Podcasts about Optimus

Trump turns 80. Musk goes trillionaire. Maine's Dem frontrunner had a Nazi tattoo. In this episode: • Trump cancels Iran air strike — no one told Iran • Jay Clayton named DNI over actually qualified Bill Py • Todd Blanch nominated as permanent AG: commits one felony a day for Trump • $70 billion for ICE — body cameras not included • SpaceX IPO: the largest in stock market history — Musk becomes the world's first trillionaire • 1 million AI data centers in orbit. 1 million Optimus humanoid robots. The endgame. • Trump visited 22 medical specialists — zero psychiatrists • Trump turns 80 and gains weight: BMI 29.9, just under "obese" • Trumpflation is official • Graham Platner leads Susan Collins by 9 points in Maine • Platner's skull-and-bones tattoo turned out to be a Nazi SS symbol • Platner's digital trail: racism, misogyny, and trivialization of rape • NYT hints Platner may have been physically abusive toward women Key figures covered: Donald Trump, Elon Musk, Todd Blanch, Jay Clayton, Graham Platner, Susan Collins, Pete Hegseth, Bill Gates, Steven Miller

Why do coaching clients so often hit a wall with "I don't know", and how can you as a coach deal with it skillfully - without panicking or falling into quick-fix patterns? In this episode of The Coaching Hub Podcast, I explore what's really happening when clients get stuck and offer practical strategies to turn these moments into breakthroughs.I break down the variety of reasons clients get stuck: nervous system activation, not feeling safe, not ready to share, a simple need for space, or even neurodiversity factors like alexithymia, which make describing emotions difficult.You'll learn why piling on more questions rarely helps, and how offering space, naming patterns, and getting curious (rather than attached to an outcome) can help clients access their own insights. If you want to empower your clients (and yourself) to move forward from uncertainty, this episode will help you reframe those moments and support change more effectively.What's inside:Why "I don't know" from clients shouldn't be seen as a red flag but as a starting point for deeper explorationGiving clients space and allowing silence can lead to meaningful self-reflection and breakthroughs “I don't know" may signal discomfort rather than a lack of desire to engageCoaches should remain curious, notice patterns, and view "I don't know" as valuable data, not a problem to fixSelf-awareness is crucial; coaches' reactions to uncertainty can impact the coaching relationship and client safetyEmbracing and exploring "I don't know" can foster growth, new insights, and a better understanding of client needs.Key Topics Covered:00:00 Intro and what to do when your clients say "I don't know"00:29 Common tendency to over-question when clients seem stuck01:02 Interpreting "I don't know" as a sign of deeper client needs or emotions01:52 The power of allowing space and silence after "I don't know"03:00 Viewing "I don't know" as valuable data or insight03:40 The importance of curiosity over repetition in questioning04:02 Working with clients who struggle to identify or articulate emotions (alexithymia)05:28 Recognising overwhelm, freeze states, and nervous system activation in clients06:36 Revising the belief that coaches must always have answers, and discussing knowledge gaps07:39 The coach's self-awareness and regulation when confronted with client uncertaintyAbout Ruth Kudzi:Ruth is the founder of Optimus Coaching Academy and a well-respected and successful coach, speaker and author who has worked across sectors including leadership, career and more recently business and mindset. She has over 10,000 coaching hours and has completed hundreds of hours of training and coaching supervision. Ruth is an MCC-level coach with the ICF and is our course director and CEO. Prior to becoming a coach, Ruth was a senior leader in education. Find out more here: https://ruthkudzi.com/ Book:How to Feel Better: 4 Steps to Self-Coach Your Way to a Happier More Authentic You eBook : Kudzi, Ruth: Amazon.co.uk: Books Connect with Ruth: YouTube: https://www.youtube.com/@RuthKudzi Facebook: https://www.facebook.com/Ruthkudzi2/ Instagram: https://www.instagram.com/ruthkudzi/ LinkedIn: https://www.linkedin.com/in/ruthkudzi/ Podcast: https://podfollow.com/the-coaching-hub-podcast About Optimus Coach Academy:Optimus offers best in class coaching training for individuals and corporates. If you want to know more about what we offer: https://optimuscoachacademy.com/coach-training We also offer business support as standard at Optimus, find out more here: https://www.optimuscoachacademy.com/ Connect with Optimus Coach Academy: Facebook: https://www.facebook.com/Optimuscoachacademy Instagram: https://www.instagram.com/optimuscoachacademy/ Linkedin: https://www.linkedin.com/company/optimuscoachacademy Produced by Winter Audio.

Is the golden age of personal development over, or is the industry simply going through a necessary shift? In this episode of The Coaching Hub, I take a closer look at how personal development has evolved and why we're starting to question the influence of gurus who were once placed on a pedestal.I explore what led to the rise of these figures, the impact of cancel culture, and how unrealistic expectations of perfection have shaped both the industry and the people within it. I also explore our tendency to outsource our thinking, and what that means for genuine, sustainable growth.More importantly, I challenge the idea that personal development is something we consume, and instead bring the focus back to personal responsibility, self-awareness, and doing the deeper work ourselves. If you've been questioning where personal development is heading, this episode offers a grounded perspective on what's changing and what actually matters moving forward.What's inside:Why the era of putting personal development gurus on pedestals is coming to an endThe impact of cancel culture and why some “gurus” have fallen from graceHow our expectations of perfection in others can hold us back from real self-growthWays that personal development is becoming more individual and meaningfulThe importance of self-reflection and personal agency over following “experts”How embracing imperfection, both in ourselves and others, is essential to authentic developmentWhy the future of personal growth means focusing less on gurus and more on our own context, relationships, and continuous learning.Key Topics Covered:00:00 Intro01:09 The rise and fall of “gurus”02:16 Cancel culture and loss of trust03:28 Why we put people on pedestals (and why it's a problem)05:25 Accepting human imperfection in personal development06:46 Moving from guru-led to self-led growth08:34 What real personal development actually looks like10:43 Refocusing on personal responsibility and agency11:28 Final thoughts & reflection.Like, comment, follow and subscribe for more inspiring conversations on growth and professional development.About Ruth Kudzi:Ruth is the founder of Optimus Coaching Academy and a well-respected and successful coach, speaker and author who has worked across sectors including leadership, career and more recently business and mindset. She has over 10,000 coaching hours and has completed hundreds of hours of training and coaching supervision. Ruth is an MCC-level coach with the ICF and is our course director and CEO. Prior to becoming a coach, Ruth was a senior leader in education. Find out more here: https://ruthkudzi.com/ Book:How to Feel Better: 4 Steps to Self-Coach Your Way to a Happier More Authentic You eBook : Kudzi, Ruth: Amazon.co.uk: Books Connect with Ruth: YouTube: https://www.youtube.com/@RuthKudzi Facebook: https://www.facebook.com/Ruthkudzi2/ Instagram: https://www.instagram.com/ruthkudzi/ LinkedIn: https://www.linkedin.com/in/ruthkudzi/ Podcast: https://podfollow.com/the-coaching-hub-podcast About Optimus Coach Academy:Optimus offers best in class coaching training for individuals and corporates. If you want to know more about what we offer: https://optimuscoachacademy.com/coach-training We also offer business support as standard at Optimus, find out more here: https://www.optimuscoachacademy.com/ Connect with Optimus Coach Academy: Facebook: https://www.facebook.com/Optimuscoachacademy Instagram: https://www.instagram.com/optimuscoachacademy/ Linkedin: https://www.linkedin.com/company/optimuscoachacademy Produced by Winter Audio.

O que faz uma estratégia atravessar diferentes cenários de mercado ao longo de cinco anos?No novo episódio do Mind Asset, recebemos Pablo Salgado e Rodrigo Koch, gestores da família Itaú Optimus. A conversa revisita a trajetória de uma estratégia que nasceu em plena pandemia e atravessou alta global de juros, volatilidade e mudanças importantes no cenário internacional, sempre buscando combinar diferentes fontes de retorno dentro de uma gestão ativa e dinâmica.Os gestores também falam sobre os momentos mais desafiadores dos últimos cinco anos, o papel dos fundos multimercado na alocação das carteiras e como estão posicionando os fundos diante de um cenário ainda marcado por volatilidade e juros elevados.Antes de investir, verifique seu perfil de investidor.

