-lateral talks on the weekend between Ukraine, the US and Russia. But he called on the U-S and Europe to keep up pressure on Russia, through sanctions.

Foreign Options for US Citizens Summary: https://www.youtube.com/watch?v=d-Jnr3Go2Gg In this conversation, Frazer Rice of Next Vantage and Judi Galst of Henley and Partners discuss the increasing interest among U.S. citizens in exploring global mobility options amidst geopolitical chaos. We delve into the distinctions between residency and citizenship, the implications of U.S. taxation, and the motivations driving individuals to seek alternative living arrangements. The discussion also covers the potential for citizenship through ancestry, popular destinations for relocation, and investment opportunities in countries like New Zealand and Australia. Judi emphasizes the importance of understanding the legal and practical aspects of relocating, as well as the need for personal exploration before making significant decisions. Takeaways Interest in global mobility has surged among U.S. citizens. Many seek residency as an insurance policy rather than leaving the U.S. Understanding residency vs. citizenship is crucial for potential expatriates. Residency can lead to citizenship but often requires time and investment. Tax implications are complex; relocating should not be primarily for tax benefits. Ancestry can provide a pathway to citizenship in several countries. Popular destinations for U.S. citizens include Europe, the Caribbean, and New Zealand. Investment opportunities exist in countries like New Zealand and Australia. Emerging markets in South America and Asia are gaining attention. Practical steps include consulting experts and visiting potential countries. Chapters 00:00 Navigating Geopolitical Chaos: The Rise of Global Mobility 02:55 Understanding Residency vs. Citizenship: Key Differences 06:06 Tax Implications and Motivations for Seeking Alternatives 08:48 Exploring Ancestry-Based Citizenship: Opportunities and Challenges 11:54 Popular Destinations for U.S. Citizens: Europe, Caribbean, and Beyond 15:10 Investment Opportunities: New Zealand and Australia 17:59 Emerging Trends in South America and Asia 20:50 Practical Steps for U.S. Citizens Considering Relocation Transcript I’m Frazer Rice. We’re certainly living in crazy political times right now, and a lot of US citizens are worried about what’s happening here and abroad. And they’re starting to think about other residencies and citizenship options. I talked to Judy Gost at Henley and Partners about what is and isn’t possible on that front. By the end of this, you’re going to understand the locations that are interesting, the difference between residency and citizenship, and why that may matter as you make choices for your retirement and your location long-term, both for yourself and for your kids. Frazer Rice (00:00.874)Welcome aboard, Judy. Judi Galst (00:03.022)Thanks for having me. Frazer Rice (00:04.244)Well, we’re in the midst of a lot of geopolitical chaos, and I think you have seen and I’ve seen a lot of interest in United States citizens looking abroad for either places to live or other situations to either get away from the chaos or try to address some other needs in their lives. What is the state of the union? assume interest has ticked up. Judi Galst (00:27.874)Yes, I’ve seen more business than I could have ever predicted, but it’s not necessarily people that are leaving the United States. For the most part, most of the clients that I’m working with are doing it as an insurance policy. A lot of the conversations I have with a client start out with them saying, I don’t want to leave the United States, but I’m feeling unsettled and the way to mitigate the way that I’m feeling is to have options. So they want to understand what if I did want to have a guaranteed right to go live in another part of the world? What is available to me? How do I pursue this? How long will it take? Frazer Rice (01:08.434)And we’ll get into some of the technical aspects here, but one of the concepts is understanding the difference between being able to reside somewhere else and being a citizen of another country, and then how that interacts with being a citizen of the United States. Maybe take us through the comparison of residents versus citizenship. Judi Galst (01:28.748)Yeah, that’s actually a really important distinction. And it doesn’t mean that one is better than the other, but they do have different benefits. And so it’s important to understand the difference. So let’s start with residents. Residents doesn’t mean the ability to have a house in another country. It means the ability to reside legally in another country. So the US passport is very strong. You can go into a lot of different countries even without having a visa. But we can’t stay there forever. We have limits, for example, in Europe. We can go in for 90 days, but then we have to leave for 90 days before we can go back in for another 90 days. So if you become a legal resident of another country, you have the ability to live there unlimited for a certain period of time. Residency is not permanent unless there’s a path to permanent residency. So usually you’re going to have to renew it and there may be some conditions in order to maintain it. Now, how frequently you have to renew it is going to vary by the country. For example, in Greece, you can become a Greek resident via a golden visa and that is good for five years and you’ll renew for another five years. In Italy, it’s good for two years. Then you renew for another three years. In Portugal, it’s good for two years. Then you renew for another three years. And as I said, there could be conditions. So in Greece, you qualify via purchasing real estate. If you sell the real estate, you’re going to lose your golden visa, not be able to renew it. In Italy, you qualify via purchasing stock. Frazer Rice (02:51.925)Right. Judi Galst (02:55.945)If you sell the stock, you’re not going to be able to renew it. You can get some travel rights by being a resident. Usually this benefit is not as important to a U.S. person because we already have really good travel benefits with our U.S. passport. But it can often be a strategy for someone from a country with a weaker passport, say even someone living in the United States that has only a Chinese passport. If they want to go into Europe, they have to get a Schenken visa. So a strategy for them might be let me become a resident of say Greece and then I gain Schengen access. Not unlimited, but I get that 90 days out of 180 days. Finally, I would say that residency can have a path to citizenship. Usually it’s a pretty arduous path. For example, in Italy, you can become a resident. You have to live in the country of Italy for six months a year for 10 years before you’d be eligible to apply. In Greece, six months a year for seven years. But there is ultimately a path in most residency programs. Frazer Rice (03:56.755)So let’s dive into citizenship, which my predilection on that is that it’s a much more permanent component, but it’s also a much more difficult process in general. Judi Galst (04:05.646)It doesn’t necessarily have to be difficult. It really depends on what program you’re doing. But you’re right. It’s a guaranteed right. It’s very difficult for a country to take away someone’s citizenship. The other big difference is that you get a passport. So in addition to gaining the ability to live in the country that you’re a citizen of, you also get another travel document. So depending upon what treaties have been done between your country of citizenship and other countries, it may really improve your mobility. Again, U.S. passport is pretty strong. you’re U.S. passport holder, unless there’s something unexpected like a pandemic when borders close to Americans, you already have a good travel document. But it can be another mobility option. Perhaps you’re going into a country you don’t want to identify as a U.S. passport holder, or perhaps you have a weaker passport and you want to travel on a secondary citizenship passport that might improve your mobility. Where citizenship is particularly powerful is in Europe. Because if you become a citizen of one country in the European Union, you gain the right to reside and work in any country in Europe. Frazer Rice (05:11.104)And just to distinguish, how does that impact UK people after they Brexited? Judi Galst (05:16.942)Sadly, with Brexit, the UK is no longer part of the EU. So many people in the UK are quite upset about this because no, you’re not going to gain the ability as a citizen of an EU country to live in the UK, nor are citizens of the UK now able to live anywhere in the European Union as they were previously. Frazer Rice (05:36.992)So let’s apply this directly to US citizens. So US citizen taxed on worldwide wealth. Let’s start with that. sure because I just got a Twitter fight with somebody who said, well, if you’re crypto, you can move away and you’re not out of the system. I’m like, that’s just no. We’ll start with that. But taxed on worldwide wealth, good passport can travel, but there are limitations as far as how long you can stay in various countries, probably around Judi Galst (05:52.622)Mm-hmm. Frazer Rice (06:06.578)Investment options, land ownership, things like that, depending on it. Where are the benefits of that U.S. person looking for another place to either reside or gain citizenship? Judi Galst (06:20.312)Well, it’s not a tax benefit. You started out with taxes and I know when someone, a client calls and says, you know, can you tell me what my options are? I’m really sick of paying us taxes. I’m like, well, this isn’t the right call for you. Yeah. So, but it’s important to understand. It doesn’t mean you’re going to be double taxed because that is a misconception that many people have about whether they should pursue a strategy of alternative residents or citizenship, because unlike the U S and Eritrea, Frazer Rice (06:22.079)Right. Frazer Rice (06:30.08)Puerto Rico that that’s it. That’s your best bet if you’re gonna try if you’re gonna try to play games Judi Galst (06:49.774)Every other country in the world, you don’t automatically become a tax resident by being a legal resident or even by being a citizen. Usually, you’re not going to trigger tax residency unless you reside 183 days in another country, but there are some exceptions. Switzerland is 90 days. Some, like New Zealand, will say it’s 183 days, but in a 12-month period, not necessarily in a year. I’m not licensed to give tax advice, so I’m giving high-level answer to this question. But in general, just by pursuing an alternative residence or citizenship, there’s no tax consequences. And if you were to become a tax resident, many of the countries that we support programs in have treaties. So it doesn’t necessarily mean that you’re going to pay double tax, but it does mean it has to be looked at. If I am talking to a client and they really have full intention of relocating to another country, immediately I want them to have a local tax consultation, which I set up for them to understand what, if any, consequences they have to be aware of. Frazer Rice (07:50.322)And those consequences can change. did an episode probably about six months ago on the change in law in the UK. And it’s a different environment than it was even six months ago for people either going in or coming out of that country as it relates to their US intersection. So I think that the summary on all of that is, look, if you’re going there, A, don’t do it for tax purposes, B, If you’re going to do it, make sure you get local tax counsel because those relationships can be complicated and will affect your planning. Judi Galst (08:25.198)Let’s talk about why people are doing it because taxes is not the strategy. And I would say, and my clients are almost exclusively Americans. So why are people calling me about this? There’s really four key motivators that tend to come up in the conversation. The first is because they do want another mobility option. They kind of have some PTSD still from the pandemic. They remember that feeling. Frazer Rice (08:27.935)Mm. Judi Galst (08:48.226)We could all work remotely. You had the vacation house in Italy or you had the private plane and all of a sudden you couldn’t take advantage of it because all the borders are closed to you and we could only stay in the United States. So some people are just realizing there is some risk to having one mobility option and they want to have an alternative. But I would say 90 % of the conversations I have there’s some reference to a plan B. People are feeling unsettled for so many different reasons. You know, I talked to people whose family fled the Holocaust. It is literally in their DNA where their family thought it could never happen here. And that comes up in every conversation with them. But I have same sex, you know, couples, have transgender clients, I have people whose family lived in other countries where they saw the fall of democracy. And then I just have a lot of wealthy clients, and they’re diversifying their assets right now. And they want to diversify their mobility. They pay a lot of money in insurance and they say, Judy, this is just another line item. Frazer Rice (09:45.896)You Judi Galst (09:46.703)I’d say some are thinking not just about themselves, but they’re thinking about protecting generational opportunity and legacy. Some say, you know, I’m a student of history and yeah, maybe it’s going to take 10, 15, 20 years, but I’ve seen this happen before. And I want to know that my kids and my grandkids are going to have options to either live a life in another part of the world for cultural or educational opportunities or in a worst case scenario, because the U.S. isn’t where they actually want to be. And finally, I’d say it fits nicely in a diversification of asset strategy, which many, many people are thinking about right now. Maybe they don’t want to hold all their money in the United States. Maybe they don’t want to all their real estate in the United States. And there can be strategies that are separate from what I do in terms of opening bank accounts in Switzerland or Singapore or other parts of the world. But really, all the programs that I do require you to move some assets. You’re either investing in stock or venture capital or private equity or real estate. So it does complement a diversification of asset strategy. Frazer Rice (10:42.911)Cool, so let’s think about, we sort of beat the tax horse to death a little bit here, but relocating versus renouncing. And different things, know, people probably come up to you with questions, do I have to fully leave? Do I have to renounce my US citizenship? How does all of that Judi Galst (10:51.608)Mm-hmm. Judi Galst (10:58.222)Great questions. So I’ve never had a client renounce. The US right now does not limit the number of passports one can have or citizenships one can have or how many residences they can have. Now, there is a congressperson who has just decided he wants to introduce some sort of bill that’s going to eliminate dual citizenship for Americans, although most constitutional scholars feel that’s like dead on arrival. But I have to acknowledge that. So no, you don’t need to renounce. And frankly, if you have a lot of money, renouncing is quite complicated and expensive, and you need really good counsel to make that very, very significant decision. In terms of relocation, almost all of the programs that we support require little to no physical presence. You’re always going to probably have to go for biometrics and give fingerprints. But a lot of these programs, you don’t actually have to come back to that country again, except to renew it. So for people that really want it as a Plan B and have no intention of really going to live in another part of the world at this stage in their lives, there’s not an obligation for you to spend time in order to maintain the ability to live in another country if you so choose. Frazer Rice (12:08.017)One thing that comes up that people ask me about and I only vaguely understand it is the concept of being able to get citizenship via ancestry. Comes up with a lot of people of Irish descent, Germany and Austrian especially. What’s the state of that and how realistic is it across different countries? Judi Galst (12:15.993)Mm. Mm-hmm. Judi Galst (12:26.767)It’s very realistic. And in fact, I’m doing German citizenship for myself. So for anyone whose family fled due to Nazi persecution from Germany and Austria, you and all future generations are entitled to citizenship. And my friends are like, why do you want German passport? But first of all, my kids got it. So my kids can go now live and work in Europe if they want, which is great, tremendous optionality. If you remember, I said before, it’s not just Germany. It’s any country in the European Union. Frazer Rice (12:30.473)Okay. Frazer Rice (12:47.956)Right. Judi Galst (12:56.899)And it’s very affordable if you actually are entitled to it. At Henley and Partners, we have established relationships with experts, lawyers in several countries that specialize in citizenship by ancestry. It’s very complex. And every country has different rules about like, it was passed down on the mother’s side, or if there was a break in the bloodline, or if it was passed a certain generation, or if there was a name change, there’s a lot of complexity to it. But clients who think they may be eligible can contact us and we will have an assessment done. And if there is a case, we’ll refer them to someone that can help them through the process. And, you know, it can cost around 5,000, 7,500 euros versus I have clients getting EU citizenship through, you know, Malta and they’re 1.5 million out of pocket. So if you can qualify via Ancestry, I’d say certainly it’s worth considering. Frazer Rice (13:50.879)Terrific. Judi Galst (13:51.311)But don’t call me and say, like, I did 23andMe and I’m Irish. Because you do actually have to produce documents. Not a humongous list of documents, but you’re going to need naturalization certificates for the descendant. You’re going to need marriage certificates, birth certificates, and other documents. Frazer Rice (13:55.187)Ha ha ha! Frazer Rice (14:10.844)So there’s definitely an exercise involved with it, but if you can legitimately trace lineage, you may have a shot. So let’s talk about what jurisdictions are popular with United States citizens. We talked a little bit about Europe, and I’m sure there’s some, let’s call it, some that are easier than others. But then Caribbean, South America, Australia, New Zealand, maybe even Asia, what comes across your desk as being Judi Galst (14:14.094)Mm-mm. Exactly. Frazer Rice (14:40.488)more reasonable than others maybe. Judi Galst (14:43.246)So I’d say clients that I’m talking to are basically going in one of four different directions. One is Europe. For residency, we’re looking at Portugal, Greece, Italy, and Malta. Those are all great programs because they require little to no time in the country to maintain the residency rights. So for people that really have no intention of spending significant time in another country, they’re really good solutions. And for citizenship in Europe, there very limited options. There’s ancestry, which we just talked about. But the concept of citizenship by investment in Europe essentially was killed by the European Court of Justice in the spring of 2025. To give a little bit of explanation, Malta used to have a citizenship by investment program. And it basically said, do these three things, make a large gift to the Maltese economy, rent a property for six years and spend somewhere around 21 days in the country. And you will have a path. to citizenship in Malta, which is an EU country. And the EU hated it. They felt it was transactional, that the passport was being sold, and they felt that people were being granted citizenship that didn’t show a tie to the country. And when this court ruling came out and deemed Malta’s program illegal, it essentially killed citizenship by investment programs in Europe. So I don’t think you’re going to see any European Union country have a citizenship by investment program, nor any country that wants to join the EU have one. But many countries in Europe have provisions in their constitution that say, if you are an exceptional person that make an exceptional contribution to our country or to humanity, we have discretionary ability to grant you citizenship. And so there are some paths to citizenship via merit, specifically through Malta and Austria right now, as well as some other places. So that’s Europe, snapshot of Europe. Let’s talk a little bit about Caribbean, which you specifically brought up. Frazer Rice (16:35.581)Right. Judi Galst (16:40.862)So Caribbean is a path to citizenship. If you remember, said citizenship, lifelong, right? Not many countries have a path to citizenship. It’s very fast. It’s very affordable. What does it give you? So there are five countries in the Caribbean that have programs St. Kitts, Antigua, Grenada, Dominica, St. Lucia. It gives you citizenship in one of those countries. A passport, another passport that you can travel on. Right now, it’s pretty strong. You can go into Europe with it, the UK, Ireland, not unlimited, same as the US, limited amount of time. Although I’m not sure the strength of the Caribbean passports is always going to be. as strong as it is today. Europe doesn’t love these programs. And I wouldn’t be surprised if the Caribbean passports tend to get weaker. However, for a client that says to me, this is purely an insurance policy. I want to cover my kids and my kids are in their 20s because a lot of times these program kids are going to need their own investment if they’re over the age of 18 or 21. Caribbean wouldn’t be a bad place for us if we felt we wanted to get out of town for a little while. Frazer Rice (17:23.23)Sure. Judi Galst (17:50.031)The Caribbean’s a great solution for a very affordable amount, maybe 400,000 for family. You can get and make an investment in real estate that you can sell in five or seven years and your entire family