We're announcing AIEWF speakers this week! Take the AI Engineering Survey!Today's guest Ethan first joined us for the LS Paper Club as the lead on NVIDIA Cosmos World Model, but then joined xAI and built Grok Imagine in 3 months:He comes back on Latent Space with some nuclear hot takes: that Video Models primarily get their intelligence from LLMs, not from training on video data, and that the next frontier for truly interactive, realtime, long-horizon world models is to work on LLMs (perhaps Interaction Models as well…)Put it this way: In the near term, the next Sora won't be a better video model, but a video agent.Generative Media may more closely follow the evolution of AI coding which went from focusing on one-shot output performance and cost, to multiturn reasoning and planning models for agents and systems that can plan, edit, test, debug, and submit PRs.At a certain point, coding models got so good that the only significant next step to improve performance was handling the orchestration of these models.Now as the performance of video models increases significantly across realism, consistency, & prompt adherence while becoming more cost efficient, the next evolution of video generation may also be systems that can plan, generate, edit, critique, and iterate across an entire creative task. In this episode, Ethan joins swyx and Vibhu to unpack what it actually takes to build frontier image and video systems: data, VAEs, diffusion transformers, audio-video alignment, inference speedups, and the hidden cost of storing and moving massive video datasets. From building NVIDIA's Cosmos world model to joining xAI as Grok Imagine was being built from zero to one, Ethan He has been at the center of some of the most important work in video generation, multimodal models, and real-time world models.We go deep on Grok Imagine, how a small xAI team shipped its first multimodal video model in three months, why iteration speed matters more than almost anything in model development, and why many of the biggest gains come from fixing tiny bugs in data and training pipelines. Flipbook: The future of VideomaxxingVideo agents are almost a sure bet to be the trend in the coming year. We end with a glance at what's beyond video agents:Flipbook caused a minor sensation this year when it was released, but most treat it as a fun demo. Ethan takes it very seriously — with the speed and cost of inference coming down every year, the future of custom video JIT UI is closer than you think. We talked about why videogen models may become the front end of AI, how generative UI could replace traditional HTML/CSS, why world models need to be real-time, interactive, and long-horizon, and why the future of video generation may depend more on language models and agents than on diffusion alone.We discuss:* Why fast iteration mattered more than meetings* Why small training bugs can drive huge model quality gains* Why coding models may make compute the bottleneck again* How image and video models are trained with synthetic captions* The role of VAEs and latent space in frontier video models* Why image models are the foundation for video models* The tradeoff between temporal compression and real-time interactivity* Flipbook, Neural OS, and the future of generative UI* Why future interfaces may go from user intent to pixels* The hidden cost of training video models: storage, egress, and GPU hours* How step distillation and consistency models (like OpenAI sCM) makes video inference orders of magnitude faster* Grok Imagine 0.9 and large-scale audio-video generation* Why audio-video alignment is harder than text-video alignment* Ethan's definition of world models* Reference-to-video, video extension, and long-context video generation* Why xAI's research communication undersells Grok Imagine* How xAI culture shaped the speed of development* AI watermarking, SynthID, and detecting generated media* Why prompt rewriting matters for video models* Grok Imagine Agent and the rise of video agents* Why language models may unlock better video generation* Robotics, physical AI, and embodied world models* Why Ethan left xAI and shifted focus toward LLMs* Self-managed context, memory, and the next frontier for language modelsEthan He* LinkedIn: https://www.linkedin.com/in/ethanhe42* X: https://x.com/EthanHe_42Timestamps00:00:00 Introduction00:01:25 From NVIDIA Cosmos to xAI00:03:24 Building Grok Imagine from Zero to One00:10:07 How Image and Video Models Are Trained00:18:53 Video Compression, VAEs, and Real-Time Tradeoffs00:22:10 Generative UI, Flipbook, and Neural OS00:32:10 The Cost of Training Large Video Models00:37:04 Distillation, GANs, and Fast Video Inference00:41:21 Audio-Video Generation and Grok Imagine 0.900:48:34 What Makes a World Model?00:55:51 Reference Videos, Long Context, and Video Memory01:00:11 xAI Culture, Research, and First-Principles Building01:09:45 AI Safety, Watermarking, and Prompt Rewriting01:13:10 Video Agents and AI-Assisted Creation01:27:32 Why Language Models Unlock Better Video01:31:15 Robotics, Physical AI, and Embodied World Models01:32:38 Why Ethan Left xAI01:34:16 Self-Managed Context and the Future of LLMs01:38:43 Ethan's Career Path and Closing ThoughtsTranscriptIntroduction: Ethan He, Latent Space, and the Path to xAISwyx [00:00:00]: We're here in the studio with Ethan He, most recently of xAI. Welcome.Ethan [00:00:10]: Thank you. Glad being here.Swyx [00:00:11]: We're also here with Vibhu. you were first coming to us or joining the latent space world because you were working on Kosmos at NVIDIA, and you did a paper. We loved it. you presented it as well, so thank you for doing that.Ethan [00:00:23]: I've actually, I also presented the MoEs twice at latent space.Swyx [00:00:29]: How did you actually hear about us? Did we reach out to you? Is that how it worked?Ethan [00:00:33]: No, actually, I-- the community. Like I realized, oh, there is this online community that people talk about AI and also learn from each other through papers every week through the Paperclip. It's very nice.Ethan [00:00:49]: I learned a lot.Swyx [00:00:49]: I think three years stop. We haven't stopped even on Christmas and New Years. many weeks I want to stop but it keeps going.Vibhu [00:00:58]: No, that was good. I think you had posted that you worked on a paper, and I was “Oh, very cool. We have Paperclip. Present then.”Vibhu [00:01:04]: But I might have reached out to you after.Swyx [00:01:05]: you-- because it's an amateur club, right?Swyx [00:01:08]: so it's very unusual and but we have sometimes paper authors come by and actually explain the paper. Today we just did, the poolside paper, which was apparently very good.Vibhu [00:01:18]: Came out yesterday.Vibhu [00:01:19]: pretty interesting, right? Fully open. They talk about everything, systems. So it's a good one. We'll, we'll recommend people to read it.Swyx [00:01:25]: Bring us up to speed on your transition to xAI, ‘cause I actually don't even know when you joined. just like tell the, tell the story about the sort of transition.From NVIDIA Cosmos to xAI: Scaling Video and World ModelsEthan [00:01:34]: Before xAI, I was working on Kosmos world model as in-- at NVIDIA. So Kosmos is, it's a giant video foundation models that can-- that aims to simulate the world and for-- it serves as a foundation of-- for all of the roboticists to build on top of. There, once I built the Kosmos one, I realized as this thing also has a scaling law similar to language model, we need to scale up the video models further. that's, that's why I realized I need to move to somewhere with much more compute resources. That's how ISwyx [00:02:13]: Than NVIDIA?Vibhu [00:02:14]: The GPU rich came themselves.Vibhu [00:02:19]: And timeline-wise, when was Kosmo? It was pretty early, right? It was open world model, open paper, everything.Ethan [00:02:25]: It was end of twenty-four.Vibhu [00:02:28]: End of twenty-four.Ethan [00:02:30]: Then at mid twenty-five, I moved to xAI. At that time-- I joined about the time when xAI was about to build video models and in multi-model models. There were no infra, no data, and no model, and it just-- as a few engineers, we built it in three months and released the first model, Grok Imagine zero point nine.Ethan [00:02:55]: And since then, I keep working on video models and move more from training and to post-training of the video models. For example, like a reference to videos, kind of like the cameo feature and, video extensions. And, before I left, I worked on a world model, leading a small team to focus on the real-time long horizon video generation.Building Grok Imagine From Scratch in Three MonthsSwyx [00:03:24]: Can you give like a rough roadmap of okay, you're on a brand-new team. Grok previously was only text, or they partnered with BFL for their image gen stuff. What do you-- what are the building blocks, right? You have compute, data you can procure somewhere. Like just what are like the sequence of things that people should think about when you're setting up a new team?Vibhu [00:03:43]: actually even deeper, not just data you can procure. You guys had to go through getting the data too, right? So you shipped it pretty fast, but yeahSwyx [00:03:51]: three months is likeVibhu [00:03:52]: From everythingSwyx [00:03:52]: actually like very surprisingly fast.Ethan [00:03:55]: One thing I say like thanks to my experience at NVIDIA, ‘cause first time when we were building Kosmos together, we built it, for about a year. So this is like the second time I do it. Roughly have an idea, what to do. I say the most important thing is the talent. Everyone were very strong and clever, very close with each other towards a common goal. So that speed up things a lot. So you reduce the communication bandwidth among people, and everyone can work towards the same goal. It's, it's like every day there's not that much meetings on the calendar, like maybe like a, like a sync a day, and after that it's, it's just all building. It was pretty fun at that time.Ethan [00:04:47]: And another thing is that xAI has very strong foundations of like data inference, model inference, and the supporting there can help the model develop a lot. When I look at, training models, I don't so actually the top important thing is like how many, how many iterations can you do, per day? and the more iteration can you do, you can, you can train the model much faster. So if you have very strong infra and you have a lot of compute, you can, you can train these models in very short period of time. That can give you a much larger buffer to, for errors, and it also gives you the opportunity to spot more bugs.Iteration Speed, Compute, and Debugging Model PipelinesSwyx [00:05:46]: What is an iteration? Is it like a few hundred steps or what are youEthan [00:05:50]: Let's say just the train-training the model, like from acquire new data and maybe design new algorithms and train a new model, maybe at smaller scale orSwyx [00:06:01]: So cycle time for like any hyperparam that you're searching.Ethan [00:06:04]: Cycle time and tune to like eval this model. Is this model better than my previous iteration?Ethan [00:06:11]: SoSwyx [00:06:11]: So it's like before you, someone had already set this up that you can iterate very quickly.Ethan [00:06:15]: I think the foundation there is extremely good forDeveloping and research models.Ethan [00:06:23]: And often I find is it-- this is kind of boring, but like a lot of the improvements does not come from new algorithms. It comes from finding small bugs here and there in the data pipeline, in the, in the model training pipeline. Those give, those give the biggest boost to the model quality.Vibhu [00:06:46]: It's interesting, right? So you say it's like small team, less communication bandwidth, but also a lot of quality is like find little bugs. It seems counterintuitive, right? You have a lot of people, you can iron out more of those, but it's interesting to see the other side, right?Swyx [00:07:00]: I also wonder, have you-- do you try using LLMs to look for bugs? I don't know.Ethan [00:07:05]: I remember at that time it was mid two thousand and twenty-five, so it's the coding model wasn't quite there yet. I remem- I remember like December two thousand and twenty-five, it was extremely good. Yeah, I've been, I've been using it at that time. It's, it's helpful. sometimes it produce codes that are kind of difficult to maintain, even though like the first time it built something extremely fast. But it gave the, like a spaghetti code, thousands of lines that I couldn't maintain, and the LLM itself couldn't figure out what's, what's wrong and how to improve on top of it. But now I find it much better. Yeah, I want to bring up another point here is now coding models are much more efficient and can help us implement stuff much faster. Compute might become a bottleneck again because previously, like if you want to train a new model, say you want to generate new synthetic data and then or write a new algorithm, it might take a few weeks. And during that period of time, you don't-- you might not have experiments to run. But now you can build that thing within a few hours, then you can immediately train a model.Ethan [00:08:24]: Now you have to have enough compute to try all of the ideas. So compute might be the bottleneck of iterating speed again.Swyx [00:08:36]: yeah, I actually, honestly, I think it's like kind of a stressful job because you're “Well, I should be trying everything, and if I'm not, then I'm not doing my job well.”Vibhu [00:08:48]: there's also the stress of you're eating thousands of GPUs per hour, which is very expensive and, compute can go to other researchers.Swyx [00:08:56]: You got the daddy Elon toVibhu [00:08:57]: You got daddy Elon.Ethan [00:08:59]: It wasVibhu [00:09:00]: But there's still finite amount of compute, like you want to use it, you want to use it well, you want more of it.Ethan [00:09:06]: That was quite stressful indeed. Yeah, I think one thing is the-- with coding models now, like a lot of these jobs can be automated, which is much better. A second, it's a, it's a marathon, so you got to maintain good health and, a regular schedule.Vibhu [00:09:28]: It's, it's hard to hear that when you shift from zero to nothing in two months.Swyx [00:09:32]: and, I think obviously the culture at xAI is very famously, people work very hard. one thing I did want to dive into, in our-- in the notes that you, that you sent ahead of time, you had specific comments about the cost of Video Gen training. presumably this is on the Colossus-1, right? the two hundred megawatt cluster. Any whatever you want to just share on that.Vibhu [00:09:54]: I think there's, there's three things we're talking about, right? So there's Video Gen, there's also the Image Gen model that you put out. Do you want to like complete the, okay, so zero to one, you have a few months. Just what are the stages of create Image Gen model?Swyx [00:10:06]: Oh, yeah, maybe I got distracted.How Image and Video Models Are Trained: Synthetic Captions, Tokenizers, and VAEsVibhu [00:10:07]: Sorry. and then, from there's Video Gen, there's Audio Gen. Would love to get into those next. But what is that first few months like? So small team, a lot of bugs, iterations, but what does it look like? Do we take something off the shelf? Do we just get data compute? What's, what's the few months like? How do you go to state-art Image Gen model? How do you just start?Ethan [00:10:28]: I cannot comment specifically how xAI did, but it's, it's a quite standard process. I can draw some, examples from Cosmos. So mainly it's building a video model, you actually need to build a image model first. And building these two models, the data you need is a hundred percent synthetic pair of language and image or language to video. Because on the, on the internet, actually, the videos don't naturally associate with text. So you can say, oh, like on YouTube, you have the title and you have the description and the commentsSwyx [00:11:11]: TitleEthan [00:11:11]: of a video, but usually they're not relevant to the video itself. And say maybe like the video is a natural scene of mountains or something, and the title is, I'm so happy today.Ethan [00:11:26]: So they have they have no correlation at all. So the first step is to, you have to generate synthetic pair of language with the videos. So you gather videos from the internet, and you use a VLM to caption the videos. So that part, here's a question, like how do you, how do you gather VLM to begin with? So if there's noSwyx [00:11:55]: You, so you fuse the model, right? LikeEthan [00:11:57]: Say if there's no like VLM exists, like how do you generate the text to the beginning, right? It's, it's impossible.Swyx [00:12:04]: I see.Ethan [00:12:05]: In the beginning, it's like you ask human to describe the video as detailed as possible.For example, you ask them to describe everything, like all objects, all characters, and all interaction and dialogues in the, in the videos. So that's in the protocol of Cosmos labeling. We require the objective we give to the labelers was that you have to describe the video as detailed as possible, such that a blind person hears a blob of text can reconstruct what the video is like from their head.Swyx [00:12:43]: Video or image? You're talking about images.Ethan [00:12:44]: Video or image, either one of them.Vibhu [00:12:47]: This was pretty common when we went from clip and DALL-E, right?Vibhu [00:12:51]: It's all training on really detailed captioning of images. So same is applied to video, but insteadEthan [00:12:57]: same appliedVibhu [00:12:57]: of using multimodal model to pass in video images and write rich descriptions, you can alsoSwyx [00:13:04]: I think there's this traditional perspective of supervised, or, very highly human curated thing. I feel like there's a unlock with unsupervised, right? Where like you have enough to bootstrap that you can just throw common corpus on it or, whatever. like unsupervised vision and language pairing, right? Like where you just have, interspersed image and text and it just learns. To me, that is the VLM breakthrough that is different from the clip, different from the LM era.Ethan [00:13:36]: It's interesting to see that you kind of need both data.Ethan [00:13:41]: For example, for theSwyx [00:13:41]: You need it to bootstrap it up. YeahEthan [00:13:43]: for the generative model training, there's also usually like a small percentage of unlabeled data. So the model is instructed to generate a video without any text instruction. That can also help the model generalize. So after this stage of generative synthetic pair, so, one important common step is to train a compressor or a tokenizer of the image or videos. So because, if you train-- If you can technically, theoretically train image or video models on pure pixels, but the problem is that the, it's, it's a lot of tokens. So like one image, it's, a thousand by a thousand, it's like one million tokens, one million pixels. It's impossible to train transformer on that. So it's, you need to train a tokenizer, which can go from image to latent space and latent space back to image.Swyx [00:14:45]: That's why we named the podcast.Swyx [00:14:48]: But, basically, you're talking about vocabulary science.Ethan [00:14:50]: so vocab.Swyx [00:14:51]: And so, what is, what is imp-- like a million is impossible?Ethan [00:14:54]: In generative models, the vocab is continuous. It's a continuous space. We can think about like you map an image to a vector. It's a, it's a fixed length vector. It's sixteen or forty-eight, something like that. And then you map that vector back to the image space. And the mapping is, has-- The mapping is patch-based. So you say you haveEthan [00:15:22]: a sixteen by sixteen patch and you match, you map that patch of pixels into this latent space.Swyx [00:15:29]: We've covered thisVibhu [00:15:30]: This is like the vision transformersSwyx [00:15:32]: VAEs,Ethan [00:15:33]: VAEs.Vibhu [00:15:34]: You basically compress your input, you do your generation, you're reasoning all that generation in smaller dimension, and then you project back out.Swyx [00:15:43]: VAE is a form compression, but I think the for me, the patching thing is from VIT, right?Ethan [00:15:48]: You can make those.Swyx [00:15:49]: Literally the, yeah, the paper is titled like sixteen by sixteen is all you need. something like that. and then I think also, people make a lot of comparisons with this kind of patching with convolutions.Swyx [00:16:02]: Which is you're, you're kind of re- reconstructing the old paradigm with the new.Ethan [00:16:05]: Actually, in VAEs, there are, there are both convolution networks and transformers. You can actually do both.Ethan [00:16:14]: After this VAE, so what you've got is you've got latent space tokens and you've got the language tokens. So now the training of the diffusion transformer, usually generative models use diffusion transformers. It is actually quite standard. It's, it's very similar to how you train a language transformer models. It's not that much difference. It's just the tokens, the visual tokens in, visual tokens out. The only difference is there's a denoising process. So you train the model to unmask some of the noise. So you add, you add random noise to the visual tokens, and then you train the model to remove those noise to generate the clean tokens. Any inference, the model can iteratively remove noise from a hundred percent noise.Swyx [00:17:12]: And then there's also, to speed things along on the tech tree of diffusion, there's CFG, and then there's, there's also, latent diffusion that, there's, there's someone in there. I think, somewhere along the line, obviously, like stability and all these other guys, pioneered a lot of this, architecture. I don't know if you want to get into that or just, or do the video side up to you.Bootstrapping Video from Image Models and Temporal CompressionEthan [00:17:37]: After you train such model, such image model, the reason it's a, it's a foundation for video models is that image models are cheaper to train, and they have much denser connection between language and text. So, sorry, language and images. For example, you train a billion, you train on a billion images, and there's a mapping from the text to the image. And the cost to train the same, like the, a billion, a billion text to a billion videos, that's much more expensive because videosNaturally have more tokens than images. Because the diffusion models, their understanding of, language purely come from this mapping. So if you don't have enough mapping, so if you only train on like a ten million videos or something, there-- you might not see enough language tokens in your training, so your model does not understand human intention enough. So that's why you really-- you train-- you first train this image diffusion models, and then you bootstrap the video model from there.Swyx [00:18:53]: One thing I did want to ask, because I-- actually, I think you're, you're the first per-- video model person I've ever talked to, I think. we've, we've like talked to Luma and all those folks. There's all these tricks in video compression where basically frame by frame there's not that much difference, so actually you don't have to regenerate or save the whole frame, right? but I think MP4 compression or something else like that.Swyx [00:19:16]: is it tempting to use that? Or as far as I can tell, everyone just treats it as, “No, we would just generate every frame.” Is that roughly the state-art?Ethan [00:19:27]: There are a few different approaches. Let's say first, like you want to just directly use MP4 compression and use that as the tokens for the transformers to train, right? So people actually have tried that, but the main challenge is the latent space for the MP4 tokens were not, were not very comprehensible for the models. It's, it's extremely hard to train on that. And there's aEthan [00:20:01]: So that's why they created VAEs, which creates more continuous, latent space, so the models can understand that latent space and learn from it much easier. Even within the VAEs, there are different difficulties of the latent space. So you can imagine something the simplest, the most naive VAE is like you have an image, and you just shuffle all of the images into a, into a vector. So you don't need to train any VAEs, right? But that latent space is extremely hard for models to train on top of. That's why there are some debate on like how do you compress the tokens. So you mentioned like you can compress frame by frame. Also, you can compress, the temporal dimension.Ethan [00:20:52]: The difference is if you compress the temporal dimension, you get a much higher compression rate. Because there's temporal redundancy between frames, because, this frame and the last frame, likely they are mostly similar, so there's only some small difference. for example, I think in 12.1 VAE, they have like a eight by eight by four compression rate. So the four temporal tokens are compressed into one tokens. That can save a lot of, save a lot of the context length. If you do it frame by frame, you have to do maybe like eight by eight by one. Your context length will be four times larger. That being said, the benefit of the frame-- per frame compression, we might come back to this later, is, real-timeness and interactivity. ‘Cause if you, if you strain the output of the model, frame by frame, you can-- the model can respond to any user request immediately. So if you have like a temporal four compression, four times compression, thenSwyx [00:22:06]: It might be laggyEthan [00:22:07]: there's a lag there in nature.Swyx [00:22:10]: So you're very pilled on this. let's just go ahead and bring it up ‘cause we have the visual prepared anyway. There's some frontier applications of real-time video gen. So Flipbook is one of the examples that went viral recently, right? What is Flipbook?Real-Time Generative UI: Flipbook, Neural OS, and Diffusion Front EndsEthan [00:22:23]: Flipbook is kind of like a web brow- web browser. You can see like it has the web bro- browser UI on top. The difference is all of the UIs are generated by generative image model in real time, and anything here are fake. But you can, you can explore inside this wor- this imaginary world. Say like we-- here we have engineering the Great Pyramid. Like the model generates this for us to understand how it works, and if we want to navigate around and understand further, we can click on some of the, some of the description here, and the model will generate a new page, new subpage describing the details we want to know about.Swyx [00:23:14]: So it's basically kind of we're playing a video, but it's pausing for our next interaction, and then it just plays the next thing based on our interaction.Swyx [00:23:23]: Which is kind of cool.Vibhu [00:23:25]: and you kind of decide your story. So this was, how do you make a pyramid? levering technique seemed interesting, right? It shows how do you take Okay, I want to know what is thisSwyx [00:23:35]: The demo, the demo tweet had more animation between frames.Vibhu [00:23:38]: I think it's just skipping,Swyx [00:23:39]: Oh, it's just skipping a lot of frames.Ethan [00:23:40]: they also have a video modeVibhu [00:23:42]: It takes a lot. There's a lot of peopleEthan [00:23:42]: but, a lot of people are using it.Ethan [00:23:45]: So it's not available.Vibhu [00:23:46]: There's a live video stream. We can try,Swyx [00:23:50]: So this is an example of the kind of future that you see at the extreme. We don't-- we're obviously not in it today.Swyx [00:23:56]: But in a world where inference is completely free this is better than generating code and text?Ethan [00:24:02]: So this is, this is a final state of where Viva will be at for word model, I think. Imagine internet doesn't exist, and then you type in google.com. Like what should, what should, what should a model show you?the model can imagine something, and this is what the model imagine. And these web pages, they completely do not exist. So I think as the inference costs come down, we are going to have generative UI for everything. If you think about how the coding model works, so they write code for a web page, and they render the code might be con- converted into binary, and the binary render the pixels on the screen. So we in machine learning, every time we have some breakthrough, obviously it's, it's more intuit. So why don't we have like user instruction to the pixel directly? So the generative UI will be user intention to the pixels directly. And say like even if I want email, let's say everyone have the same interface, but I want, I want it slightly different. I want the email to show to me like a TikTok, so I can swipe left and right for the emails. And or maybe you want something else. We can have completely different things. Or like I have I'm looking at, Instagram stories, and I don't like the Like button. I always may click it. And, generative UI resolved it. So it's going to be a revolutionary replacement of the interface. So in the future, we might have much more powerfulEthan [00:25:50]: LLMs and coding models running behind the scene. And in the, in the front-end, the diffusion model will actually be the front-end to show stuff to you. That's how I imagine it.Swyx [00:26:02]: Diffusion front-end, deterministic back-end.Swyx [00:26:04]: Something like that. I find that very expensive, but,Vibhu [00:26:08]: I find it interesting you called LLMs writing code on the back end deterministic, but okay.Swyx [00:26:14]: you write it onceVibhu [00:26:15]: Compare it toSwyx [00:26:16]: And then you execute.Ethan [00:26:17]: If you think about the cost, say, let's say H100 costs $1 per hour, and if you use this eight hours a day and thirty days, so, every month you're paying this two forty, you'll actually not wanna pay for that. That's even more expensive than Cloud Code Max. But if you think about the compute costs come down like two times every year, and I think the future will likely arrive like within few years.Vibhu [00:26:49]: It's everything, right? compute cost comes down, compute gets faster, model gets smarterEthan [00:26:54]: More efficientVibhu [00:26:54]: model gets smaller.Swyx [00:26:55]: I don't know why you say two times, ‘cause I think it's like 100 times. In language models, it is roughly one hundred to a thousand times every twelve to eighteen months, for the same given level of LMSys, ELO.Vibhu [00:27:08]: That's a net of everything, right? That's model performance alongside compute. So different than just compute costs come down. But, a very interesting future.Swyx [00:27:19]: So the web designers will have to shout out that accessibility is an issue, right? how do you deal with screen readers or whatever. But yes, this is higher bandwidth storytelling than anything you can possibly generate with code, right? So I think that's the rough idea.Ethan [00:27:34]: And I'd like to add a little bit that so human naturally have the maximum bandwidth when we are looking at things, look at videos, and we also have maximum output bandwidth when we are talking. So in the future, it might be something like we talk to AI models, and the AI model responds back with a generative UI. So that would be the maximum input and output bandwidth to interact with AI models before neural link happens.Vibhu [00:28:06]: And it's also very custom, right? Some people are very visual, some people are not as visual, right? They prefer the text. But the best thing about generative UI, right, it can also be text.Swyx [00:28:17]: There's another project that we wanted to highlight, which is the Neural OS. Kinda similar idea, but here you're literally operating, simulating an operating system with a video model.Swyx [00:28:27]: and you can play Doom, you can do Firefox. I find this like mildly less impressive, obviously, because it's an OS that I can run.Swyx [00:28:37]: But here everything is imagined.Vibhu [00:28:40]: I was, used to the Command+W to close the Firefox tab. It didn't crash. That's why I saidSwyx [00:28:45]: It's too immersive.Vibhu [00:28:46]: It's, it's too immersive for me.Swyx [00:28:47]: Too immersive.Vibhu [00:28:48]: I wanted to close the tab.Vibhu [00:28:49]: But yes, I can play generated diffusion.Swyx [00:28:51]: this is shockingly fast.Swyx [00:28:54]: Because I remember there was a demo about like maybe one to two years ago. Someone tried to do the first-person shooter with a image model. There was no consistency. It was very slow. But here it looks like realistically it's-- this is Doom.Vibhu [00:29:07]: I think there's two sides to that, right? There's okay, what is running a game? The heavy part of it is actually the game engine, all the lighting, all that stuff, the graphics. This is just kind of video, right? Like we've solved consistency. This is still, it looks like a few years old image generation. There's some temporal consistency, but it's, it's kind of just images stitched together as frame video. But it's a good visual representation to pi- to picture the future you wanna see, right? that's, that's what I see in these more so.Ethan [00:29:38]: This reminds me of how the video models gets better and better. So Neural OS is kinda if you just look at it feels like it's just a crappy version of the, like the Windows we could have, right? And, but the difference is, so the model, this model is overfitted on the existing operating systems. It can generate nothing different than that. But it's actually also similar to video models. So when we are training these video model, image model, we train them on internet. There's no imaginary supernatural stuff on the internet. But once we train this model, you can prompt the model to generate something supernatural that have never existed in the data set. So if you train your Neural OS or neural computer on the standard screen recordings on the entire internet. The model can imagine completely new interface to interact with the computer.Swyx [00:30:43]: This is one of those things that is magical to me. usually generalizing out of distribution is bad, but somehow we have learned some kind of internal world model that you say, this plus, but it looks like rainbows and butterflies, it'll do it and it will kind of make sense.Swyx [00:31:03]: So yeah, that's kind of cool. Yeah, I don't know if there's any comment more on there. I do, I do wanted to, I did wanted to touch a little bit more on the model architecture stuff, which I think you were getting. It's, really fascinating. We don't get a chance to talk about this enough. So one of the papers that we covered, we've covered every annual, segment anything release. and I don't know if you follow-- you're a computer vision guy, so youEthan [00:31:26]: I knowSwyx [00:31:27]: . So they did memory attention, which is kind of interesting. And I always think, anything where you can, across the temporal dimension, keep some consistency, I think it's, very fascinating, and I don't know if Basically, does that-- the CV side bleeding into video gen side, I think is underexplored, right? we talk about it for labeling, but actually you can borrow the architecture itself.Ethan [00:31:50]: There's, there's also complete different approaches, right? you brought up the term world model, so we went from video model to world model. There is diffusion, but there's also other approaches that people are doing. So maybe we get into those after as well,?Swyx [00:32:03]: He has a whole definition of world models and stuff. I feel like we threw a lot at you. Whatever you want to comment on.Why Video Models Are Expensive: Storage, I/O, and Training ScaleEthan [00:32:10]: I think one thing that we should actually comment back on is okay, so we were talking about the steps to train image gen to video model. One thing we don't see as much of is okay, you brought up the delta in training data, right? SoEthan [00:32:24]: you won't have as much a video model might not generalize, but what is the cost of training a large video model? So we know for LLMs roughly, okay, even like the poolside thing that came out today, right? It's a Gemma level model trained on roughly forty trillion tokens at this many H200s over this much time, right? You can see what is the exact cost of that. So how many GPU hours over how much H200 costs? So how do we do the back-end math of, same thing for video models, image models. How do you, how do you kind of break that down? I can share some back-envelope calculation. So surprisingly, video models is-- the cost is very-- is comparable to language models and obviously the largest scale is language model, maybe like a medium scale to language models. I said just storing the videos alone, it costs a lot. You can, you can maybe look up on AWS or something.Ethan [00:33:20]: You really, say if you have a billion videos and let's say, let's just say like each video, like five megabyte, then you need five petabyte to just store those videos. And also remember we talk about you use a VAE to compress the videos, and you also need to store, typically you need to store those continuous feature, in-- also in your storage. That's also comparable size with the videos themselves. So just storing these videos and the features is tens of petabytes alone. And,Swyx [00:33:58]: I just, I just looked up the calculation. Five petabytes on S3 Standard is one hundred K per month.Ethan [00:34:05]: AndSwyx [00:34:05]: It's comparableEthan [00:34:05]: and you needSwyx [00:34:06]: AndEthan [00:34:06]: And then like tens of petabytes, two hundred K. And even more expensive is you have the ingress and egress.Swyx [00:34:13]: Oh, yeah.Ethan [00:34:14]: Like you-- through the internet. You have to just to download those videos, I believe it's, it's more expensive on AWS than just storing those videos.Swyx [00:34:25]: Storing, yeah.Ethan [00:34:25]: And each training runs, you probably need to pull them once. If you train multiple times, it's, it's even more than that. So it's like just storing the network, those costs is just, it would be a few, a few millions per month to just storing everything, not to mention the GPU cost.Ethan [00:34:45]: AndSwyx [00:34:45]: my side tangent, the compute rental, like GPU rental is very efficient. There's one side, okay, you can be XAI and build your data center. Should we not just build our, storage compute as well? LikeEthan [00:34:57]: Of courseSwyx [00:34:57]: cloud cost compared to just,Ethan [00:34:59]: You save so muchSwyx [00:35:00]: store. Yeah, exactly.Swyx [00:35:01]: Especially with like egress and stuff. So.Ethan [00:35:04]: That's a good idea, but it also comes to-- there are some of its own challenges.Swyx [00:35:09]: Of course, of course.Ethan [00:35:10]: like people who build the GPU data centers, they might not expect this much, storage. And yeah, people build storage, typically they just build it somewhere with just CPUs.Swyx [00:35:23]: I just looked it up. Five-- AWS only charges for egress, not ingress. Tier five for five petabytes is two hundred and thirty K.Ethan [00:35:32]: Even more expensive than the storage.Swyx [00:35:34]: But storing is per month, right? You check in, then you cannot check out. so it's so cool. It's okay. So there's that side.Ethan [00:35:41]: So the TLDR, my backhand mathSwyx [00:35:42]: Data is larger than you think. Yes.Ethan [00:35:44]: my backhand math of GPU hours times GPU cost is also very much, I'm missing some storage.Swyx [00:35:49]: You're also-- you're basically like also more IO bound than normal training.Swyx [00:35:55]: Yes. ‘Cause like data loading, so caching everything, it becomes super important.Ethan [00:36:00]: So in Cosmos, we did a lot of optimizations to make it not IO bound. So, speaking of the training, actually training the model, the GPU cost, if you look up like the open source model, how big these video models are, I think like LTX has nineteen B parameters. That's a dense model. And people are also exploring, MoEs, so it might be twenty B active and, like a hun- hundreds B, total. So that's, that's even-- that's similar size as medium-sized LLM models. And if you, if you look at number of tokens-Uh, we disclose that in Cosmos. It's also like tens of trillions of tokens on the visual tokens. So putting this together, the cost of, training these video models, it's actually comparable with LLMs. Not to mention, the infra is slightly different from LLM, so it might be less efficient to train these models.Inference Speedups: Step Distillation, Consistency Models, and GANsSwyx [00:37:04]: Do you get the benefits of traditional diffusion speed-up? So for, images, there's LCM, LoRAs for, fine-tuning. There's, there's a lot of stuff that's beenEthan [00:37:15]: Flow matching.Swyx [00:37:16]: there's flow matching. There's a lot of stuff that's been done. there's some overlap that applies to diffusion on the inference side and stuff or?Ethan [00:37:23]: so the difference-- the inference side is a completely different story.Ethan [00:37:28]: I think for the training side, it might be a little bit hard to reduce that cost. And for the inference side, the biggest gain is from the distillation of these models. You can-- It's called step distillation, slightly different from knowledge distillation in LLMs. So you-- Typically, for flow matching models, you need like 100 steps or something. Like a distortion model even need even more, like 1,000 steps to generate a good image or video. A step distillation is try to learn to generate fewer step from the model itself. It's kind of like now we-- you use the full model to generate in 100 steps, and then you take a model that only generate 10 steps and let that model to learn from the perfect one.Ethan [00:38:25]: why this workSwyx [00:38:27]: Strong to weak seemingly.Ethan [00:38:28]: It is. It's kind ofSwyx [00:38:29]: DistillationEthan [00:38:29]: kind of like strong to weak. the-- from the modeling perspective, the strong model, the teacher model is trying to model the image and videos of inter-internet, and that distribution is extremely complex. But the step distilled model is just trying to learn from the teacher. The teacher is a model, and the size is fixed, as the distribution is much simpler than the whole internet. That's the intuition I have why step distillation can work. So usually these models serve in productions, they only run in a few steps. In Cosmos, I believe we have, we have like four step and eight steps. If you do some simpler task, image-image translation, it can even run in fewer step, like one step in Cosmos Transfer.Swyx [00:39:22]: I think this is the same intuition that guides a lot of the consistency model work. I sent you a link for, SCM. I don't know if you covered that. To me, that was actually one of, the most impressive papers I've ever seen from OpenAI.Swyx [00:39:34]: That this is the unifying grand concept of consistency models. I don't know if you have any comments on this.Ethan [00:39:41]: So there are, there are a few different approaches,Swyx [00:39:46]: Oh, yeah. Here it is.Swyx [00:39:47]: Two steps versus twenty or 100 steps, whatever. It's already done.Ethan [00:39:52]: So there are, there are a few different approaches, for example, consistency model, and there are also Actually, we shouldn't forget GAN. So GAN, actually, that was, that was the OG ofSwyx [00:40:05]: OGEthan [00:40:05]: step distillation ‘cause it trained just one step to begin with. So actually, a lot of, uh-- For example, there's a distribution matching distillation which use, which uses GAN, as one of the laws for distillation. It-- GAN just tells you, “Hey, generate an image,” and thenEthan [00:40:31]: it has a discriminator to tell, is this image real or not? So the model, the model just need to learn one of the distribution, not the full distribution. Because in training, the model is asked to reconstruct the ground truth image from the internet, which is extremely hard. And in-- When you're training GAN, it's a step process. It's just a, “Hey, you generate image. Does this image look as real as the image from the internet?” Which is a much simpler task. And, yeah, combining a lot of these approaches together, people typically do that, like consistency model and distribution matching and GAN, and we can get these few step models.Audio-Video Generation and Time AlignmentSwyx [00:41:21]: Then there's one step I wanted to add, which is audio and video.Ethan [00:41:26]: So, Grok Imagine zero point nine, I believe it's, it's a first audio video transmodel deployed at a large scale. SoSwyx [00:41:39]: And that was your first model?Ethan [00:41:40]: that was, Grok Imagine's first model. It's, it's audio video, joint generation. I think the hard part is, the modality alignment, ‘cause before this transmodel, we have, we have text to video alignment. We have this, correspondence between text and video. Typically, most of the VLMs, they understand images and videos. Video's very rare, and they don't understand audio mostly. And if you look at the audio generation on the LLM side, you can talk to them perfectly fine, but if you ask them to sing a song or something, it typically is not very good. Also, they don't have, they don't have music either. The hard part is thatUh, actually audio has two component. It has like a discrete component, a continuous component. The discrete component is like the language.Ethan [00:42:44]: So when we speak, it's just, someSwyx [00:42:47]: It's an ASR issue, yeah.Ethan [00:42:49]: It's, it's text token with some characteristics, I would say.Ethan [00:42:54]: But musicSwyx [00:42:56]: I think the speech guys would disagree with this.Swyx [00:42:57]: Like disfluencies and then,Vibhu [00:43:00]: There's tones you can get angry.Ethan [00:43:01]: Well, I say largely.Ethan [00:43:03]: the mu- but the music is completely different. It's, it's very continuous, and you cannot model them like discrete tokens in language models. this is like the hard part for models is, not to mention we have to align text, video, and audio together.Ethan [00:43:26]: SoVibhu [00:43:26]: How?Ethan [00:43:28]: So significant-- some significant challenges are like-- So first, like we talk about as the VLMs, they cannot understand most of them cannot understand audio.Ethan [00:43:39]: So you have to have some way to do the synthetic data generation for audio. You have to caption the model, and that involve, that involve synthetic data and human data effort a lot. And not just surprisingly, most of the LLMs are very bad at recognizing, like the beat, tone, and the details of the of music. They can, they can give some general prediction of which song is this, but it's very hard to describe the details of the music. like we mentioned in image generation, like you have to describe image as detailed as possible so that someone blind can reconstruct that. So here is like someoneVibhu [00:44:32]: DeafEthan [00:44:32]: someone deaf can reconstruct how the music sounds like without actually listening to it. Maybe you can think of it need to have the-- or they call the script.Vibhu [00:44:49]: Subtitles, yeah.Ethan [00:44:49]: You gotta have all the details of the music, and the dialogue.Vibhu [00:44:55]: So is the challenge there typically stuff like music and audio, or is it just Like is there a baseline? Okay, there's enough data where we can understand, narration, conversation, but there's nuances in audio that's where you hit all the data issues or is it just from stage zero, you just do it all right?Ethan [00:45:15]: So one important thing is like the alignment. So the model, the model has to know like the video and audio, the, uh-- it has to have a time-based alignment, like at which time step the video and the audio token correspond to each other. But we actually don't have this kind of alignment for most of the other modalities. If you think about like text and image, text and video, they are loosely aligned. So you can, you can have a description of what's going on in the video, but you don't have to exactly, You typically don't have exact description, oh, at, time step one second like what happened?Vibhu [00:46:02]: It's veryEthan [00:46:03]: At time step two second what happenedVibhu [00:46:03]: coarse. Yeah.Swyx [00:46:05]: So what was the ideal time step? You have to oblate it, and then it's like four seconds or something.Ethan [00:46:09]: So that comes down to how you design the model to, for the model to be aware of as a time, as a time modality. So the model is like a time aware. And that's something pretty unique if you think about LLMs. So if you ask LLM to complete a task, say they, uh-- you ask them and they will say, “Oh, this task will probably take twelve hours to complete,” and they come back in one hour. Say “I've already spent two days on this and I've exhausted everything.”Ethan [00:46:47]: So the LLMs them-themselves, they don't have a sense of time there.Vibhu [00:46:53]: I actually don't think that's just them not having a sense of time. I think it's somewhat based, right?Vibhu [00:46:58]: Like you tell someone, “Okay, go work on this feature. Go implement this,” there's a general understanding you would have of how long that would take without LLMs working at LLM speed, right? So you think back like two years ago, if I tell you to like build me like a new front end for latent space, have a search bar, have all this, you'll estimate that it'll take a few days, right?Vibhu [00:47:19]: So you tell an LLM, “Go build this.” It'll take me a few days. But I think it's somewhat grounded as opposed to them not having the best-- Not saying that they have a great understanding, but I think that example is like you can see where it comes from, right? You're trained on all over the text.Swyx [00:47:35]: They're, they're trying to estimate what a human would say.Vibhu [00:47:37]: because that's what the, that's what the data kind of represents. It's not themEthan [00:47:41]: It came from the corpus on the internet. People have a estimate of how much time.Vibhu [00:47:45]: And not even just in direct like training samples, right? Just your world understanding of tokens of how long stuff takes, right? Go read a book. It'll take you a while, right?Vibhu [00:47:56]: Even if you do nothing but read a book, it takes a few days. So yeah, LLM, I read it took me a few hours.Vibhu [00:48:01]: It'll take me a few hours to go through this research. But this is a tangent.Swyx [00:48:05]: Somewhat, yeah.Swyx [00:48:06]: This is a train of thought I haven't really expressed until now is, which is basically like a full world model must also be recursive, meaning that the participant in the world model must also be aware that they have a world model. which is like this whole recursive thing down the, down the line. but yes, and that the world model can be wrong and that they need to update it and blah. Yeah. We've, argued this on the, newsletter as well, that there needs to be sort of recursive or adversarial world models.World Models: Real-Time, Long-Horizon, Interactive VideoVibhu [00:48:34]: just, to ask, how do you define world model?Swyx [00:48:38]: Oh, yeah, let's go there.Ethan [00:48:40]: SoVibhu [00:48:40]: So just for context, we talked about, video generation, and then there's a-- if you say there's a distinction between world models, what's your, what's your definition? How do you see the two?Ethan [00:48:53]: So disclaimer, I'm not going to debate, what is world model. Yeah. there are many definitions, so I'll just talk about my definition. Since I came from the multi-model, multi-model domain, so mainly talking from video. So world model is like real-time interactive long horizon videos. So there are three parts. so we-- let's talk about them one by one. So the so interaction, so we just, we just look at Facebook and neural computer. So the interaction part of it, so you, world model can allow you to interact with them through keyboard, mouse, and maybe also voice. So these all is-- all is a modality. You can, you can interact with the model, and the model should respond reasonably. Second part is real time. So once you, once, say, you move your mouse, if, say, the world model generate a game, how fast can the game respond? So if you're like professional CS: GO players- -my say, oh, you have to respond- He's beginner within sub ten milliseconds or- Yeah even less. So that's not most of the- No, sixty FPS. Let's go. Oh, three hundred FPS. Oh, five hundred FPS. Wait. okay, yeah. I didn't do the math, but yeah, okay. Uh- Yeah, three hundred FPS, that's a three millisecond. So you have to respond- Oh, s**t. Okay. YeahEthan [00:50:29]: within a millisecond. Most of the video models cannot do that. Yeah. And, but if you, say, if you have a video model that is, say, like a digital human, the response time might be more generous. Maybe typically, for real-time voice interaction, it's like two hundred millisecond. So that's, that's much more generous. But even two hundred millisecond is pretty, it is pretty tricky, ‘cause remember we mentionedEthan [00:51:01]: you have this, temporal compression coming from the VAE. So if you, if you don't compress the temporal dimension, your sequence length is going to explode. So if you want to have this real-time, real-timeness in your model, you have to do is one context problem. And the third part is long horizon, ‘cause we-- if you're not going to just play with, video games just, a few seconds, most video models only a few seconds. We're going to play with minutes, hours. The model have to be able to generate long-form content.Ethan [00:51:42]: So putting these three together, it's, real-time, long horizon interactive videos. I think the final state will be, for example, like a video, a video version of Playbook, where you can, you can interact with, a neural computer. You move your mouse, and you click on the generative interface, and it will reply to you through pixels- generating in real time. But getting there, it's, it's a very long way to get there. So one of the first step, at Grok Imagine, where I led a small world model team there, was to build video extension. So, video extension- it's the first step of interactivity. Yeah. It's, it's the first step. Yeah. So it's the first step- You have it here, video editing, yeah. Yeah. Yeah. So the first step is because, this unlocks long horizon videos. Typically, for most of the video generation models, you give it a prompt or an image as an initial frame. You generate video, that's it. That's just, one time, done. And some creators would try to, use the last frame as a first frame for the second video. It can-- sometimes it works, but if you do it a few times, it says the quality would decrease. And- It doesn't have that context- Yeah over the full video, so the temporal- Yeah, exactly. Yeah, ‘cause you only gave it the last frame, of course, right? Yeah. Exactly. And- it's actually a pretty fun hack. if you've seen like- Oh, no, he's saying something better. Yeah. And for example, like Vue, I remember Vue 3 has like a second context of the last video. It is slightly better than using the last frame, but it has the same problem-- similar problem that it, the quality would decrease. if you extend a few times to, one minute, the video quality would look much worse than the first video. Second, another problem is that the model doesn't have long-range knowledge of, what's happening before. Say, if they generate some dialogue, some, two people speaking, and their voice might change, over some time, especially if the second conditioning, it does not cover the previous context. So these are the core challenges. So the Grok Imagine video extension, it has historical context of all of the previous generated videos. It can, It has, it has the context of, who is speaking and what objects have appeared and everything, having that to generate the next video. So if we naively do this, you can imagine, just, put all of the previous history video tokens into the context. The context lens will easily explode. Especially for video models, that can be like a few, a few million context, I would imagine- context lens. Yes.Yeah.Swyx [00:54:58]: Let's run with that.Ethan [00:54:59]: for example, like in Cosmos, I think just five seconds of video is like a fifty K or sixty K number of tokens. So like if you do, if you do fifty second, that's a five hundred K tokens. If you do longer than that, easily explode. This long horizon, problem was the first step we're trying to solve world model. It turns out people, yeah, people love video extension. Like a lot, a lot of the creators love using video extension to create longer form videos. This is the part I liked that you have a, you have an intermediate step toward the final goal instead of just a straight shot to the final version very much.Swyx [00:55:48]: But I can see you have a strong vision of where we want to end up.Long Context, Redundancy, and Efficient Interactive VideoVibhu [00:55:51]: Does it seem like it's an efficiency issue? okay, we're at a few million tokens context,. If you draw the parallel to language models, we had very short context, two thousand, eight thousand, then, you scale it up one million, ten million. sure, there's effective context, but at the end of the day, it's just what's it worth? sure, there's a whole training data side. In video, it might be slightly easier ‘cause we have a hundred million token video, right? Just take a movie with the full context there. Like is this efficiency from an inference standpoint that like it's expensive, but we know how to solve it? Or like why is this not the approach? So like my broader point was on your second point of world models, you say it needs to be interactive and live, right? You should be able to play a game and see the interaction live. So one thing I see with research is a lot of what you actually serve is different than what you build, right? So we talked about distillation. You train big model, you distill it, you do quantization, speculative decoding. We do all this stuff to serve it efficiently. Should we not just have a solution, like a world model that can interact well, do inference optimization, serve it, distill it secondary, so make it real time after you solve it? So like a-- another parallel is say, continual learning, right? What we need is someone to solve it and show it works inefficiently. Give it a few years, people will make it efficient. Same thing with regular attention, right? It worked. Over a few years, people have different forms of attention, and we've scaled it to be efficient at log context,? So kind of two things there, right? One is it seems like it works. You've scaled it. Can we not just scale it a lot more efficiently over time? Do we need a separate approach if this works? And same thing with interaction, right? if we can get it done, like if we can solve some way that it works, we can solve making it more efficient from an inference standpoint later.Ethan [00:57:53]: that's actually a very good point. So in videos, there's actually a lot of redundancies. So we solve a lot of the pixel redundancy from VE, but there's more redundancy in long range and long horizon videos. Say, if a character appear in the first clip and then it disappeared, it only reappear at the end of the video, you probably don't need the-- the context, like in the middle of the generation. So you only need that character, where you need. So that's why, I helped build another feature. It's a reference video.Vibhu [00:58:36]: Is it here?Swyx [00:58:36]: is it the same model release or different one?Ethan [00:58:39]: It's a different one.Ethan [00:58:41]: You probably need to search onSwyx [00:58:43]: I'll find itEthan [00:58:43]: X reference to video.Ethan [00:58:46]: So reference video allow you to like upload up to seven images as condition and generate the video. Say, if like I want-- it can, it can be characters or objects or even scenes. Say like I want, I want condition on, Sean's selfie and holding a bladeSwyx [00:59:07]: We have a dogEthan [00:59:08]: or whatever.Swyx [00:59:08]: We put the dog in the thing.Ethan [00:59:09]: you can put them there and the video models will generate the video from and copies the context over. So that can solve a lot of the problems there, like the long context problem. It doesn't need to have a very long context, but it's-- I feel like it's an intermediate solution. The modelSwyx [00:59:29]: It's cheating.Ethan [00:59:30]: the model should be able to like selectively know, where should I draw the references. So say if I want to generate a movie, I generate it autoregressive, like a ten second at a time or something. And now this character appear, I can look back to where it first appear and, bring that back. Yeah, this one, I put the references. Yeah, that's, Optimus, Einstein myself, Annie.Vibhu [01:00:02]: Oddly enough, I used Grok Search to find it, and it pulled your LinkedIn post. But yeah we found it.Ethan [01:00:08]: Interesting.Vibhu [01:00:10]: ButxAI's Underrated Work, Culture, and WatermarkingSwyx [01:00:11]: this is a problem. This is not your fault, but like XAI doesn't communicate all this work that you do very well because they just have the model release and then that's it. But actually, these details are very good.Swyx [01:00:22]: As far as I understand, everything you just described is state-art, like no one else has done it.Vibhu [01:00:30]: A lot of-- yeah, I have a lot moreSwyx [01:00:32]: And then, and then you just put this blog post with the cookies. I'm this is not enough,?Swyx [01:00:37]: but I, obviously this is like the high level numbers that people want to know. But no, okay, soVibhu [01:00:42]: And I wonder, like part of that is also some labs don't share research into what happens. And ifSwyx [01:00:50]: No, but this is literally bragging about how good they are, right?Swyx [01:00:54]: Like, why would you not say that you are capable of extending with full context? this is not a secret sauce. This is like we did the work. yeah, I don't know.Ethan [01:01:02]: different labs have slightly different communication styles.Swyx [01:01:07]: Anyway, if anyone from XAI is listening we are always happy to help you tell your story. Yeah, okay, so you did references, and I think, I think kind of the point you're, you're making is it is sort of like a kludge, right? this is-- you can do seven, but what about 100?Swyx [01:01:23]: Right? Then you need a completely different thing.Ethan [01:01:26]: So I think it's-- this is, a mechanism to, select the context from the history, and you might not put the entire history into the context. for example, there's a paper called Frame Pack, which haveEthan [01:01:41]: a heuristic that the latest history, the last one second, I put the entire history, and the history before that, I would, compress it and makes the video smaller. So they follow this pattern, this build overall pattern that the maximum sequence length is fixed. So the further you are from the current frame, you have a smaller image. So this is just a heuristic. I think it can be more automatic. The model is aware like which history part of it can be select. So this part of the research is actually being actively, worked on by a lot of people. It's also quite interesting. I feel this is actually, this part of long context is a little bit ahead of the LLM part.Ethan [01:02:31]: So for example, like in LLMs, if you-- so contexts keep growing. Let's say if you call tool and the tool call history is extremely long, that's still in context, and keep growing, keep growing. Even if you switch the topic to something else, the whole context was there. There are some agentic harnesses that help you to, say, prune the tool results and, prune Like when you, when you query a file, only show like the top 200 lines or something. Those were very heuristic-driven.Swyx [01:03:08]: For listeners, we did a write-up on the cloud code, leak where there are eight different kinds of pruning, including like you prune the tool results and all that. So you can, you can read up on that kind of thing.Ethan [01:03:17]: I think, one breakthrough in continual learning might be like a way to automatically, manage its own context.Swyx [01:03:27]: These are all heuristics, and they will be replaced by machine learning.Ethan [01:03:30]: InterestinglyVibhu [01:03:32]: TheEthan [01:03:32]: the same thing is being researched in both LLMs and video models.Vibhu [01:03:36]: The interesting thing is also like in the paper you showed, it's actually happening at the model level, right? Compared to like language models, sure, we have base attention, but we'll do our own compression, we'll do our own pruning, which is separate from model error.Vibhu [01:03:49]: Eventually, it all just boils in, hopefully.Swyx [01:03:52]: I think this is a form of like attention, but like also know sort of reasoning attention. I feel like that's different than normal attention.Swyx [01:04:03]: Does that, does that make sense?Ethan [01:04:04]: It's, it's different in the sense that attention, not to mention, set sparse attention aside,