can gain citizenship. So that’s Caribbean. I can pivot to something else that you want to ask a question. OK, so I actually love the program that New Zealand has out right now, especially for a high net worth person. Frazer Rice (18:05.342)Okay, no, let’s try Australia and New Zealand. Judi Galst (18:18.414)I think every high net worth person should do New Zealand. And for a couple of reasons. First of all, it’s purely investment driven. You have to move a lot of money. So it has to be for a high net worth person because they’re going to move three million US dollars to be invested in private equity, venture capital and private credit in New Zealand for around a three year period. And children up to the age of 25, provided that they’re single and not working full time can be included in that investment. There’s very little time that the family needs to spend in New Zealand. As soon as you move the money there, you gain the right to live unlimited in New Zealand. But the main applicant only has to do 21 days, and the other family members only have to enter and exit for one day in the first year. At the end of three years, provided you didn’t invest in things that have a longer holding period, but from an immigration perspective, you can liquidate your investment. And then you can become a permanent resident. So you have a lifelong right at any time to relocate to New Zealand, or you never have to go back again. English speaking, good healthcare, good education. You could have a life there, unlike I don’t think people really want to envision spending 10 years in the Caribbean. But 10 years in New Zealand, you know, there’s many industries and many things that you could be doing. And you could have a quality of life, maybe not akin to the United States, but good. So I love the New Zealand program. Australia used to have a citizenship by investment program. They do not have one any longer. There is a route that they extend to people, which they call sort of like a talent visa. So there are certain sectors that are important to Australia and they would very much like to attract talent in those sectors. Usually it’s younger talent. So when I’m talking to a client that’s over 55, it can be difficult to get you approved for it. But I’ve had people over 55 that have gotten approved. And if you have the background that Australia deems valuable, they’ll grant you a five-year visa for you and your family at no cost. Children have to be under the age of 18 or financially dependent up to age 23 to be included. But this is a visa that’s only good for five years. And if you don’t contribute to Australian society, it’s not getting renewed. Judi Galst (20:38.082)But I’ve had people from Hollywood, I’ve had songwriters, I’ve had producers, directors, people in private equity that specialize in sectors that are important to Australia. People in finance have been approved. So it’s worth considering if the idea of being able to live in Australia means something to you. Interestingly with that visa, you can also live in New Zealand. Frazer Rice (20:58.095)Okay, it’s one of those things too. If people aren’t forcing you to say, don’t hate me because I’m beautiful, that might not be a good route, but if you are talented or bring something to bear, it may be worth taking a stab at. Is it reciprocal? If you’re in New Zealand, can you go to Australia? Got it. So let’s pivot to Asia and or South America, which you hear about Singapore, you hear about… Judi Galst (21:16.194)No. Good question. Frazer Rice (21:27.131)Other different sort of haveny types of places where people place their wealth or establish family offices and South America I think is, know, think about like Uruguay and places like that which, you know, have the reputation of being the Switzerland of South America. What’s the state of play there? Judi Galst (21:44.527)So I have actually had a few clients that have done residency in Uruguay. They don’t have a formalized program, although I think a more formalized program is going to come out of there. Henley and Partners actually has a government advisory line of business, so we design a lot of these programs and we’re very active in South America. There’s a lot of interest in South America to have citizenship and residence by investment programs, so I think you’re going to see a lot coming from that region in the near term. But Uruguay does have a path to residency. You have to spend time there. Frazer Rice (21:58.611)Mm-hmm. Frazer Rice (22:12.893)Judi Galst (22:13.251)And they don’t tell you exactly how much. Yeah. But most of my clients went with the expectation that maybe they’d have to stay for 30 days and they ended up getting the visa approved faster. You have to go back every year for a period of time or not renew renewing it. But yes, there is a path in Uruguay and more in Central America. People are doing Panama. Frazer Rice (22:36.637)Costa Rica. Judi Galst (22:37.773)Costa Rica is really interesting, very affordable. know we wanted to talk a little bit about the range, but in Costa Rica, you can gain temporary residence by demonstrating you have $2,500 a month in passive income. Many people will have that with interest and dividend income. Or you could invest $150,000 in real estate. It’s a temporary residence for two years, and then you renew for another two years. But at three years, you can transition to permanent residence. As a temporary resident, cannot work for a company in Costa Rica, so you’d have to be able to work remotely. And then once you become a permanent resident, that requirement disappears. Once you are approved, you do have to pay into Social Security in Costa Rica that gives you access to health care. So it’s about $300 per application per month. But Costa Rica is very interesting, I think. Frazer Rice (23:26.67)As we go back, pivot back to Asia, are there any countries with Singapore or others that are possibilities for people in the US? Judi Galst (23:33.722)So Singapore is a possibility. However, you have to move a family office with over 200 million there, or investment levels are around 30 million, and you have to relocate, and the ability to renew it is contingent upon how much time you spend in Singapore. So I would say a very niche client could do Singapore. A more affordable option might be Thailand, which you can get a residence permit very… Frazer Rice (23:44.125)Mm-hmm. Frazer Rice (23:52.605)To be sure. Okay. Judi Galst (24:00.782)Inexpensively. mean, a five-year permit for $25,000. Frazer Rice (24:05.159)Wow. And to round out our tour of the world here, Middle East countries, maybe the UAE, you hear about that as a place where a lot of Europeans go to move their wealth. Is that becoming popular with United States citizens? Judi Galst (24:16.463)Mm-hmm. Judi Galst (24:22.381)Golden Visa in Dubai is very popular. Honestly, not so much among Americans. It’s usually people from other parts of the world. mean, my firm has 70 offices around the world and we do a lot of UAE Golden Visas. I don’t have a huge amount of interest from Americans. I’ve done a couple of them. It’s not hard. You do have to spend time, like 30 days as part of the process there. Frazer Rice (24:26.525)Mm-hmm. Judi Galst (24:46.703)You can invest in real estate at 550,000, but there’s like 19 different visa types. You can set up a company. If you’re a member of YPO, Young Presidents Organization, they’re deemed talented and they don’t even make an investment. So, you know, it’s an option and we could certainly help it. But to be honest, I don’t see huge demand among Americans. Frazer Rice (25:03.259)Interesting. So let’s round this out a little bit here. For a U.S. citizen who is feeling unsettled or is just curious what’s out there. They want the ability to go live in Madeira, buy a place there. And to be able to go unfettered or something like that. What’s a good thought process or sequence of events for them to go through in order to make that happen? Judi Galst (25:31.344)I mean, we don’t charge for consultations. So I don’t know if you’re going to share my email at the end of this, but just hit me up. To me, any client conversation is about educating. This is generally a new topic for someone. It’s very rare that someone calls me and they really understand what is available to them and also what would be a good fit for them. They may not understand if they want to include their children. There are going to be some that are going to be better fits for them than other based on the ages of the kids. They may not understand how much time they have to spend in a country to make it happen. How much it’s going to cost, and just learn about it. Learn what your options are. I can usually pretty quickly. Once I understand a client’s objectives, tell them. This is a strategy that I think makes sense for you and exactly how it would Frazer Rice (26:14.206)And it strikes me too, that for people who are exploring different places, it’s probably a good idea to have visited them first before just jumping in, jumping in feet first and sort of solving a problem without understanding what actually implementing the solution looks like. Judi Galst (26:21.111)Yeah. Yeah. Judi Galst (26:29.177)For sure. I because many of the clients that I work with are of higher wealth, they usually have done a fair amount of traveling. So the idea of envisioning, know, residency in Italy, they’ve been to Italy. But when I talk to clients, especially about the Caribbean, where they might be investing in real estate and they have to decide between which country makes the most sense, I always tell them they should try and go because it can be a lifestyle decision. And they want to see where they could actually envision themselves if, in fact, they triggered this insurance policy. Frazer Rice (26:58.59)Judy, great stuff. Here it is. Put your email out there in case people want to reach out and find out more. Judi Galst (27:05.099)Okay, amazing. So my email is my first name, Judy, J-U-D-I dot my last name, GALST, G-A-L-S as in Sam T, at henleyglobal.com, H-E-N-L-E-Y, global.com, or you can give me a call at 646-856-3712. Frazer Rice (27:29.406)Great stuff. We’re going to have that in the show notes too so people can look on webpage, etc. to get that information. Thank you so much. It’s something, you know, when you’re at the desk and dreaming wistfully about what life looks like, what you’re done working, if you’re done working, my calculation is I’ll be able to retire when I’m 127. But it’s great just to sort of envision what that looks like. the expertise is out there. Thanks for being on. Judi Galst (27:56.047)My pleasure. HENLEY & PARTNERS DAVID LESPERANCE ON CITIZENSHIP DIVERSIFICATION DAVID LESPERANCE ON US EXPATRIATION https://www.amazon.com/Wealth-Actually-Intelligent-Decision-Making-1-ebook/dp/B07FPQJJQT/ #familyoffices #citizenship #residency #residencybyinvestment #citizenshipbyinvestment #austriancitizenship #newzealand #portugalproperty #portugalresidency #uscitizens #stkitts #malta #eucitizenship #wealthcitizenship #Californiawealthtax #puertorico #puertoricotax

with @PalmerLuckey @cdixonIn this special episode — our 100th on the a16z crypto show! — Chris Dixon interviews Palmer Luckey (founder of Anduril; founder of Oculus VR and designer of the Oculus Rift) to talk about the future of technology, belief, and building.What does it take to build hardware at scale? Where are many of today's tech bottlenecks? And what's the case for optimism about the future despite growing geopolitical turmoil, regulatory constraints, and other blockers to innovation? The candid, wide-ranging conversation covers crypto, banking, and stablecoins, as well as modern warfare, the U.S.–China technology race, AI and manufacturing, and much more. Dixon also digs into company building in good times and bad with Luckey; the conversation was recorded live at our Founders Summit. Highlights:0:00 — Introduction2:08 — Early Oculus: Why VR was hard8:02 — Bitcoin & early crypto days9:49 — The Facebook acquisition13:36 — How successful was VR, really?18:59 — Starting Anduril20:01 — Hiring for mission ("Don't Work at Anduril")23:59 — How Anduril works (product dev, org design)27:47 — How Palmer stays ahead of the curve33:00 — The US-China technology race34:40 — What Putin understood early about AI39:45 — Stablecoins & banking risk45:00 — Politics as bottleneck47:00 — Future of technology: AI, fusion, quantum50:23 — Automation, abundance, and optimism53:23 — Ukraine, drones, and the reality of warFollow a16z crypto for more...X: https://x.com/a16zcryptoLinkedIn: https://www.linkedin.com/showcase/a16zcrypto/posts/Spotify: https://open.spotify.com/show/7pMZvsNXEnb0CYcPiDQywEApple Podcasts: https://podcasts.apple.com/us/podcast/web3-with-a16z-crypto/id1622312549Youtube: https://www.youtube.com/@a16zcrypto