In a conversation with CancerNetwork®, Nathan Goodyear, MD, spoke about the role that exercise and lifestyle intervention can play in the treatment of patients with cancer. He described how prescribed exercise may serve as a biologically interventional therapy that can help prolong longevity, reduce the risk of recurrence; and supplement the efficacy of standard therapeutic approaches like chemotherapy, immunotherapy, and surgery.Goodyear, an integrative medicine physician at the Williams Cancer Institute, pointed to literature indicating the potential benefits of structured exercise programs across different cancer populations. For example, data from the phase 3 CHALLENGE trial (NCT00819208) highlighted a lower risk of death and reduced recurrence following a 3-year structured program among patients with stage II and III colorectal cancer. Furthermore, the OPTIMUS trial (NCT02950324) demonstrated that a short-term exercise program that takes place before surgery or alongside chemotherapy can increase CD8-positive T-cell infiltration while decreasing immunosuppressive cells, effectively turning “cold” tumors “hot.”Additionally, Goodyear addressed some preconceptions surrounding the potential role of exercise in oncologic care, defending it as a prescribable therapy that necessitates a deliberate, properly applied approach to achieve success among patients. He discussed the importance of structuring individualized exercise-based regimens by considering performance status and other physical patient characteristics. He also noted how exercise intervention may mitigate immunosenescence and accelerated aging may be associated with one's disease and anti-cancer therapy. “Surgery, chemotherapy, and radiation…have efficacy; there's no question about that. They also promote senescence and accelerated aging. What if we're able to bring in these therapies that can work to break those cycles, like exercise?” Goodyear stated. “If it improves the outcome, helps the patient heal better, empowers their immune system in intended [and] direct ways that are reproducible in the research, and if it helps to block that accelerated aging, we reengage the immune system, countering the immunosenescence that is accelerating that process called inflammation.”References Courneya KS, Vardy JL, O'Callaghan CJ, et al. Structured exercise after adjuvant chemotherapy for colon cancer. N Engl J Med. 2025;393(1):13-25. doi:10.1056/NEJMoa2502760 Rayner CJ, Bartlett DB, Allen SK, et al. Prehabilitation during neoadjuvant chemotherapy results in an enhanced immune response in esophageal adenocarcinoma tumors: a randomized controlled trial. J Sport Health Sci. 2025;14:101063. doi:10.1016/j.jshs.2025.101063