Following two years of above-average production, U.S. apples are expecting another pretty good year, with Washington doing it's share of the lifting.

In episode 1993, Jack and Miles are joined by English professor, author of The New Mutants: Superheroes and the Radical Imagination of American Comics, and host of Nerd from the Future, Ramzi Fawaz, to discuss… Trump Continues To Prove The Haters Right, Minneapolis PD Says Off Duty Officers (Of Color) Being Targeted by ICE, The Right Is Trying To Claim Star Trek and more! Trump: We've done more than any other administration has done by far—in terms of military, in terms of ending wars, in terms of completing wars, nobody's really seen very much like it. Trump: These are professional agitators and professional people that want to see our country do badly. But that's not happening because we have the hottest country. Trump: I'm glad my finger wasn't in that sucker. That could have done some damage. But you know what? I wouldn't have shown the pain. Trump: "Your lover isn't going to be killed anymore, so you can act like a real lover. You can walk right through the middle of the town. And DC is beautiful again too." Minneapolis PD Says Off Duty Officers (Of Color) Being Targeted by ICE William Shatner eats a bowl of cereal while driving and more star snaps William Shatner boldly devours cereal while driving his SUV in Studio City Stephen Miller Has a Truly Rancid Star Trek Opinion William Shatner Pokes Fun at Stephen Miller for Calling on Him to Control ‘Star Trek’ Franchise How Stephen Miller Rode White Rage from Duke’s Campus to Trump’s West Wing ‘Star Trek: Starfleet Academy’ Debuts With Positive Reviews And Political Nonsense Musk and Hegseth vow to “make Star Trek real” but miss the show’s lessons "Star Trek is inherently right wing and Christian and no amount of modern rewriting or changing of canon can remove that." Elon Musk and Stephen Miller’s culture war against Star Trek is built on ignorance Hollywood Flashback: ‘Star Trek’ Showed TV’s First Interracial Kiss in 1968 How ‘Star Trek’ Survived the Vietnam Era and Took Over the World Star Trek's Prime Directive Had A Grim Real-Life Inspiration William Shatner responds after Ted Cruz says Captain Kirk was likely a Republican LISTEN: PARTO NATURALE by MarteSee omnystudio.com/listener for privacy information.

In episode 1992, Miles and guest co-host Blake Wexler are joined by comedian and producer of the monthly Facial Recognition Comedy show, Pallavi Gunalan, to discuss… President Pump Fake? One of Our Sh*ttiest Senators Is Being Sued Under A HOMEWRECKER LAW? Analyst Warns: The Bank Of England Should Prep For Aliens, Brooklyn Beckham Calls Out His Famous Parents and more! “Let Me Speak Your Language, Trump—F* Off”: EU Lawmaker Explodes in Parliament Over Greenland | AC1G BESSENT: I'd tell everyone sit back. Take a deep breath. Do not retaliate. Do not retaliate. Kilmeade: Greenland Will Cost GOP The Midterms Kyrsten Sinema Faces ‘Homewrecker’ Lawsuit for Alleged Affair With Former Bodyguard Bank of England must plan for a financial crisis triggered by aliens, says former policy expert The Disclosure of Aliens Could Cause a Bitcoin Rush, Former Bank of England Analyst Says Brooklyn Beckham Calls Out His Famous Parents Brooklyn Beckham: ‘I do not want to reconcile with my family’ David Beckham breaks silence after son Brooklyn Beckham post LISTEN: Chill Me Out by Masayoshi TakanakaSee omnystudio.com/listener for privacy information.