The new Transformers pinball machine from Stern Pinball is finally here… but so is the controversy surrounding it. In this episode, we dive deep into the rumor that Optimus Prime was originally supposed to transform in the game, where that rumor may have started, how the pinball community ran with it, and whether content creators helped create expectations that Stern themselves never officially promised. We also discuss: • The origins of the transforming Optimus rumor • How leaks and insider information shape community expectations • Whether Stern Pinball is being unfairly criticized • The responsibility of pinball media and influencers • The hypocrisy surrounding “confirmed” rumors • Why social media and speculation can spiral out of control • The actual gameplay, code, and features people are LOVING about Transformers And to be fair, we also talk about the positives:

Are traditional coaching models becoming obsolete in the age of AI and increasingly personalised approaches? In this episode of The Coaching Hub, I explore how the coaching landscape is evolving and ask whether well-known models like GROW, CLEAR, and OSCAR still hold their place in a world where AI can already replicate structured conversations.I reflect on my own journey with these frameworks and why they continue to matter, particularly as foundational tools for newer coaches building confidence and structure in their practice. At the same time, I share the growing shift towards more individualised, relational coaching that moves beyond rigid models and into deeper, more human-centred work.We'll look at the balance between structure and flexibility, the limitations of a “one-size-fits-all” or “sausage machine” approach, and what it really means to create transformational impact for clients today. If you're thinking about how to evolve your coaching style and stay relevant in a changing industry, this episode offers a clear, honest perspective on what's next.What's inside:Why traditional coaching models like GROW, CLEAR, and OSCAR are still widely used but increasingly questionedHow coaching model training provides confidence and structure for new coachesThe limitations of rigid, model-first approaches, especially given individual client needs and neurodiversityWhy AI can easily coach using models, making transactional, model-based coaching less distinctiveThe importance of moving toward more personalised, relational, and flexible coaching frameworksHow organisational settings may still benefit from a model-led approach for consistency and easeWhy effective coaching now centres on tailoring the experience to each client, fostering true transformation and awareness.Like, comment, follow and subscribe for more inspiring conversations on growth and professional development.Key Topics Covered:00:00 Intro00:34 Why traditional models are being questioned02:03 The impact of AI on coaching models03:22 Why models still matter for new coaches04:07 Moving toward flexible, client-centred coaching05:59 When models still work (especially in organisations)07:49 Knowing when to use (or let go of) models09:00 Transformational coaching beyond frameworks09:52 Final thoughts & wrap-upAbout Ruth Kudzi:Ruth is the founder of Optimus Coaching Academy and a well-respected and successful coach, speaker and author who has worked across sectors including leadership, career and more recently business and mindset. She has over 10,000 coaching hours and has completed hundreds of hours of training and coaching supervision. Ruth is an MCC-level coach with the ICF and is our course director and CEO. Prior to becoming a coach, Ruth was a senior leader in education. Find out more here: https://ruthkudzi.com/ Book:How to Feel Better: 4 Steps to Self-Coach Your Way to a Happier More Authentic You eBook : Kudzi, Ruth: Amazon.co.uk: Books Connect with Ruth: YouTube: https://www.youtube.com/@RuthKudzi Facebook: https://www.facebook.com/Ruthkudzi2/ Instagram: https://www.instagram.com/ruthkudzi/ LinkedIn: https://www.linkedin.com/in/ruthkudzi/ Podcast: https://podfollow.com/the-coaching-hub-podcast About Optimus Coach Academy:Optimus offers best in class coaching training for individuals and corporates. If you want to know more about what we offer: https://optimuscoachacademy.com/coach-training We also offer business support as standard at Optimus, find out more here: https://www.optimuscoachacademy.com/ Connect with Optimus Coach Academy: Facebook: https://www.facebook.com/Optimuscoachacademy Instagram: https://www.instagram.com/optimuscoachacademy/ LinkedIn: https://www.linkedin.com/company/optimuscoachacademy Produced by Winter Audio.

O Senado volta a ter protagonismo em Roma. Porém, tem seu holofote roubado por um dos maiores conquistadores dos últimos tempos: quem? como? Veja bem. Mais.Contate-nos: vejabempodcast@outlook.comPIX: e3257213-46ea-4c97-9740-4c6f268baa0fReferências:⁠78-80 - Nerva -Optimus Trajan⁠⁠⁠ – podcast, The History of Rome⁠⁠⁠⁠Trajan: Optimus Princeps⁠⁠ - YouTubeTrajan - Rome's Last Conqueror Documentary - YouTube⁠⁠Pax: Guerra e Paz na Era Dourada de Roma⁠ --livro, kindle

In this episode we answer emails from TJ, Jose and Optimus Bill.  We discuss the foibles of trying to catch up via investment picking if you are behind on retirement, debunk CAPE-style and other crystal ball forecasts from "experts" that Level Two investors often fixate upon, lay out practical growth-tilted allocations that can beat narrative-driven investing and invite you all to contact Optimus Bill about your Risk Parity Radio listening habits.And THEN we our go through our weekly and monthly portfolio reviews of the eight sample portfolios you can find at Portfolios | Risk Parity Radio.Additional Links:FI Service Corp DC Charitable Event:  DC Double PlayFather McKenna Center Donation Page:  Donate - Father McKenna CenterMichael Batnick Critique of CAPE Ratio "Predictions":  Stocks Are More Expensive Than They Used to BeAccumulating With a Golden Ratio Portfolio Article:  Minimize Your Miss – Portfolio ChartsCatching Up to FI Episode 100:  0️⃣ From Zero to Hero: A Late Starter's Guide to the Galaxy

The EV industry is evolving fast and so are the biggest automotive and technology festivals in America. On this episode of the Turn Down For Watt Podcast, we sit down with BJ Birtwell, founder of Electrify Expo and the newly expanded Demo Days Festival, to discuss why one of the nation's largest EV-focused touring events is transforming into a massive future mobility festival featuring EVs, hybrids, robotics, Tesla Optimus, air mobility, eVTOL aircraft, autonomous technology, and hands-on demos from major OEMs and brands like Tesla, Rivian, Lucid, Ford, Toyota, and more.