Concern about the possibility of a Chinese attack against Taiwan has surged in recent years. Wargames and research studies have focused primarily on identifying gaps in US and allied capabilities with the goal of strengthening deterrence. A relatively understudied question, however, is the potential consequences for China if a military operation against Taiwan were to fail. To address this gap, the German Marshall Fund led a study of the possible costs that China would incur across four different, but interrelated areas: the Chinese economy, the military, Chinese social stability, and international costs.GMF commissioned four papers on these key areas. We considered two scenarios that could realistically take place in the next five years. In the first scenario, a minor skirmish escalated into a multi-week maritime blockade of Taiwan by China. Although several dozen members of the Chinese and Taiwanese military were killed, US intervention eventually forced China to de-escalate.  In the second scenario, a conflict escalated into a full-fledged invasion, with Chinese strikes on not only Taiwan but also U.S. forces in Japan and Guam. After several months of heavy fighting, Chinese forces were degraded and eventually withdrew after suffering many tens of thousands of casualties.The authors found that the costs to China of a failed military action against Taiwan would likely be considerable. We believe their findings are important and warrant wide dissemination. In this podcast, we'll discuss the report's major conclusions and implications. Then we'll talk about the potential impact of a failed Chinese attempt to take Taiwan on China's military capabilities and the possible international costs that Beijing could face.  Our next two China Global podcasts will examine the implications of a failed military operation against Taiwan for China's economy and social stability.Our guests today are Zack Cooper and Joel Wuthnow. Zack is a senior fellow at the American Enterprise Institute and lecturer at Princeton University.  Joel is a senior research fellow in the Center for the Study of Chinese Military Affairs within the Institute for National Strategic Studies at NDU. Joel's paper and this interview reflect only his personal views and not those of the National Defense University, the Department of War, or the US government.Timestamps: [00:00] Introduction  [03:22] Implications for China, the United States, and Taiwan [06:31] Actions to Strengthen Deterrence  [08:50] Evaluating Costs and Risks for Chinese Decisionmakers[11:46] Lessons Learned for the PLA [14:05] Steps to Avoid Another Attack [17:14] Intensifying Frictions between Party and Military?  [19:53] Anticipating US Intervention as a Military Variable [22:49] Countries and Organizations Likely to Respond to China[25:55] Potential Diplomatic Actions and Costs[31:50] A Treaty Alliance with Taiwan [34:44] Why International Costs Matter to China

European leaders have condemned threats levelled by the United States to impose new tariffs on countries who don't cooperate with U-S ambitions to take over Greenland. U-S President Donald Trump has continued with his plans to acquire control of the Arctic island, prompting warnings about the future of NATO and transatlantic ties.

Canada has been lockstep with the U-S when it comes to trade policy with China for years.Last week, that changed. PM Mark Carney's new tariff deal with China will allow 49,000 Chinese EVs into the country every year for three years at a dramatically-slashed tariff of only 6.1 per cent—in exchange for a reduced tariff on canola seeds, lobster, crab and other agricultural products exported to China.It's a deal that's earning praise from the prairies but disdain in Ontario.Host Cristina Howorun sits down with Flavio Volpe, President of the Automotive Parts Manufacturing Association and one of the architects of CUSMA, to discuss the implications this deal could have on the EV market, the 90,000 jobs in the auto sector and tariff and trade negotiations with the States. We love feedback at The Big Story, as well as suggestions for future episodes. You can find us:Through email at hello@thebigstorypodcast.ca Or @thebigstory.bsky.social on Bluesky

European leaders have condemned threats levelled by the United States to impose new tariffs on countries who don't cooperate with U-S ambitions to take over Greenland. U-S President Donald Trump has continued with his plans to acquire control of the Arctic island, prompting warnings about the future of NATO and transatlantic ties. We talked to Dr. Ronja Kempin from the Stiftung für Wissenschaft und Politik in Berlin about the current geopolitical situation. - Die aktuelle geopolitische Lage ist derzeit durch den Konflikt um Donald Trumps Ansprüche auf Grönland geprägt. Dadurch wird auch das Verhältnis der eigentlich engen Verbündeten der EU und der USA auf eine harte Probe gestellt – Dies nicht zuletzt durch die jüngst angekündigten Strafzölle, die zuerst Trump und dann die EU angekündigt haben. Darüber sprechen wir mit Dr. Ronja Kempin von der Stiftung Wissenschaft und Politik in Berlin.

The U-S evacuate personnel from a mideast base as Trump mulls over military strikes on Iran. A Jewish group urges changes and a delay to the government's hate speech reform. - トランプ大統領がイランへの軍事攻撃を検討する中、アメリカは中東の基地から職員を退避させています。連邦政府が起草した　ヘイトスピーチ規制法をめぐり、国内のユダヤ人団体は法案の練り直しを求めています。また、オーストラリア・イマーム評議会からも批判の声が出ています。

The U-S evacuate personnel from a mideast base as Trump mulls over military strikes on Iran. A Jewish group urges changes and a delay to the government's hate speech reform. Recorded 16 January. - トランプ大統領がイランへの軍事攻撃を検討する中、アメリカは中東の基地から職員を退避させています。連邦政府が起草した　ヘイトスピーチ規制法をめぐり、国内のユダヤ人団体は法案の練り直しを求めています。また、オーストラリア・イマーム評議会からも批判の声が出ています。 2026年1月16日収録。

In episode 1989, Jack and guest co-host Mort Burke are joined by comedian, Blake Wexler, to discuss… At Least Zohran Is Getting Busy, The Trump Administration Wants Us To Believe That They Havana Syndrome-d Venezuela, Finally An App To Ensure You’re Not Dead, Now Stranger Things Fans Are Convinced That ChatGPT Wrote The Finale and more! US used powerful mystery weapon that brought Venezuelan soldiers to their knees during Maduro raid: witness account This Pain-Inducing Acoustic Device Used to Control Crowds in Azerbaijan Might Be U.S.-Made How to Dodge the Sonic Weapon Used by Police Are You Dead?: The viral Chinese app for young people living alone An App Called ‘Are You Dead?’ Is Climbing the Apple Charts A record share of Americans is living alone Why humans are increasingly choosing to live alone Rising numbers of people found long after death in England and Wales – study The Backlash Against Netflix’s ‘Stranger Things 5’ Documentary, Explained Stranger Things Fan Tweet: "is that a f**kin chatgpt tab i see" LISTEN: Victory Lap by Fred again.., Skepta, PlaqueboymaxSee omnystudio.com/listener for privacy information.

On this Salcedo Storm Podcast: Retired, Lt. Colonel Tony Shaffer is a Newsmax Contributor. And he's the President of Project Sentinel.

On this Salcedo Storm Podcast:Chris talks about the great skill and power of the U.S. military. He also dives into to the Democrat domestic terrorists targeting ICE.

Protests over Iran spread to cities around the world. Meanwhile, President Trump has been warning the U-S would get involved if protesters are shot. Greg and Holly discuss the latest and whether military action from the US could backfire.

The year 2026 begins with the end of pretense.In this kickoff episode of Boiler Room, host Bryan “Hesher” McClain is joined by Adam “Ruckus” Clark, Bazed-Lit Analyzer, Mystical Pharaoh, and Mark Anderson to examine what can no longer be denied: the transition from covert influence to open empire.As Venezuela becomes the focal point of overt regime-change language—justified openly by energy control and corporate interests—the illusion of “rules-based order” collapses in real time. From U.S. officials admitting the real motives at the UN, to media outlets reframing intervention as necessity, the show traces how empire now speaks plainly.But this isn't just a foreign policy story.The same logic driving intervention abroad is returning home: militarized enforcement, narrative warfare, AI-driven propaganda, and government behavior increasingly treated as spectacle—甚至 wagered on in real-time prediction markets.From the symbolic death of MTV to deepfake politics, from Greenland to Cuba, from ICE violence to betting on press briefings, this episode sets the tone for the year ahead.No illusions.No euphemisms.No disguises.Welcome to open empire.Support:Support BOILER ROOM & ACRPatreon (Join and become a member)Shop BOILER ROOM Merch Store

An Iranian-Canadian tells us she's ready to accept all the risks associated with a U-S military intervention, if it means ousting the regime threatening her loved ones' lives.We reach a Minneapolis council member -- who tells us why she's urging her constituents to keep the pressure on the federal government as ICE agents remain in the city.Yesterday on this show, the chief of Pimicikamak Cree Nation had some tough questions for Manitoba Hydro about the outage that has displaced his community. Today, Manitoba Hydro responds.Earlier this week, Donald Trump and Gustavo Petro seemed to be mending things, but we'll play you part of a recent BBC interview with the Colmbian president in which he's pulling no punches when it comes to his view of the United States.The costume designer for "Heated Rivalry" says she had no inkling show would take off like it has -- let alone spark an obsession with one piece of clothing in particular.A canine Houdini cracks two locks to make his get-away from a shelter -- and back into the arms of his owner. As It Happens, the Friday Edition. Radio that admires his escape claws.

In this weekend's episode, three segments from this past week's Washington Journal. First, the aftermath of the U-S military operation in Venezuela that toppled President Nicolas Maduro dominated Washington this week. We speak with Republican Randy Fine of Florida – a member of the Foreign Affairs Committee – about the future of U-S involvement in that country. The U-S military strikes and the capture of Maduro has raised all kinds of legal and constitutional questions. We break it all down with Creighton University Professor and international law expert Michael Kelly Finally, this week also saw the 5th anniversary of the January 6th attack on the Capitol. We chat with Associated Press Congressional Reporter Mary Clare Jalonick about the legacy of that day which she examines in her book "Storm at the Capitol: An Oral History of January 6th." Learn more about your ad choices. Visit megaphone.fm/adchoices

In episode 1984, Jack and Miles are joined by comedian and host of Rebrand, Mort Burke, to discuss… Trump: People Say I’m Jealous But My Kink Is Just Karma, Benny Johnson: Venezuela Rigged The 2020 Election! So Yeah! Trump Health? John Krasinski Laid The Groundwork For Venezuela Attack and more! U.S. plan to ‘run’ Venezuela clouded in confusion Benny Johnson: Venezuela Rigged The 2020 Election! So Yeah! Trump Health? John Krasinski Laid The Groundwork For Venezuela Attack Jack Ryan clip about Venezuela gets viral amid capture of Nicolàs Maduro. Did ‘Jack Ryan’ Predict U.S.’ Venezuela Intervention? Co-Creator Carlton Cuse Reacts To Season 2 Clip Going Viral, Shares Hopes For “Stability And Peace” Amazon's 'Jack Ryan' TV series lambasted for promoting Venezuela 'invasion' Jack Ryan is the Latest TV Show to Film at CIA Headquarters How Does Amazon's 'Jack Ryan' Compare to Real Life at the CIA? LISTEN: 4 Raws by EsDeeKidSee omnystudio.com/listener for privacy information.