What does true mastery look like in coaching, and how do Master Certified Coaches set themselves apart? In this episode of The Coaching Hub, I share the reality of mastery beyond flashy techniques or the latest tools. Whether you're an aspiring coach, pursuing your MCC, or just curious about the journey to excellence, this episode offers a candid look at the path toward coaching mastery.I reflect on my own experience training as a Master Certified Coach and dive into the difference between coaching the person versus coaching the problem and why genuine presence, partnership, and curiosity matter more than any model or framework.You'll discover how the best coaches bring deep self-awareness, embrace not knowing, and use space and precision in service of their clients. I unpack why mastery is a never-ending journey, how every coaching conversation is shaped by our state of being, and why human connection and transformational impact can never be replaced by AI.What's insideCoaching mastery isn't about techniques or tools, but presence with clientsCommitting to continuous learning and growth, not believing oneself to be a finished "master"The importance of a growth mindset in becoming an MCCBeing present and non-judgmental allows for deeper conversationsA defining trait of master coaches is focusing on the person as a wholeMastery takes a holistic, human-to-human approach that cannot be replicated by AI.Key topics covered:00:00 Intro00:03 Mastery in coaching is about who you are in the room01:02 Focusing on continually learning and evolving01:54 Master coaches are deeply committed to improving their craft02:39 Recognising and addressing blind spots03:34 Adopting a growth mindset for improvement04:23 Mastery is about coaching the person05:30 Operating without attachment to being right07:12 Genuine curiosity reveals deeper insights.08:13 At Master level, questions are sharper 11:13 Courageous challenge defines transformational coaching.About Ruth Kudzi:Ruth is the founder of Optimus Coaching Academy and a well-respected and successful coach, speaker and author who has worked across sectors including leadership, career and more recently business and mindset. She has over 10,000 coaching hours and has completed hundreds of hours of training and coaching supervision. Ruth is an MCC-level coach with the ICF and is our course director and CEO. Prior to becoming a coach, Ruth was a senior leader in education. Find out more here: https://ruthkudzi.com/ Book: How to Feel Better: 4 Steps to Self-Coach Your Way to a Happier More Authentic You eBook : Kudzi, Ruth: Amazon.co.uk: Books Connect with Ruth: YouTube: https://www.youtube.com/@RuthKudzi Facebook: https://www.facebook.com/Ruthkudzi2/ Instagram: https://www.instagram.com/ruthkudzi/ LinkedIn: https://www.linkedin.com/in/ruthkudzi/ Podcast: https://podfollow.com/the-coaching-hub-podcast About Optimus Coach Academy:Optimus offers best in class coaching training for individuals and corporates. If you want to know more about what we offer: https://optimuscoachacademy.com/coach-training We also offer business support as standard at Optimus, find out more here: https://www.optimuscoachacademy.com/ Connect with Optimus Coach Academy: Facebook: https://www.facebook.com/Optimuscoachacademy Instagram: https://www.instagram.com/optimuscoachacademy/ LinkedIn: https://www.linkedin.com/company/optimuscoachacademy Produced by Winter Audio.

This week on Wanye's World we open the show with a little business and IT talk and the growth of World Wide Sports Magic. We talk some Oilers and the coaching issues. Wanye tells us how close we may be to finding Bigfoot, Ocean vs Space and an unofficial update on Optimus 3 Hosted on Acast. See acast.com/privacy for more information.

This week the show briefly becomes obsessed with magical swords and loopholes around enchantments, and despite one of the episodes being named 'Alpha/Omega' we manage to only make a single Omegaverse joke! This datatrack contains discussion of the following topics; burger crimes, Excalibur cheat codes, bandicootmaxing, Optimus becoming daddy, Megatron's creative problem solving, particle effects, magical weapons throughout fiction, Medieval fujoshis, and sword dates. Noise Space | Discord | Patreon This podcast is powered by Pinecast.

Play NowThe Twincast / Podcast starts up its 401st episode by diving right in to the latest upcoming products shown off by Takara Tomy. Missing Link Ironhide and New Legends Nemesis Star Convoy lead the way. After the crew opines on those, the larger slate of reveals from this year's Shizuoka Hobby Show are discussed, including Beast Wars II Galvatron, Gigastorm, Missing Link Ratchet, Overgear Optimus Prime and even the AM-T Star Saber inspired by the character's design in the IDW comics. The latest Hot Wheels collaborative products are talked about next, followed by a quick look at the upcoming new Wild King toys. Listener questions about anniversaries and Optimus Prime toys lead into a quick installment of the Bragging Rights segment to close out the episode.

One of the biggest pieces of news on Tesla's conference call was an update on Optimus. Tesla is building a 1M unit per year factory in Fremont, California that is set to start production in July or August of this year. They are also building a 10M unit production line in Austin that is set to begin production next summer! Things are heating up! They are keeping the prototypes secret because of copy cats, but one thing is forsure Optimus is making crazy progress and will be coming soon. The big reveal when production starts is going to be crazy.My X:   / gfilche  HyperChange Patreon :)   / hyperchange   Disclaimer: I'm long Tesla this is not financial advice.

▶ 2026知識衛星高峰會 ◀ 知識解構，價值重構，讓 AI 時代的機會，向你傾斜

SpaceX could invest as much as $119 billion to build its Terafab, a proposed semiconductor manufacturing and computing fabrication facility in Texas, according to local documents tied to a tax abatement request. The Commissioners Court of Grimes County will hold a public hearing on June 3 to consider the deal.The notice states that SpaceX would locate the multi-phase, vertically integrated Terafab at the Gibbons Creek Reservoir and surrounding areas between Austin and Houston. The project's initial phases could involve up to $55 billion in investment, with total spending potentially reaching $119 billion if additional phases move forward.The project represents a collaboration between SpaceX, Tesla and xAI, all companies in which Elon Musk serves as CEO. According to a social media post shared by Tesla, the Terafab would manufacture 1 terawatt of chip output per year, which the company claimed would exceed the combined output of all global chipmakers.Neither the companies nor local officials announced a timeframe.SpaceX revealed Terafab in March and described it as “the next step towards becoming a galactic civilization.” The company said the plant would make chips for Tesla's electric vehicles and Optimus humanoid robots, as well as SpaceX spacecraft.Tesla designed the bipedal, autonomous Optimus robot to perform dangerous, repetitive and boring tasks. According to CBS News, Musk said the automaker could begin selling the robot in 2027.As for space travel, the Terafab aligns with SpaceX's broader plans involving solar-powered AI satellites and future space operations.#SpaceX #Tesla #xAI #ElonMusk #Semiconductors #ChipManufacturing #AI #DataCenters #Manufacturing #Texas #TechNews #FactoryOfTheFuture #Terafab #Robotics #Optimus #ArtificialIntelligence #IndustrialAutomation #SemiconductorIndustry #FutureOfManufacturing #Automation #AdvancedManufacturing #SpaceTechnology #Computing #SupplyChain #Gigafactory #MadeInAmerica #SmartManufacturing #TechInnovation #IndustrialNews #Engineering

Tech companies are betting big on robots that look like humans and do human jobs. Why, robot? This episode was produced by Avishay Artsy, edited by Jolie Myers, fact-checked by Gabriel Dunatov, engineered by David Tatasciore, and hosted by Sean Rameswaram. An Optimus humanoid robot showcased at a Tesla booth. Photo: CFOTO/Future Publishing via Getty Images. Listen to Today, Explained ad-free by becoming a Vox Member: vox.com/members. New Vox members get $20 off their membership right now. Transcript at ⁠vox.com/today-explained-podcast.⁠ Learn more about your ad choices. Visit podcastchoices.com/adchoices

Paris Marx is joined by Ben Tarnoff and Quinn Slobodian to discuss their new book Muskism which explores how Elon Musk exemplifies a new economic system shaping our lives, similar to Fordism in the twentieth century. Ben Tarnoff & Quinn Slobodian are the authors of Muskism. Ben is a writer and technologist based in Massachusetts and the author of Internet for the People. Quinn is professor of international history at Boston University, and the author of books like Crack-Up Capitalism. Tech Won't Save Us offers a critical perspective on tech, its worldview, and wider society with the goal of inspiring people to demand better tech and a better world. Support the show on Patreon. The podcast is made in partnership with The Nation. Production is by Kyla Hewson. Also mentioned in this episode: For listeners who are feeling extra academic, here is the Milton Friedman economics paper, “The Methodology of Positive Economics.” Quinn discusses his struggle to find any reporting on Jared Leto and the Optimus robot media stunt (that goes deeper than commenting on the virality).

In today's edition of BizNews Daybreak, we dive into President Trump's strategic move to sustain an economic blockade on Iran over military alternatives. The global market faces further upheaval as the UAE announces its departure from OPEC following decades of membership. Meanwhile, King Charles III delivers a historic address to the US Congress, emphasizing the critical importance of the Atlantic partnership and unyielding support for Ukraine. In local news, environmental activists challenge Eskom's nuclear expansion in the Eastern Cape, citing threats to unique biodiversity and water security. Plus, Alec Hogg analyzes Tesla's $25 billion gamble on the "Optimus" robotic future and unpacks the latest quarterly tech earnings.

Description: In this episode, we take a deep dive into Tesla's Q1 2026 earnings call, featuring critical insights from Elon Musk and Vaibhav Taneja. The conversation centers on the technical limitations of Hardware 3 for unsupervised Full Self-Driving (FSD) and Tesla's massive $25 billion capital investment strategy. We explore the production roadmap for the highly anticipated CyberCab and semi-trucks, alongside the surging demand for Megapack energy storage. The team also breaks down the complexities of the Optimus robot project and the delicate balance Tesla must strike between futuristic innovation and investor expectations. Finally, we look ahead to upcoming discussions on bipartisan EV initiatives and practical urban carbon footprint solutions. Support the Show: https://www.supportkilowatt.com/ Other Podcasts: Beyond the Post YouTube Beyond the Post Podcast Shuffle Playlist 918Digital Website News Links: Tesla Q1 2026 Earnings Call Livestream *Show Art Created By Gemini Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Episode OverviewThe Butcher Shop goes deep on one of the Great Lakes predator fly world's most distinctive patterns in this conversation with Eli Berant, the Michigan-based fly designer and founder of Great Lakes Fly. Eli is the creator of the Optimus Swine — a reverse foam head-embedded, side-kicking musky streamer that has been turning heads and producing fish since around 2009. In this episode, host Marvin Cash walks Eli through the full arc of the pattern: the lake musky problem it was designed to solve, the unconventional decision to reverse a foam popper head to create a slower fall and a pronounced glide-bait wiggle, the material choices that define the fly's profile and movement and the step-by-step construction logic from spinner bait hook to laser dub head.The conversation covers the full Swine family — the original 8–9 inch version on a 6/0 Mustad, the scaled-down Swine Junior for river smallmouth and stripers, the fettuccine-foam Pot Belly Swine for subsurface river applications, and the articulated Maximus Swine and Maximus Swine Junior, which remain something of a "secret menu" offering. Eli also addresses color selection by region — from olive-and-pink for fired-up Tennessee muskies to the Wisconsin-proven Willen's Villain black-white-yellow combo and his own favorite Mardi Gras pattern — and breaks down his preferred line and leader systems for lake musky versus river smallmouth applications. Throughout, the discussion grounds fly design theory in direct, tactical fishing application.Key TakeawaysHow reversing a foam popper head toward the rear of the hook creates a slower fall rate and induces the Optimus Swine's distinctive side-to-side glide-bait action.Why proportionality in bucktail application — specifically how much material per section and how many sections — is the most common failure point for tiers attempting the Swine for the first time.How to tune the Pot Belly Swine's fettuccine foam piece by removing individual strips to achieve neutral balance and proper swim orientation before fishing.Why a jerk-strip retrieve with a sinking line (350–450 grain tip) is the preferred delivery system for lake musky, allowing the sink tip to hold depth while the fly kicks side to side on each pull.When to dial back retrieve aggression and employ a stutter-strip or extended pause with the Swine Junior, particularly during cold-water conditions when bass are holding and waiting.Why sharing newly discovered synthetic fly tying materials openly — rather than hoarding them — is essential to keeping those materials in production and available to the broader tying community.Techniques & Gear CoveredThe Optimus Swine is designed around a jerk-strip retrieve that drives its foam-induced side-to-side action, and Eli breaks down exactly how to execute it — stripping two feet with the line hand in alternating pulls, roughly like ripping a bag open. For lake musky, he runs a 10-weight with a 350–450 grain sinking tip, paired with a short 3–4 foot leader from loop to fly — a butt section of 40-pound to wire, finished with cross-lock snaps for fast fly changes. River smallmouth and striper applications drop to a 7- or 8-weight with a 200–350 grain tip depending on conditions. Construction-specific details are substantial: Mustad 32608 spinner bait hook (6/0 for the original), Rainy's Mini Me medium foam popper head reversed and goop-set with silicone adhesive, synthetic yak hair blended with flash for the tail, grizzly saddle feathers as flanks, Magnum Flashabou, everyday bucktail applied in top-and-bottom sections, laser dub for the head, and 1/2-inch eyes pressed and held in a two-touch goop cure process. Anadromous Fly Company tungsten carbide scissors get a specific callout as Eli's go-to cutting tool for heavy production tying.Locations & SpeciesThe Optimus Swine was developed specifically for lake musky, with Lake Saint Clair in Michigan serving as the primary proving ground — a relatively snag-free fishery that allows anglers to fish sinking lines freely across the water column. The pattern's documented multi-species versatility extends to Great Lakes migratory species, pike, lake trout, stripers on the East Coast and river smallmouth, including Eli's personal use of the Swine Junior on Lake Saint Clair for targeting large smallmouth by eliminating the smaller fish. Color selection is explicitly regional in the episode: olive-and-pink for fired-up Tennessee fish, pink-and-chartreuse or the Willen's Villain black-white-yellow for Wisconsin tannic water, and Mardi Gras (pink, chartreuse, black head) as a broadly effective pattern.FAQ / Key Questions AnsweredHow does the reversed foam popper head make the Optimus Swine swim differently than other musky flies?Positioning the foam head toward the rear of the hook — rather than at the front — reduces the fly's sink rate compared to a traditional epoxy-head pattern and shifts the center of buoyancy rearward. This causes the fly to kick side to side with a pronounced glide-bait cadence on a jerk-strip retrieve, rather than simply pushing water or diving. The effect is amplified when fishing a sinking tip, which holds the running line low and forces the rear of the fly to tip upward and roll on each strip.What are the most common mistakes tiers make when tying the Optimus Swine?Eli identifies two primary failure points: applying bucktail in clumps that are too large, which destroys proportionality, and using too much laser dub in the head, which throws the silhouette out of balance. The fix for bucktail is learning to feel the correct bundle size — roughly the width of a toothpick at the pinch, the width of a popsicle stick at the ends — and building five top-and-bottom sections before reaching the laser dub head on the original Swine. Managing the laser dub means stacking it, pulling off loose fibers and removing material rather than adding more.How do you tune the Pot Belly Swine to swim correctly for river applications?Because the Pot Belly Swine uses fettuccine foam strips in place of the reversed popper head, Eli ties in more foam strips than needed — six to eight — and tells buyers they may need to remove one to four strips to get the fly to balance and swim true. The goal is first to eliminate any spin or tilt, then to dial in the side-to-side action. This is the same principle as Barry Reynolds's flash philosophy applied to buoyancy: put in more than you need because you can always remove it, but you can't add it once the fly is finished.What line and leader setup does Eli prefer for lake musky with the Optimus Swine?For open lake musky fishing on snag-light water, Eli runs a 10-weight with a Scientific Anglers sinking tip in the 350–450 grain range, specifically preferring striper-style lines with a long 26–28 foot tip section. Leaders are intentionally short — 3–4 feet total from loop to fly — built with a 2-foot 40-pound butt section going straight to wire, then a cross-lock snap at the fly. The short leader keeps the fly in the sink tip's depth zone and maximizes the kicking action on the jerk-strip retrieve.How should retrieve style change when downsizing to the Swine Junior for smallmouth or stripers?Moving to the smaller patterns calls for a less aggressive retrieve cadence overall, but Eli emphasizes breaking out of monotonous repetition — consciously varying the retrieve is as important as the base technique. Key adjustments include a stutter-strip (half-length pulls done twice in quick succession) and extended pauses, which become particularly effective in cold water when bass are holding and watching the fly. The foam piece in all Swine variants allows the fly to hang suspended during a pause without sinking, which is the primary trigger for following fish.SponsorsThanks to TroutRoutes for sponsoring this episode. Use ARTFLY20 to get 20% off of your TroutRoutes Pro membership.Related ContentS1, Ep 2: The T-Bone: A Deep Dive with Blane Chocklett - The Butcher ShopBONUS: Shack Nasties and the Drunk & Disorderly: A Winter Chat with Tommy LynchBONUS: Crafting The Nut Job: A Deep Dive with Brendan RuchBONUS: A Deep Dive into the Swingin' D: Techniques and Tips with Mike SchultzS6, Ep 124: Fly Tying with Chase SmithConnect with Our GuestFollow Eli on Instagram.Follow the...

A.M. Edition for April 23. Tesla shares have slipped off-hours despite surprising Wall Street with better revenues - and rising car sales. WSJ's Becky Peterson says investors are worried about the price tag for Elon Musk's AI plans, including the new Optimus robot. Plus, Defense Secretary Pete Hegseth has fired Navy Secretary John Phelan in the latest shakeup at the Pentagon. And Senators approve a budget plan to fund DHS, which will hand ICE and Border Control an additional $70 billion, despite Democratic opposition. Luke Vargas hosts. Sign up for the WSJ's free What's News newsletter. Learn more about your ad choices. Visit megaphone.fm/adchoices

Today's top stories, with context, in just 15 minutes.On today's podcast:1) The US and Iran are locked in a battle for control of the Strait of Hormuz after failing to meet for a fresh round of peace talks, with both sides blocking the waterway to gain leverage during an extended ceasefire. The US is maintaining a naval blockade on ships going to and from Iran’s ports, which Iran calls a violation of the ceasefire, while Iran is keeping Hormuz closed to almost all other international traffic. The blockade and closure of the Strait of Hormuz have led to increased oil prices and concerns about supply shortages and a global inflation crisis2) Navy Secretary John Phelan was fired after clashing with top leaders at the Pentagon, including over administration efforts to revive US shipbuilding. Phelan was pushed out after butting heads with Defense Secretary Pete Hegseth and Deputy Defense Secretary Stephen Feinberg over President Trump’s focus on a new US “Golden Fleet." Undersecretary Hung Cao will replace Phelan, according to Pentagon spokesman Sean Parnell, who said Phelan was stepping down “effective immediately."3) Tesla anticipates billions of dollars in additional spending this year to support Elon Musk’s ambitions to transform the company into an AI and robotics company. Capital expenditures this year will exceed $25 billion, roughly three times last year’s outlay, to be put toward a dramatic expansion of factory operations and initiatives such as production of its Optimus humanoid robot. The investments will support production of key products including Cybercab, Semi and an updated version of its Megapack battery storage system, with Tesla remaining on track to start making these products.See omnystudio.com/listener for privacy information.

Het kabinet ziet mogelijk af van de versobering van de kleinschaligheidsinvesteringsaftrek (KIA), na verzet in de Tweede Kamer van onder meer VVD, JA21, SGP en ChristenUnie. Daarmee blijft een belangrijk fiscaal voordeel voor ondernemers bestaan, maar moet het minderheidskabinet op zoek naar 100 miljoen euro dekking voor eerder aangekondigde energiesteun en andere maatregelen, waarover vandaag verder wordt gedebatteerd bij de Voorjaarsnota. In Washington praten Libanon en Israël met Amerikaanse bemiddeling over een verlenging van het staakt-het-vuren, terwijl in Zuid-Libanon dagelijks nog aanvallen en vernielingen worden gemeld. Libanon wil een langere wapenstilstand en een einde aan Israëlische vernielingen in het bezette zuiden, Israël zoekt vooral garanties rond het ontwapenen van Hezbollah, terwijl de uitkomst sterk afhangt van de druk van de VS en de link met onderhandelingen met Iran. Tesla verrast beleggers met beter dan verwachte kwartaalcijfers: de nettowinst stijgt met 17% tot 477 miljoen dollar en de omzet neemt 16% toe vergeleken met een jaar eerder. Het bedrijf kondigt voor 2026 kapitaaluitgaven van 25 miljard dollar aan voor onder meer robotproject Optimus, eigen chipproductie en robotaxi’s, terwijl de markt afwacht hoe deze investeringsgolf en de groeiende vraag naar elektrische auto’s zich vertalen in toekomstige winstgevendheid. Deze omschrijving is met AI gemaakt en gecontroleerd door een BNR-redacteur. See omnystudio.com/listener for privacy information.

Tesla Q1 2026 earnings are hot off the press, and the biggest announcement wasn't in the results but on the conference call. Elon Musk and Tesla are now upping CAPEX guidance to $25B in 2026! That's right, they want to spend $25BILLION this year on investments on AI and robotics. This means they will be losing free cash flow and could mean they will have to raise more capital in the future. All to fund ambitious projects like Cybercab, Robotaxi, Optimus training, Terafab AI Chips & more! What are your thoughts on this massive news!??!My X:   / gfilche  HyperChange Patreon :)   / hyperchange   Disclaimer: Nothing in this show is financial advice I'm long Tesla.

Going live for the release of Tesla's Q1 2026 Shareholder Letter as the company reports earnings today! Updating investors on financials, Robotaxi, Optimus, Cybercab and more! $TSLA investors join in!

Tesla (TSLA) sales are flat for a third year in a row, according to Steve Westly. He makes the case that the Mag 7 giant needs to accelerate its AI, robotaxi, and Optimus prospects as some investors continue to question Tesla's fundamentals. Tom White turns to an example options trade for Tesla ahead of the company's earnings after Wednesday's closing bell. ======== Schwab Network ========Empowering every investor and trader, every market day.Options involve risks and are not suitable for all investors. Before trading, read the Options Disclosure Document. http://bit.ly/2v9tH6DSubscribe to the Market Minute newsletter - https://schwabnetwork.com/subscribeDownload the iOS app - https://apps.apple.com/us/app/schwab-network/id1460719185Download the Amazon Fire Tv App - https://www.amazon.com/TD-Ameritrade-Network/dp/B07KRD76C7Watch on Sling - https://watch.sling.com/1/asset/191928615bd8d47686f94682aefaa007/watchWatch on Vizio - https://www.vizio.com/en/watchfreeplus-exploreWatch on DistroTV - https://www.distro.tv/live/schwab-network/Follow us on X – https://twitter.com/schwabnetworkFollow us on Facebook – https://www.facebook.com/schwabnetworkFollow us on LinkedIn - https://www.linkedin.com/company/schwab-network/About Schwab Network - https://schwabnetwork.com/about

Play NowThe Seibertron.com Twincast / Podcast is back with Episode 399. The crew starts with a deep dive into the Takara Tomy web comic providing some insights on how G2 Menasor fits into their New Legends line. Age of Primes Commander Class Armada Jetfire, including the Requiem Blaster, then moves the discussion into new toy talk. Studio Series discussion follows, prompted by the launch of new pre-orders for upcoming products in that lineup. Leak lists then get some attention as they always seem to this time of year, with early indicators of what fans and collectors can expect in 2027. Listener questions then start conversation about how Megatron should be handled in any future installments of the live-action movie franchise, before deciding who the most canine Transformer would be that doesn't turn into a dog. "Bragging Rights" then brings this show to a close.

To conclude the Crisis of Command storyline, we have a heroic rescue from the solo battlin Optimus Prime. Its definitely not an easy fight, but why would it be?If you'd like to contact the guys, they'd love to hear from you!Morethanmeetstheseguys@gmail.comhttps://discord.gg/sKr8jwaAvhIf you'd like to toss a buck or more per episode, we'd adore and say nice things about you. You don't have to, as we'll still gladly hang out with you guys and gals every week, but we appreciate any help! patreon.com/user?u=69144181

Shoot us a Text.Episode #1318: Penske climbs the dealer ranks as consolidation continues, Tesla sends off Model S/X with a pricey Signature Series, and $4 gas is pushing consumers online.Automotive News' 2026 Top 150 Dealer ranking saw some notable movement as acquisitions and stronger same-store sales reshaped the leaderboard.Penske Automotive moved to No. 2, bumping AutoNation to No. 3, while Lithia holds the top spot yet again.Penske's growth was fueled in part by high-volume California and Texas store acquisitions now fully counted in 2025 results.The Top 150 sold 4.14M vehicles, increasing their share of total U.S. sales to 27%.Despite selling more cars, the Top 150 owns fewer rooftops overall—continued consolidation in action.81 groups moved up overall and 23 gained double-digit spots.Public retailers increased their share of Top 150 sales to 34.3%, highlighting their growing influence.14 currently represented at ASOTU CON: Lithia, Holman, Ourisman, LaFontaine, DARCARS, Walser, McGovern, Zeigler,  RML Automotive, American Motors Group, CMA, Huffines, Casa, Preston Auto Group.Tesla is closing the chapter on its flagship sedans and SUVs with an ultra-exclusive, invite-only “Signature Series” run. With just 350 units and premium pricing, it's a nostalgic—and pricey—farewell to the brand's roots.Tesla will build just 350 units (250 Model S, 100 Model X), available only via invite to select owners.Exclusive Garnet Red paint, gold badging, and numbered interiors highlight the collector-focused design.The pricing reflects rarity, with the Model X Signature hitting $159K—about a $30K premium.These models will mark the end of Model S/X production as Tesla shifts factory capacity toward Optimus robots.Elon Musk previously called it an “honorable discharge,” closing a chapter that started in 2012.Rising gas prices are pushing more shoppers to skip store trips altogether. A sharp spike in online spending suggests convenience—and avoiding the pump—is becoming a bigger factor in buying decisions.Online spending jumped 20% in March, far above typical monthly gains, as gas prices topped $4.Orders rose 12% and average order value increased 8%, showing bigger and more frequent purchases.In-store shoppers are consolidating trips, making fewer visits but spending more per trip.83% of consumers cite gas as a top cost concern, with many shifting to online to avoid driving.“When gas crosses a psychological price threshold, the math changes,” said Omnisend's Marty Bauer.Join Paul J Daly and Kyle Mountsier every morning for the Automotive State of the Union podcast  as they connect the dots across car dealerships, retail trends, emerging tech like AI, and cultural shifts—bringing clarity, speed, and people-first insight to automotive leaders navigating a rapidly changing industry.Get the Daily Push Back email at https://www.asotu.com/JOIN the conversation on LinkedIn at: https://www.linkedin.com/company/asotu/

On today's episode, we discuss everything from Trump's drawn‑out war with Iran to Tesla's fast‑evolving self‑driving software, future robotaxis, and the coming wave of home robots. James opens by grilling the “Fearsome Threesome” futurists—Glenn, Mark, and Dwayne—on whether Trump can find an off‑ramp in Iran, how turning the Strait of Hormuz into a toll canal might work, and why “breaking their arms but not fixing their government” risks long‑term instability. The conversation then pivots to Tesla's latest Full‑Self‑Driving updates: better road graphics, parking‑spot memory, “smart summon” for rainy‑day pickups, quirky voice commands, and an almost comical obsession with avoiding animals—even if that means a squirrel or armadillo gets priority over a human who “should know to move.” From there, they explore Tesla's broader ecosystem, including third‑party Supercharger build‑outs at abandoned gas stations, vehicle‑to‑vehicle communications, the Cybertruck's rear‑steer “crab walk,” and rumors of a Cyber‑SUV lurking in drone footage over Giga Texas. In the second half, the panel zooms out to Musk's Optimus robots and a future where bots clean garages, wash cars, cook lasagna, and free people to write novels or tend showpiece yards, while Mark warns that such freedom will still demand new kinds of responsibility and “management” of our machines. They close with a whirlwind tech‑finance lightning round—crypto as an “office commodity” with lost wallets and off‑grid keys, quantum‑computing races over qubit designs, AI‑driven corporate layoffs that Wall Street perversely rewards, and college students whose ChatGPT‑assisted assignments are homogenizing their voices in the classroom. Don't miss it!

Play NowThe Seibertron.com Twincast / Podcast kicks off Episode 398 reviewing the new teaser images of G2 Menasor and MPG Secret Agent Arcee from Takara Tomy. The cast then reviews the Autobot Symbol Optimus Prime, not metaphorically, but the literal Optimus Prime with an Autobot Symbol alt mode. Listener questions then continue the show starting with a prompt for examples where death was handled correctly in the Transformers Universe. The show then returns to lighter topics like new Autobot leader alt modes, head swaps to create new toy characters, and assembling Autobot/Decepticon strike teams. Listener questions finish up on debate if Beast Wars Megatron was an analytical thinker or a goon when choosing an alt mode. The recurring "Bragging Rights" segment then brings an episode to a close.

The gap between Elon Musk's ambitious promises for Tesla and the practical realities of the company's performance leading into 2026. While Musk continues to tease future innovations like the "Cybervan" and a steering-wheel-free Cybercab, the texts highlight significant delays in electric vehicle sales growth, the Optimus robot program, and the mass production of the Tesla Semi. Regulatory challenges are also central, as California officials clarify that Tesla's ride-hailing service is currently a standard chauffeur operation rather than a true autonomous "robotaxi" network. Furthermore, analysts express skepticism regarding the technical safety and data transparency of Tesla's self-driving software compared to competitors. Collectively, the reports portray a company transitioning toward artificial intelligence and robotics while struggling to meet previously established industrial and autonomous milestones.

Follow Forbes Talks Tesla on Thursday reported quarterly vehicle deliveries that fell below Wall Street's expectations, the latest sign of a disrupted electric vehicle market as Elon Musk's automaker shifts its focus toward robotaxis and humanoid robots. Key Facts Tesla said Thursday it delivered just over 358,000 vehicles in the first quarter, below the automaker's compiled consensus of 365,645 and consensus analyst projections of 381,000, according to FactSet. That marks a 14.3% decline from the December quarter (418,227), but a 6.2% year-over-year growth from Q1 2025 (337,000), when Tesla reported its fewest quarterly vehicle deliveries since 2022. Model 3 and Y vehicles accounted for nearly 342,000 of Tesla's quarterly deliveries, down nearly 19% from the previous quarter (406,585). Shares of Tesla declined 3.4% shortly after trading opened on Thursday. What To Watch For Tesla will report Q1 earnings after market close on April 22, the company said. The automaker is expected to report quarterly revenue of $22.9 billion and $0.41 earnings per share, representing what would be annual growth of 18.6% and nearly 52%, respectively. Key Background Tesla's vehicle delivery reports are often cited as insight into the automaker's sales ahead of its earnings reports. The latest quarterly slide in deliveries follows a broader decline in electric vehicle demand: EVs represented roughly 12% of the U.S. market in September, an all-time high, but that dropped to 6% by January, according to Cox Automotive. Tesla remains the market leader in the U.S., however, even as it faces growing competition from Chinese automaker BYD, which surpassed Tesla as the world's largest EV maker. Musk said in January the company would end production of its flagship Model S and Model X cars, announcing Tesla would use its production line in Fremont, California, to manufacture the company's Optimus humanoid robots. Earlier this week, Musk said orders for the S and X vehicles had “come to an end.” Forbes Valuation Musk is by far the world's richest person, with an estimated net worth of $823.8 billion as of Thursday. He's expected to soon become the world's first trillionaire, after his SpaceX filed confidentially for an IPO on Wednesday, leading the way for what will likely be the largest-ever market debut. Musk, who owns about 43% of SpaceX, would become the first person to be chief executive of two companies valued at $1 trillion after the aerospace firm's listing. Read the full story on Forbes: By Ty Roush https://www.forbes.com/sites/tylerroush/2026/04/02/tesla-misses-vehicle-delivery-estimates-as-ev-market-struggles/ Learn more about your ad choices. Visit megaphone.fm/adchoices

The global healthcare system is heading for a catastrophic workforce shortage — 10 to 15 million workers short by 2030. In this episode, financial analyst and Tesla expert Cern Basher (@cernbasher) returns to break down how humanoid robots like Tesla's Optimus could be the only scalable solution, why New Zealand is uniquely positioned to lead a healthcare robotics pilot, and what the Terafab chip factory and Digital Optimus mean for the timeline. From fleet learning to the privacy concerns, from Moxy robots already in 25+ US hospitals to Elon Musk's vision of billions of robots powered by space-based AI — this is the conversation that healthcare, tech, and policy leaders need to be having right now. IN THIS EPISODE: The global healthcare workforce crisis — why no recruitment drive can fix it Robots already in hospitals: Moxy's 1M+ deliveries across 25 US hospitals Tesla Optimus: where the technology is right now The three components: physical body, AI brain, and language model Fleet learning explained — 40 robots learn in 6 months what takes a human 25 years Why New Zealand is the ideal proving ground for healthcare robotics Digital Optimus and Macrohard: the software robot that runs businesses Terafab: Tesla's $25B chip factory with SpaceX and xAI This episode is sponsored by MitoSynergy Copper 1+ Most copper supplements use poorly absorbed oxidised forms that can actually increase free radical damage. MitoSynergy's patented BioCopper1 (Cunermuspir) is a copper-niacin chelate that delivers reduced copper (Cu1+) directly to your mitochondria, supporting ATP production at cytochrome c oxidase. I've been personally testing MitoActivator EX and have noticed a real difference in training power and energy output. Try it: https://mitosynergy.com/lisaTamati (10% off with this link) ABOUT CERN BASHER: NZ-born financial analyst and one of the most influential voices on X at the intersection of AI, Bitcoin, Tesla, and macroeconomics. Follow Cern: @cernbasher on X and YouTube -------------------------------------------- PTL SIGNAL — AI, Tech, Bitcoin and Markets: https://ptlsignal.com Free founding member access to our AI-powered Financial Document Analyzer — earnings transcripts, annual reports, Fed minutes analysed in 60 seconds. Take control of your health and unlock the secrets to a longer, healthier and more vibrant life: https://www.lisatamati.com/healthspan-hacks-course/ SHOP Longevity Supplements: https://shop.lisatamati.com -------------------------------------------- PODCAST — Pushing The Limits: https://www.lisatamati.com/podcast https://podcasts.apple.com/nz/podcast/pushing-the-limits/id1207975008 https://open.spotify.com/show/6mc5BfQispXYMxd4AaYXYL FOLLOW LISA: Instagram: @lisatamati X/Twitter: @lisaytamati YouTube: @LisaTamati Enquiries: support@lisatamati.com Website: www.lisatamati.com Shop: shop.lisatamati.com PTL Signal: ptlsignal.com

On today's episode, we discuss how rapidly advancing technologies—especially autonomous vehicles, AI, and humanoid robots—are about to reshape everyday life, work, and city infrastructure. James, Glenn, Mark, and Dwayne open with PJ's Coffee banter and then dive into self‑driving cars, using Waymo's 170 million driverless miles and dramatically lower accident rates alongside James's own Tesla “deer detection” stories to argue that human driving will become too risky and expensive to insure, especially in big cities. They explore knock‑on effects like cheaper robo‑taxis, fewer personally owned cars, reclaimed urban parking real estate, and drone delivery networks that could make one‑hour Amazon drop‑offs routine, even in smaller markets. From there, the conversation shifts to compute and energy, as they talk about data‑center power demand, Musk's proposed Terrafab chip complex as a domestic rival to TSMC, and massive new gas‑fired plants and possible micro‑nuclear solutions being built to feed AI workloads in Louisiana. The crew also reacts to multi‑agent AI systems like Grok, jokes about “investing in real intelligence,” and walks through how Tesla's Optimus robots could share vision data with cars, work in factories and restaurants, and eventually handle home tasks from cooking to yard work—upending both jobs and household roles. While they repeatedly muse about “Skynet” and communist China's ambitions toward Taiwan, they ultimately frame this wave of automation as a huge opportunity for regions that adapt quickly, universities that pivot into robotics and AI, and individuals willing to offload drudgery to machines and focus on higher‑value work and relationships. Don't miss it!