Pick the headline that best describes the story:VenezuelaTrump's Hint to Oil Executives Weeks Before Maduro Ouster: ‘Get Ready'Maduro overthrow could help these U.S. oil companies recover assets seized by VenezuelaTrump makes it clear shocking Venezuelan regime change is largely about oil: ‘They stole our oil … We're going to make a lot of money'US oil giants mum after Trump says they'll spend billions in VenezuelaUS oil companies gain after capture of Venezuela's MaduroA group of about 20 US investors is already planning a trip to Venezuela in MarchMaduro falls, Bitcoin rises: The 1,671% surge that hit before Wall Street woke upAI-generated content spreads after Maduro's removal — blurring fact and fictionElonElon Musk's X faces probes in Europe, India, Malaysia after Grok generated explicit images of women and childrenElon Musk's X faces regulatory probes in Europe, India and Malaysia after its Grok chatbot began generating deepfake explicit images, some depicting child sex abuse.Elon Musk After His Grok AI Did Disgusting Things to Literal Children: “Way Funnier” Elon Musk's Grok AI faces government backlash after it was used to create sexualized images of women and minorsMusk's xAI faces backlash after Grok generates sexualized images of children on XWoman felt 'dehumanised' after Musk's Grok AI used to digitally remove her clothesElon Musk plans 'high-volume production' of Neuralink brain chips and says he wants to automate the surgical procedureTesla Loses EV Crown to BYD After Second Annual Sales DropAIChildrenTech Giants Pushing AI Into Schools Is a Huge, Ethically Bankrupt Experiment on Innocent Children That Will Likely End in DisasterChildren Falling Apart as They Become Addicted to AIOpenAI's child exploitation reports increased sharply this yearPsychosisDoctors Say AI Use Is Almost Certainly Linked to Developing PsychosisWoman Suffers AI Psychosis After Obsessively Generating AI Images of HerselfMan Describes How ChatGPT Led Him Straight Into PsychosisAI Godfather Warns That It's Starting to Show Signs of Self-PreservationDisturbing Messages Show ChatGPT Encouraging a Murder, Lawsuit AllegesOpenAI Reportedly Planning to Make ChatGPT “Prioritize” Advertisers in ConversationBillionairesThe world's richest people just had their best year yetAI boom adds more than half a trillion dollars to wealth of US tech barons in 2025There are more self-made billionaires under 30 than ever before—11 of them have made the ultra-wealthy club in the last 3 months thanks to AIJamie Dimon made $770 million last year. 2026 could be even better for banksEasing rules and a rebound in dealmaking are reshaping the landscape for U.S. banks, with bigger profits likely aheadThreat of California Billionaire Tax Draws Criticism From UltrawealthyBill Ackman slams California wealth tax as ‘expropriation' of private propertyBill Ackman Blasts Ro Khanna For Defending Billionaire Tax: 'Lost His Way'Peter Thiel and Larry Page are preparing to flee California in case the state passes a billionaire wealth tax, report saysTech billionaires threaten to flee California over proposed 5% wealth taxBari Weiss yanking a 60 Minutes story is censorship by oligarchy Speed Round Dumb or Good Rating (1-10)Dumb2 former Hinge execs are building an app to make it easier to plan hangouts with your friends 10Boeing (BA) CEO “is a Nonsense Guy,” Says Jim Cramer 2Some men may downplay climate change risks to avoid appearing feminine 0New research provides evidence that men who are concerned about maintaining a traditional masculine image may be less likely to express concern about climate change. The findings suggest that acknowledging environmental problems is psychologically linked to traits such as warmth and compassion. These traits are stereotypically associated with femininity in many cultures. Consequently, men who feel pressure to prove their manhood may avoid environmentalist attitudes to protect their gender identity. The study was published in the Journal of Environmental Psychology.CEO of local public company to step down after nearly 10 years 9Malcolm Gladwell tells young people if they want a STEM degree, ‘don't go to Harvard.' You may end up at the bottom of your class and drop out 6OpenAI CEO Sam Altman says he is ‘envious' of Gen Z college dropouts who have the ‘mental space' and time to build new startups 9This 22-year-old college dropout with an AI powered YouTube empire makes $700,000 a year and works just 2 hours a day 1Trump Mobile says its first-ever smartphone is delayed, and the government shutdown is to blame 3The college-to-office path is dead: CEO of the world's biggest recruiter says Gen Z grads need to consider trade jobs with no degree required 4ChatGPT gets ‘anxiety' from violent user inputs, so researchers are teaching the chatbot mindfulness techniques to ‘soothe' it 7Good?Minimum wage just went up in 19 states—workers in one state are getting a $2 an hour raise 8Judge says Trump administration must continue funding consumer watchdog Consumer Financial Protection Bureau 5Angry town halls nationwide find a new villain: the data center driving up your electricity bill while fueling job-killing AI 8Bernie Sanders and Ron DeSantis speak out against data center boom. It's a bad sign for AI industry 9Mitt Romney says the U.S. is on a cliff—and taxing the rich is now necessary ‘given the magnitude of our national debt' 7Microsoft CEO Begs Users to Stop Calling It “Slop” 10Man Operating Robot Accidentally Makes It Kick Him Directly in the Nutsack 9MATT1WE MISSED THE PREDICTIONS:Crypto: Tom Lee Predicts $250K Ethereum Price as BitMine Adds to $13 Billion Stash, Grayscale Predicts Bitcoin Will Reach New All-Time High by March 2026Stocks: Every Wall Street Analyst Now Predicts a Stock Rally in 2026 - EVERY!AI: ‘Godfather of AI' Geoffrey Hinton predicts 2026 will see the technology get even better and gain the ability to ‘replace many other jobs', Amazon's Alexa chief predicts an end to doom scrolling: the next generation is ‘going to just think differently', In 2026 CFOs predict AI transformation, not just efficiency gainsAs millions of Gen Zers face unemployment, CEOs of Amazon, Walmart, and McDonald's say opportunity is still there—if you have the right mindset - I PREDICT NOT HAVING A JOB IS YOUR OWN FAULTOily oil: Oil experts predict slight rise in gas prices as global tensions mountBlowhards: Elon Musk predicts double-digit US growth by 2026, Treasury Secretary predicts historic merger could make 2026 a ‘very good year' Trump advisor predicts Miami will dethrone NYC as financial capital under new progressive mayorOpenAI's CEO Sam Altman says in 10 years' time college graduates will be working ‘some completely new, exciting, super well-paid' job in space

SummaryIn this episode, Clayton Cuteri delves into significant current events, focusing on the U.S. military operation in Venezuela and the Minnesota fraud case. He explores the underlying economic motivations, the role of Israel, and the distractions posed by ongoing allegations against political figures. The conversation also touches on the importance of empowerment and spiritual wealth, encouraging listeners to seek personal growth and understanding in a complex world.Clayton's Social Media LinkTree | TikTok | Instagram | X (Twitter) | YouTube | RumbleTimecodes 00:00 - Intro00:30 - Major Stories of the Week: Venezuela and Minnesota Fraud01:23 - U.S. Military Operation in Venezuela: The Facts03:11 - Understanding the Layers of Venezuela's Situation06:27 - The Economic Motivations Behind Venezuela's Takeover15:22 - The Role of Israel in Venezuela's Political Landscape18:08 - Distractions and Allegations: The Epstein Files21:05 - America's Military Role in Global Politics24:48 - Minnesota Fraud: A Local Perspective33:40 - Empowerment and Spiritual WealthIntro/Outro Music Producer: Don Kin IG: https://www.instagram.com/donkinmusic/Spotify: https://open.spotify.com/artist/44QKqKsd81oJEBKffwdFfPSuper grateful for this guy ^NEWSLETTER - SIGN UP HEREBecome a supporter of this podcast: https://www.spreaker.com/podcast/traveling-to-consciousness-with-clayton-cuteri--6765271/support.Official Traveling to Consciousness Website HEREALL Indigo Education Podcasts HEREMy Book: The Secret Teachings of Jesus HERE

On CNN's State of the Union, Dana Bash presses Senate Intelligence Committee Chairman Tom Cotton about President Trump saying the U-S is now “running” Venezuela. Next, Democratic Sen. Chris Murphy tells Dana that the Trump administration “lied to our face” about pursuing regime change in Venezuela. Then, House Judiciary Committee Chairman Jim Jordan tells Dana that he “trust[s] the president to make decisions that are in the best interest of Americans” in Venezuela. After, House Intelligence Committee Ranking Member Jim Himes tells Dana that Jordan “gave the game away” and that “America can see the fact that they no longer have a Congress.” Finally, Dana talks with former NATO Supreme Commander Adm. James Stavridis and former Deputy DNI Beth Sanner about what comes next after Maduro's ouster in Venezuela.   Learn more about your ad choices. Visit podcastchoices.com/adchoices