Arm moves closer to owning the silicon layer, OpenAI sharpens its enterprise strategy, and a wave of geopolitical and market pressures exposes what is really driving the AI race. Patrick Moorhead and Daniel Newman unpack how compute constraints, capital intensity, and supply chain risk are starting to dictate who can scale, who can compete, and who gets left behind as the industry shifts from experimentation to execution.  The handpicked topics for this week are: Arm Unveils AGI CPU — First-Ever In-House Chip, Co-Developed with Meta: Arm steps into direct silicon production with its AGI CPU, raising questions about vertical integration, ecosystem neutrality, and how this move reshapes competition across the data center landscape (The Decode) OpenAI Kills Sora and Doubles Down on Enterprise: OpenAI pivots away from experimental consumer products to focus on enterprise adoption, signaling a sharper push toward monetization and long-term business sustainability (The Decode) Tesla, SpaceX, and xAI Introduce Terafab: A new manufacturing and compute initiative highlights the growing importance of vertically integrated infrastructure in scaling AI and advanced technologies (The Decode) AI Data Center Moratorium Act: House senators introduced the AI Data Center Moratorium Act, proposing a federal pause on new AI data center construction until comprehensive regulations are established (The Decode) RSAC 2026 Signals the Rise of Agentic AI Security: This year's conference underscores a shift toward securing autonomous systems, with agentic AI emerging as a new frontier in cybersecurity strategy (The Decode) The Flip: Can Tesla Actually Build a Semiconductor Fab? Or Is Terafab a $10 Billion Fantasy? The debate centers on whether Tesla can successfully build its own semiconductor fabrication facility to produce custom AI chips for its autonomous vehicles and Optimus robot fleets (The Flip) Intel & AMD CPU Shortage Causes Stock Surges, Along with Dell & HPE Gains: A global shortage of Intel and AMD CPUs sent both stocks surging on March 25 (Bulls and Bears) Qualcomm Is Downgraded From Outperform to Market Perform: Bernstein downgraded Qualcomm on March 26 from Outperform to Market Perform, cutting the price target from $175 to $140 with the pointed note that "investors can buy actual AI winners instead" (Bulls and Bears) NVIDIA Remains Rangebound Despite Strong Demand Signals: Even with continued demand for GPUs, market performance reflects uncertainty around valuation, supply, and future growth expectations (Bulls and Bears) For a deeper dive into each topic, please click on the provided links. Subscribe to our YouTube Channel so you never miss an episode. Disclaimer: The Six Five Pod is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and reference share prices, but nothing discussed should be taken as investment advice. We are not investment advisors. The Decode Arm Unveils AGI CPU — First-Ever In-House Chip, Co-Developed with Meta https://newsroom.arm.com/news/arm-agi-cpu-launch https://www.cnbc.com/2026/03/24/arm-launches-its-own-cpu-with-meta-as-first-customer.html https://www.reuters.com/business/arm-jumps-new-ai-chip-drive-billions-annual-revenue-2026-03-25/ https://x.com/PatrickMoorhead/status/2036426991650353243 https://x.com/danielnewmanUV/status/2036490290341748793 https://x.com/danielnewmanUV/status/2036544293616361841 https://x.com/PatrickMoorhead/status/2036537560059572518 https://x.com/PatrickMoorhead/status/2036537564304384282 https://x.com/PatrickMoorhead/status/2036537566485463459 OpenAI Kills Sora and Doubles Down on Enterprise https://x.com/PatrickMoorhead/status/2036564346419998767?s=20 https://www.nytimes.com/2026/03/24/technology/openai-shutting-down-sora.html https://www.reuters.com/technology/openai-set-discontinue-sora-video-platform-app-wsj-reports-2026-03-24/ https://www.cnbc.com/2026/03/24/openai-shutters-short-form-video-app-sora-as-company-reels-in-costs.html Tesla, SpaceX, and xAI Introduce Terafab https://www.reuters.com/business/autos-transportation/musk-says-spacex-tesla-build-advanced-chip-factories-austin-2026-03-22/ https://www.datacenterdynamics.com/en/news/elon-musk-announces-terafab-20bn-factory-will-make-chips-for-spacex-orbital-data-centers-and-tesla-vehicles/ AI Data Center Moratorium Act https://apnews.com/article/data-centers-ai-electricity-sanders-aoc-65651bd28c3d911d18eeb46cd54f4c75 https://www.axios.com/2026/03/25/sanders-aoc-data-center-moratorium-bill https://www.wired.com/story/new-bernie-sanders-ai-safety-bill-would-halt-data-center-construction/ https://www.pbs.org/newshour/politics/ocasio-cortez-and-sanders-push-bill-to-impose-ai-data-center-moratorium https://www.sanders.senate.gov/press-releases/news-sanders-ocasio-cortez-announce-ai-data-center-moratorium-act/ https://www.washingtonpost.com/opinions/2026/03/25/bernie-sanders-artificial-intelligence-claude/ RSAC 2026 Signals the Rise of Agentic AI Security https://x.com/PatrickMoorhead/status/2036028385856512294 https://www.crn.com/news/security/2026/10-hot-new-cybersecurity-tools-announced-at-rsac-2026 https://newsroom.cisco.com/c/r/newsroom/en/us/a/y2026/m03/cisco-reimagines-security-for-the-agentic-workforce.html https://www.prnewswire.com/news-releases/35th-annual-rsac-conference-opens-flagship-event-in-san-francisco-on-monday-302721594.html The Flip Can Tesla Actually Build a Semiconductor Fab? Or Is Terafab a $10 Billion Fantasy? https://evwire.com/p/tesla-terafab-is-difficult-but-probably-not-rocket-science-says-elon-musk https://www.cbsnews.com/news/terafab-elon-musk-chips-semiconductors-what-to-know/ https://www.linkedin.com/pulse/i-wrong-musk-chip-design-im-skeptical-manufacturing-patrick-moorhead-8ckve https://x.com/PatrickMoorhead/status/2035729688098656427 https://electrek.co/2026/03/22/tesla-spacex-terafab-chip-factory-ai-desperation/ https://www.foxbusiness.com/technology/musk-says-tesla-spacex-build-advanced-chip-manufacturing-facility Bulls and Bears Intel & AMD CPU Shortage Causes Stock Surges, Along with Dell & HPE Gains https://asia.nikkei.com/business/tech/semiconductors/supply-crunch-in-intel-amd-cpus-deal-fresh-blow-to-pc-and-server-makers https://www.investopedia.com/amd-and-intel-are-leading-a-chip-stock-rally-wednesday-here-is-why-intc-11934148 https://www.tomshardware.com/pc-components/cpus/pc-makers-face-shortages-of-intel-and-amd-cpus-that-stretch-up-to-six-months-lead-time-for-orders-jumps-from-just-two-weeks-in-the-face-of-ai-demand https://www.benzinga.com/markets/tech/26/03/51458770/chip-shortage-2026-why-cpus-from-intel-and-amd-are-getting-harder-to-find https://x.com/danielnewmanUV/status/2036779062740525380 https://www.fxleaders.com/news/2026/03/25/dell-technologies-hits-all-time-high-above-178-as-ai-server-demand-fuels-record-rally/ http://markets.chroniclejournal.com/chroniclejournal/article/marketminute-2026-3-25-the-ai-factory-architect-dell-technologies-solidifies-dominance-as-hardware-infrastructure-becomes-the-new-gold https://247wallst.com/investing/2026/03/16/record-ai-orders-pushed-dell-to-a-33-4-billion-quarter-and-wall-street-still-has-doubts/ https://www.marketwatch.com/data-news/hewlett-packard-enterprise-co-stock-outperforms-competitors-on-strong-trading-day-1cde62cf-7e48d1fb4647 https://www.marketbeat.com/instant-alerts/hewlett-packard-enterprise-nysehpe-trading-up-92-whats-next-2026-03-25/ https://www.morningstar.com/news/marketwatch/20260325221/super-micro-dell-and-hpe-have-been-red-hot-stocks-this-week-whats-behind-the-big-moves https://www.investing.com/news/analyst-ratings/evercore-isi-raises-hp-enterprise-stock-price-target-on-ai-demand-93CH-4581751 http://markets.chroniclejournal.com/chroniclejournal/article/finterra-2026-3-25-the-networking-transformation-a-deep-dive-into-hewlett-packard-enterprise-hpe-in-2026 Qualcomm Is Downgraded From Outperform to Market Perform https://investing.com/news/stock-market-news/bernstein-downgrades-qualcomm-says-investors-can-buy-actual-ai-winners-4582017 https://intellectia.ai/news/stock/qualcomm-faces-2026-challenges-amid-20-billion-buyback NVIDIA Remains Rangebound Despite Strong Demand Signals https://www.thestreet.com/investing/stocks/goldman-sachs-sends-blunt-message-on-nvidia-stock-after-gtc https://247wallst.com/investing/2026/03/24/nvidias-gtc-developments-were-far-bigger-than-the-market-realizes/  

 Tesla's strategic shift toward a future defined by artificial intelligence and modular vehicle platforms. Internal documents and public teasers suggest the development of a CyberSUV or CyberVan, designed to leverage the Cybertruck's unique stainless steel architecture for larger families and commercial use. This expansion coincides with the launch of the Model Y "Project Juniper" refresh, which introduces refined aesthetics and hardware to the world's top-selling electric vehicle. Beyond consumer cars, Elon Musk is prioritizing the Optimus humanoid robot and specialized autonomous units like the Cybercab and Robovan. To support this massive computational demand, Tesla is constructing a "Terafab" semiconductor facility in Texas to produce proprietary AI chips. These moves reflect a broader transition from traditional automaking to a comprehensive robotics and autonomy ecosystem.

This week on Autonomy Signals, Grayson Brulte and Rob Grant discuss Tesla Optimus delays driven by China's rare earth export controls, the EU's push to slow AI regulation and what it means for autonomous vehicles, and Waymo's potential expansion into Canada.China's Ministry of Industry and Information Technology (MIIT) has classified humanoid robot actuator components as dual-use technology, requiring foreign manufacturers to share technical specifications to obtain export licenses. Tesla relies on Chinese suppliers for the specialized rare earth magnets that give Optimus its 22-degree hand dexterity, and with China controlling 90% of that supply, delays could persist.AUTNMY AI's proprietary AI algorithm, OMEGA, analyzed the impact of a potential export ban, which could increase the price from $46,000 to produce Optimus parts in China to $133,000 if all production moves to America. If this were to happen, it would lead to a delay in Optimus, and this is further compounded by an FTC investigation into whether over 60% Chinese component content disqualifies Tesla's made-in-America branding.Then there is the MIIT's March 2nd humanoid robot standardization directive, which requires Chinese suppliers to prioritize domestic manufacturers such as Unitree and Xiaomi over foreign customers including Tesla, which creates an additional supplier prioritization risk on top of the export control risk.Closing out the show, Grayson and Rob discuss Waymo's potential Canadian expansion, examining lobbying records that show Waymo Co-CEO Tekedra Mawakana met with Toronto council staff to discuss ride-hail, goods delivery, and commercial operating authorizations. OMEGA also discovered lobbying records showing Waymo has been lobbying British Columbia to change the laws to allow L4 autonomous vehicles, pointing to a potential Vancouver expansion.Episode Chapters00:00 AUTNMY AI00:24 Signal 1: Potenial Tesla Optimus Gen 3 Delay23:35 Signal 2: Europe Delays Classifying L4 Autonomous Vehicles as High Risk48:45 Signal 3: Waymo Eyes Canadian Expansion51:29 Closing--------About The Road to AutonomyThe Road to Autonomy is the definitive media brand covering the Autonomy Economy™. Through our podcasts, newsletter, and proprietary market intelligence, we set the narrative for institutional investors, industry executives, and policymakers navigating the convergence of automation, autonomy, and economic growth.Join institutional investors and industry leaders who read This Week in The Autonomy Economy every Sunday. Each edition delivers exclusive insight and commentary on the autonomy economy, helping you stay ahead of what's next.Subscribe today for free: https://www.roadtoautonomy.com/ae/See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Play NowEpisode 397 of the Seibertron.com Twincast / Podcast begins with the cast providing impressions and their preordering decisions for recent listings including the Collaborative FIFA Optimus Prime, an awkwardly timed patriotic Star Eagle, and the collector-centric Monstructor boxed set. Prompted by the dramatic story events within, a lengthy discussion occurs about Skybound's 30th issue of their Transformers comic book series, leading to criticism about its execution and pacing in the midst of praise for Robert Kirkman's willingness to try something new. This leads to a listener question about the brand's pivot into focusing on its legacy, with the cast providing potentially unpopular suggestions for how to re-boot the franchise in radical new directions. Listener questions about Overgear and triple-changers round the topics out before the episode concludes with the recurring "Bragging Rights" segment.

In this episode of The Brainstorm, Brett, Nick, and Sam are joined by Frank Downing to discuss how XAI is shifting gears, aiming to turn that raw power into practical, revenue-generating products. Elon Musk's latest moves reveal a daring strategy: leveraging Tesla's custom chips, SpaceX's orbital infrastructure, and a close-knit integration of AI and hardware. The question is, can XAI reinvent itself with this vertical integration, or will it become just another seller of compute?If you know ARK, then you probably know about our long-term research projections, like estimating where we will be 5-10 years from now! But just because we are long-term investors, doesn't mean we don't have strong views and opinions on breaking news. In fact, we discuss and debate this every day. So now we're sharing some of these internal discussions with you in our new video series, “The Brainstorm”, a co-production from ARK and Wolf.financial, and sponsored by Public. Tune in every week as we react to the latest in innovation. Here and there we'll be joined by special guests, but ultimately this is our chance to join the conversation and share ARK's quick takes on what's going on in tech today.Key Points From This Episode:The key to competitive advantage now lies in how models are packaged and integrated for real-world application rather than just their raw capabilities.Success depends on seamlessly combining technological capabilities with the right distribution channels and user interfaces.Dominance in AI is increasingly driven by proprietary hardware and custom chips that reduce reliance on expensive, profit-margined cloud providers.To learn more about WOLF: https://wolf.financialTo learn more about Public: https://public.com/

Recorded live at https://www.abundance360.com/ Peter H. Diamandis, MD, is the Founder of XPRIZE, Singularity University, ZeroG, and A360 Elon Musk is the cofounder and CEO of Tesla, cofounder of SpaceX, and xAI. – My companies: Apply to Dave's and my new fund: https://qr.diamandis.com/linkventures... Go to Blitzy to book a free demo and start building today: https://qr.diamandis.com/blitzy _ Read the Solve Everything Paper Get access to metatrends 10+ years before anyone else:https://qr.diamandis.com/metatrends Connect with Peter: X Instagram Listen to MOONSHOTS: Apple YouTube – *Recorded live on March 11th, 2026 *The views expressed by me and all guests are personal opinions and do not constitute Financial, Medical, or Legal advice. Learn more about your ad choices. Visit megaphone.fm/adchoices

In the Electrek Podcast, we discuss the most popular news in the world of sustainable transport and energy. In this week's episode, we discuss Rivian R2, Lucid's counterattack, and Tesla's 'Digital Optimus'. The show is live every Friday at 4 p.m. ET on Electrek's YouTube channel. As a reminder, we'll have an accompanying post, like this one, on the site with an embedded link to the live stream. Head to the YouTube channel to get your questions and comments in. After the show ends at around 5 p.m. ET, the video will be archived on YouTube and the audio on all your favorite podcast apps: Apple Podcasts Spotify Overcast Pocket Casts Castro RSS We now have a Patreon if you want to help us avoid more ads and invest more in our content. We have some awesome gifts for our Patreons and more coming. Here are a few of the articles that we will discuss during the podcast: Musk confirms xAI-Tesla joint ‘Digital Optimus' project — after saying Tesla didn't need xAI Rivian reveals full R2 lineup and pricing, starting at $57,990 with a $45K RWD model coming later Rivian is phasing out the R1S Dual Standard, its most affordable SUV, ahead of the R2 Lucid (LCID) reveals Cosmos and Earth SUVs as first midsize EVs, starting under $50,000 Lucid takes aim at the Tesla Cybercab with Lunar, a two-seat EV robotaxi concept BMW's flagship i7 EV is getting a Neue Klasse upgrade: Here's our first look BYD is open to building cars in Canada and acquiring a rival automaker BYD's luxury EV with 5 min fast charging and nearly 500 miles of range is headed overseas Honda is scrapping three of its most important EVs for the US, including the Acura RSX Aptera (SEV) raises $6.3M through warrant exercise to fund solar electric vehicle validation Here's the live stream for today's episode starting at 4:00 p.m. ET (or the video after 5 p.m. ET: https://www.youtube.com/live/aLUDy28zGmw

This week on Another Pass, Sam and Case are joined by Nic Woolfe to roll out and revisit Transformers: The Movie (1986)! We dig into the film's bold tonal shift, unforgettable soundtrack, and the shocking moments that left an entire generation of kids staring at the screen in disbelief. Does this animated cult classic still have the touch… or does it dare to be stupid? Another Pass Full Episode Originally aired: March 13, 2026  Music by Vin Macri and Matt Brogan Podcast Edited by Sophia Ricciardi Certain Point Of View is a podcast network brining you all sorts of nerdy goodness! From Star Wars role playing, to Disney day dreaming, to video game love, we've got the show for you!  Learn more on our website: https://www.certainpov.com Support us on Patreon! patreon.com/CertainPOVMedia  Join us on Discord: https://discord.gg/wcHHer4 PODCAST SHOWS: ▶ Another Pass - https://www.certainpov.com/another-pass-podcast Notes Transformers Movie Overview and Legacy The 1986 Transformers movie remains a cult classic largely due to its bold character deaths and iconic soundtrack despite its flaws (00:00). Cult Classic Status from Bold Choices (00:39) The film's decision to kill most of the original cast early was a major risk that shaped its lasting appeal. This choice was tied to marketing a new toy line, forcing the removal of older characters to introduce new ones. The high character death count was unusual for 1980s cartoons but resonated as a memorable and impactful narrative choice. Sam Alicea emphasized the movie's unique "music video" style and violent tone as key to its charm. Soundtrack as a Defining Feature (00:29) The film features a heavy metal soundtrack with songs like "The Touch" and Weird Al Yankovic's contribution, which remain popular. Sam highlighted the soundtrack's role in making the movie enjoyable even when used as background during chores. The soundtrack's placement sometimes felt forced but added to the overall energetic vibe of the film. The music helped cement the movie's nostalgic value and cultural impact beyond just the story. Animation Quality and Style (00:32) The animation is a mix of highly detailed hand-drawn sequences and inconsistent lower-budget scenes. Key moments like Unicron's introduction and the transformation of Galvatron show impressive craftsmanship. Some sequences, such as Autobot City's transformation, defy logic but maintain visual excitement. The movie's aesthetic reflects typical 1980s animation with a blend of impressive and sloppy elements. Character and Design Highlights (00:14, 00:33) The introduction of new characters like Hot Rod and Cup served marketing goals but received mixed reception; some preferred legacy characters instead. The Dinobots, especially Grimlock, were fan favorites noted for their personality and screen presence. Decepticons like Galvatron and Starscream stood out, with Starscream's treacherous antics praised. New designs leaned toward smoother, rounded shapes compared to the original blockier forms, signaling the toy line shift. Strategic Marketing and Toy Line Impact The movie functioned primarily as a vehicle to launch a new toy line, influencing story and character decisions deeply (00:12, 00:16). Toy Line Rollout Drives Plot and Character Fate (00:12) Killing off legacy characters cleared the stage for a new roster designed to sell fresh toys. Characters like Cup were introduced primarily to support new toy sales rather than story needs, frustrating some fans. Hot Rod's arc was designed to establish a fresh hero while maintaining toyetic appeal. The shift to new designs reflected a strategic move to modernize the brand and stimulate consumer interest. Balancing Narrative and Marketing Needs (00:16) The movie's story and character choices sometimes sacrificed cohesion for toy marketing goals. Legacy characters like Perceptor and the Dinobots were retained to maintain continuity and fan connection. Some characters, like Ultra Magnus, were less favored due to lack of narrative effectiveness but existed for toy continuity. The film's pacing and tone reflected the tension between storytelling and commercial objectives. Voice Cast and Star Power Usage (00:24) Leonard Nimoy's casting as Galvatron was a notable stunt but was not sustained beyond the movie. Orson Welles' last role as Unicron added gravitas despite limited credit and technical challenges with his audio. Voice changes and stunt casting decisions were influenced by budget and potential marketing impact. The cast choices reflect a blend of marketing intent and creative ambition. Narrative and Character Development Critiques The movie's storytelling assumes audience familiarity and presents mixed character arcs, impacting accessibility and engagement (00:38, 00:59). High Barrier for New Viewers (00:38) Nick Wolfe identified the movie as not beginner-friendly, expecting viewers to know extensive Transformers lore. The lack of exposition around key characters and events made it confusing for newcomers. Proposed solutions included adding narrated backstory and flashbacks to ease new viewers into the plot. This gap likely contributed to the movie's commercial failure despite strong fan following. Character Arcs and Roles (00:16, 00:59) Hot Rod's character is a flawed hero whose leadership rise felt unearned, creating mixed audience reception. Cup's role was criticized for being a new character inserted mainly for toy reasons rather than story depth. The pitch suggested replacing Cup with a legacy character like Ironhide to strengthen narrative bonds. Optimus Prime's death remains a pivotal emotional moment but complicates continuity and future storytelling. Supporting Characters and Dynamics (01:06, 01:09) Grimlock's interactions with Hot Rod added needed conflict and character development. Supporting characters like RC and Springer were noted as effective background players without overwhelming the story. Starscream's consistent treachery was highlighted as a strong character trait enhancing the villain dynamic. There was consensus that some Decepticon characters disappeared too quickly, weakening the villain ensemble. Proposed Improvements and Alternate Pitch Nick Wolfe's detailed pitch aimed to make the movie more accessible, coherent, and emotionally resonant without losing core elements (00:39, 00:42). Introductory Narration and Flashbacks (00:42) Suggested opening with Optimus Prime narrating key backstory events to orient new viewers. Including flashbacks to important episodes would provide context for the war, characters, and stakes. This would bridge the gap between fans and newcomers, enhancing story clarity. It also sets up the importance of Energon and Autobot City more clearly. Expanded Character Development and Role Reassignments (00:43, 00:51) Proposed making Cup more like Ironhide or Tanker (an original draft character) to improve narrative depth. Hot Rod would be shown as a capable fighter before his failures, building audience empathy. Grimlock and Hot Rod's relationship would be deepened with conflict and eventual teamwork. Optimus Prime would be critically wounded but kept alive on life support, preserving his legacy. Streamlined Plot and Group Dynamics (00:50) Consolidated the Junkion and Quintesson arcs into a single planet scenario for simplicity. Divided Autobots into clear factions with distinct challenges to create focused narrative threads. Added heroic sacrifices and teamwork moments to raise stakes and emotional impact. The pitch ended with Optimus evolving into Ultra Magnus, aligning with toy line needs while preserving character continuity. Maintaining Cult Appeal While Improving Accessibility (01:03) The pitch carefully retained key emotional beats like Optimus Prime's near-death and Hot Rod's rise. It balanced new viewer guidance with fan service to preserve the movie's iconic moments. Suggested minor tweaks to Unicron's scale and lore to reduce confusion without major changes. The approach aimed to prevent later franchise regrets about character handling seen in season three. Fan and Host Perspectives on Movie's Legacy The hosts and guest expressed strong affection for the movie's nostalgic and cultural value despite its flaws (00:59, 01:00). Embracing the Movie's 1980s Roots (01:00) Sam Alicea stressed the film's authentic 80s vibe, embracing both its high-quality animation and its rough edges. The movie was seen as a time capsule of 80s animation and storytelling norms, including its willingness to embrace trauma. The soundtrack's energy and the movie's unapologetic style were key to its enduring love. There was reluctance to change the movie too much, preserving its unique charm. Appreciation of Character Moments and Humor (01:09) Starscream was praised for his consistent jerkiness, providing comic relief and memorable villainy. The Decepticons' internal conflicts were contrasted with the Autobots' camaraderie, enriching character dynamics. The hosts lamented the disappearance of classic Decepticons when newer ones appeared. The blend of action, humor, and character quirks contributed to the movie's lasting appeal. Community and Ongoing Engagement (01:12, 01:16) Nick Wolfe and hosts highlighted fan communities like the Certain Point of View Media Discord for ongoing discussions. References to other Transformers properties like Beast Wars show layered fan engagement across generations. The show's Patreon and related projects encourage deeper fan interaction and content creation. The continued interest in the movie reflects its significance beyond initial box office performance. Distribution and Community Outreach The podcast promotes broader engagement through Patreon, additional shows, and social media presence (01:16). Patreon Support and Exclusive Content (01:16) The show thanks executive producer-level patrons by name, recognizing their financial support. Patreon offers early episode clips, essays on geek culture, and D&D topics to supporters. Essays like "Never Go Full Ranger" provide added value and deepen listener engagement. This support sustains the podcast's production and community activities. New Shows and Guest Hosting Opportunities (01:17) The launch of "Trade School," a comic book guest-hosted show, expands the network's content diversity. The format encourages fans to share positive takes on trade paperbacks in brief episodes. This initiative fosters community involvement and fresh perspectives. The network invites submissions, broadening participation from listeners. Social Media and Contact Channels (01:13, 01:16) Hosts provide social media handles and highlight the Discord server as a key interaction hub. Nick Wolfe shares his Reddit and Discord activity, emphasizing low-pressure involvement. The Discord serves as a central place for fan discussion and host engagement. Listeners are encouraged to tag hosts for responses and participate in the fan community. Upcoming Episode and Network Branding (01:18) The next podcast episode will cover "Highlander 2: The Quickening," maintaining a focus on cult and flawed films. The show's production credits and branding reinforce a professional and creative identity. The network's website and YouTube presence offer additional access points. This continuity supports sustained audience growth and brand recognition.    

In this episode, the mates, along with guest Ben Horowitz, explore Elon Musk's shift to lunar AI data centers, mass drivers, O'Neill cylinders, Dyson swarms, and Optimus robots pioneering space. Get notified once we go live during Abundance360: https://www.abundance360.com/livestream  Get access to metatrends 10+ years before anyone else - https://qr.diamandis.com/metatrends   Peter H. Diamandis, MD, is the Founder of XPRIZE, Singularity University, ZeroG, and A360 Ben Horowitz is a cofounder and general partner at Andreessen Horowitz (a16z), NY Times bestseller author, and creator of the a16z Cultural Leadership Fund. Salim Ismail is the founder of OpenExO Dave Blundin is the founder & GP of Link Ventures Dr. Alexander Wissner-Gross is a computer scientist and founder of Reified – My companies: Apply to Dave's and my new fund:https://qr.diamandis.com/linkventureslanding   Go to Blitzy to book a free demo and start building today: https://qr.diamandis.com/blitzy   _ Connect with Peter: X Instagram Connect with Ben X Instagram Linkedin Learn about a16z Connect with Dave: X LinkedIn Connect with Salim: X Join Salim's Workshop to build your ExO  Connect with Alex Website LinkedIn X Email Substack  Spotify Threads Listen to MOONSHOTS: Apple YouTube – *Recorded on February 13th, 2026 *The views expressed by me and all guests are personal opinions and do not constitute Financial, Medical, or Legal advice. Learn more about your ad choices. Visit megaphone.fm/adchoices

What happens when AI bots get their own social network, Silicon Valley execs cozy up to power, and Apple takes a cut from creators? This week's panel calls out the bold, bizarre, and often problematic ways tech's biggest players are reshaping everything from AI assistants to your everyday privacy. There's a social network for AI agents, and it's getting weird Moltbook is the most interesting place on the internet right now Exposed Moltbook Database Let Anyone Take Control of Any AI Agent on the Site Pentagon clashes with Anthropic over military AI use, sources say Salesforce signs $5.6B deal to inject agentic AI into the US Army Angry Norfolk residents lose lawsuit to stop Flock license plate scanners SpaceX wants to put 1 million solar-powered data centers into orbit Elon Musk reportedly wants a June SpaceX IPO to align with his birthday, the planets Tesla hits a grim milestone: its second straight year of decline Tesla says production-ready Optimus robot is coming soon Microsoft reports strong cloud earnings in Q2 as gaming declines What We Learned From Meta, Microsoft and Tesla Apple tells Patreon to move creators to in-app purchase for subscriptions by November Apple CEO Tim Cook 'heartbroken' after repeated ICE killings in Minneapolis A rival smart glasses company is suing Meta over its Ray-Ban products TikTok, YouTube, and Meta are headed to court for a landmark trial over social media addiction The 'Social Media Addiction' Narrative May Be More Harmful Than Social Media Itself TikTok users freak out over app's 'immigration status' collection — here's what it means A Waymo hit a child near an elementary school in Santa Monica Autonomous cars, drones cheerfully obey prompt injection by road sign Samsung's TriFold phone will cost $2,899 in the US Groundhogs are bad at predicting weather, but they're valuable animal engineers Satellites encased in wood are in the works Belkin reminds users that its Wemo smart home products are shutting down this week Host: Leo Laporte Guests: Gary Rivlin, Devindra Hardawar, and Victoria Song Download or subscribe to This Week in Tech at https://twit.tv/shows/this-week-in-tech Join Club TWiT for Ad-Free Podcasts! Support what you love and get ad-free audio and video feeds, a members-only Discord, and exclusive content. Join today: https://twit.tv/clubtwit Sponsors: monarch.com with code TWIT Melissa.com/twit helixsleep.com/twit canary.tools/twit - use code: TWIT expressvpn.com/twit

We've been covering what's happening in Minnesota, and the killing of Alex Pretti, all week on The Verge. To begin this episode, Nilay explains why — and why so many others seem to feel the same way right now. After that, the hosts talk about the CEO-studded screening of Melania Trump's documentary last weekend, the disastrous public appearance from Tim Cook, and whether Cook and other CEOs have any other option but to capitulate to the Trump administration. Then it's time for some gadgets: we talk about the super-foldy, super-expensive Samsung Galaxy  Z Trifold, the Clawdbot / Moltbot phenomenon, and whether Google can finally put Chrome OS and Android together the right way. Finally, in the lightning round, it's time for Brendan Carr is a dummy, Tesla's anti-car pivot, Apple's design hires, and more. Further reading: On the ground in Minneapolis after the killing of Alex Pretti  I grew up with Alex Pretti  Creators and communities everywhere take a stand against ICE  It doesn't matter if Alex Pretti had a gun  Why won't anyone stop ICE from masking?  Tim Cook, Andy Jassy, and AMD CEO Lisa Su are at the White House for a VIP screening of the Melania doc. Tim Cook had ‘a good conversation' with Trump about deescalation  Cook in 2020: Speaking up on racism From The New York Times: Amazon's $35 Million ‘Melania' Promotion Has Critics Questioning Its Motives From The Hollywood Reporter: ‘Melania' Set for a $3 Million Opening Despite Amazon's $35 Million Marketing Push Here's Tim Cook hanging out with accused rapist Brett Ratner at the Melania screening What TikTok's new owners mean for your feed  TikTok USA is broken  TikTok is still down, here are all the latest updates  TikTok is still struggling in the US due to a “cascading systems failure.”  TikTok US is mostly back up and running  TikTok blames its US problems on a power outage  Oracle admits it broke TikTok. Congress doesn't seem to know if the TikTok deal complies with its law  Is New TikTok banning the word “Epstein” in DMs? Not really.  TikTokers are heading to UpScrolled following US takeover  Mark Zuckerberg is all in on AI as the new social media  Meta is stopping teens from chatting with its AI characters  Bluesky is testing ‘live' features to take on X  Best gas masks The Samsung Trifold will cost nearly three grand  Google just leaked a first look at Android for PC in action  Chromebooks train schoolkids to be loyal customers, internal Google document suggests  Moltbot, the AI agent that ‘actually does things,' is tech's new obsession Clawdbot's bad day  I used Claude to vibe-code my wildly overcomplicated smart home The FCC's Late Night Comedy Show Tesla discontinuing Model S and Model X to make room for robots  Tesla says production-ready Optimus robot is coming soon  Tesla hits a grim milestone: its second straight year of decline Elon Musk invests $2 billion in Elon Musk Hang on, there's a Trump Phone Ultra coming too?  Halide co-founder Sebastiaan de With is joining Apple's design team  The Stream Deck-packed gaming keyboard is a monster of good ideas Subscribe to The Verge for unlimited access to theverge.com, subscriber-exclusive newsletters, and our ad-free podcast feed.We love hearing from you! Email your questions and thoughts to vergecast@theverge.com or call us at 866-VERGE11. Learn more about your ad choices. Visit podcastchoices.com/adchoices