Podcasts about cs go

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,

Moses returns to Snake & Banter predicting Mouz's roster overhaul is already a disaster, with xertioN as IGL a cautionary tale waiting to happen.   Into The AM: Premium apparel built for long flights and tournament weekends. Get an additional 10% off on top of sale prices at https://intotheam.com/lfn Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

karrigan's HLTV interview has Falcons fans asking questions before the IEM Cologne Major even begins. Thorin, Mauisnake, and STYKO break it all down in Counter-Strike.   Raycon — Upgrade your listen with Raycon's Open Earbuds or Everyday Classics. Go to https://buyraycon.com/snakeopen to get 15% off your order.   Polymarket — Livetrade on CS2 today on Polymarket: https://polymarket.com/?via=lastfreenation-eeux Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Bienvenue dans ce 98e épisode du BlaBla Royal, votre podcast dédié à l'univers des jeux vidéo, du gaming et de l'eSport!Cette semaine, Lulune, Dan et Marc reçoivent une invitée d'exception : Stéphanie Harvey, aussi connue sous le nom de MissHarvey, l'une des figures les plus importantes du gaming au Québec et dans le monde.Multiple championne de Counter-Strike: Global Offensive (CS:GO) et grande gagnante de Big Brother Célébrités 2022, elle partage avec nous son parcours impressionnant dans l'univers du eSport professionnel.Au programme de cet épisode :

On explore comment la triche s'est installée jusque sur la scène e‑sport, avec un focus sur CS:GO et les cas de joueurs pris en flagrant délit en direct. Matthieu retrace l'affaire depuis 2014 et explique les méthodes employées, des logiciels d'aide à la visée aux parades pour contourner la détection. On analyse aussi la riposte de Valve (VAC, contrôle en tournois) et ce que ces scandales changent pour l'e‑sport pro.En plateau Michaël de Marliave — animateur Matthieu Lambda — chroniqueur Tiffany Souterre — chroniqueuse➤ Pour découvrir Mammouth IA : https://mammouth.ai/➤ Pour le Merch Micode et Underscore_ : https://traphic.fr/collections/micode⚠️ Précommandes avant le 15 Janvier ! Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.

Linktree: ⁠https://linktr.ee/Analytic⁠Join The Normandy For Additional Bonus Audio And Visual Content For All Things Nme+! Join Here: ⁠https://ow.ly/msoH50WCu0K⁠Dive into the Notorious Mass Effect podcast with host Analytic Dreamz as we dissect Counter-Strike 2 (CS2) – Valve's free-to-play tactical FPS masterpiece that replaced CS:GO in 2023. Built on the Source 2 engine with sub-tick architecture for precise hit registration and revolutionary volumetric smokes, CS2 delivers unmatched gunplay in MR12 competitive modes, Premier rankings, and a thriving economy centered on weapons, utilities, and over 1,000 skins.Analytic Dreamz explores the game's massive player base exceeding 1.8 million concurrent peaks in 2025, billions in revenue from cases and Prime upgrades, persistent cheating controversies, and the seismic October 2025 update introducing Trade-Up Contracts for knives and gloves – crashing the skin market by billions overnight and reshaping virtual asset values.From pro losses to improved accessibility, this segment analyzes CS2's performance, economy shock, and future implications. Support this podcast at — https://redcircle.com/analytic-dreamz-notorious-mass-effect/donationsPrivacy & Opt-Out: https://redcircle.com/privacy

Down to Dorf - 5 Jahre Jubiläum! Gaming Special & Burger-Test

Watch The X22 Report On Video No videos found (function(w,d,s,i){w.ldAdInit=w.ldAdInit||[];w.ldAdInit.push({slot:17532056201798502,size:[0, 0],id:"ld-9437-3289"});if(!d.getElementById(i)){var j=d.createElement(s),p=d.getElementsByTagName(s)[0];j.async=true;j.src="https://cdn2.decide.dev/_js/ajs.js";j.id=i;p.parentNode.insertBefore(j,p);}})(window,document,"script","ld-ajs");pt> Click On Picture To See Larger Picture Canada is sinking fast, 1 in 4 Canadians work for the government. UK study is turning out to be true, SNAP receipts don’t want to work. Trump voids Biden’s green new scam.  Job numbers are portraying what is really going on, the deportations of illegals is opening up jobs. Trump is building the narrative to get rid of the Fed. The [DS] knows that Trump and team are coming after them, there is no escape. The FBI arrested the J6 pipe bomber, they have known this entire time. Kash and team are building respect showing American they will arrest the true criminals. The criminal syndicate has poisoned America, Trump is in the process of curing it with the people. Economy https://twitter.com/WallStreetMav/status/1996577094294266268?s=20  who vote for every dollar they can squeeze out of the people who actually built Canada. (function(w,d,s,i){w.ldAdInit=w.ldAdInit||[];w.ldAdInit.push({slot:18510697282300316,size:[0, 0],id:"ld-8599-9832"});if(!d.getElementById(i)){var j=d.createElement(s),p=d.getElementsByTagName(s)[0];j.async=true;j.src="https://cdn2.decide.dev/_js/ajs.js";j.id=i;p.parentNode.insertBefore(j,p);}})(window,document,"script","ld-ajs"); https://twitter.com/EndWokeness/status/1996350793721717150?s=20 Watch: Trump Signing Car Industry EO, Which Seeks to End ‘Burdensome’ Green Energy Regulations  In an event in the Oval Office, President Trump, along with representatives of car manufacturers and dealerships, announced a new executive order rescinding Biden CAFE tailpipe emissions standards, which raised costs on the companies and the cost of car prices for consumers. CAFE stands for “Corporate Average Fuel Economy.” An easy definition, via the National Highway Traffic Safety Administration, is a regulation: …[on] how far our vehicles must travel on a gallon of fuel. NHTSA sets CAFE standards for passenger cars and for light trucks (collectively, light-duty vehicles), and separately sets fuel consumption standards for medium- and heavy-duty trucks and engines. NHTSA also regulates the fuel-economy window stickers on new vehicles. Ending these regulations is part of complying with Trump’s “Unleashing American Energy” Executive Order and the [Transportation] Secretary's “Fixing the CAFE Program” Memorandum, according to the department’s homepage. On Wednesday, the president rattled off the investments in the United States that the auto companies have made since Trump 47 began in January. In one case, he cracked up the room by joking that they could do better than the billions of dollars they are promising: Trump added that it’s part of getting rid of more of the “Green New Scam,” too. https://twitter.com/townhallcom/status/1996322694099345577?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1996322694099345577%7Ctwgr%5E31d2137f89d95733dae2dffe91ce3f002937e448%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fredstate.com%2Fbeccalower%2F2025%2F12%2F03%2Fwatch-trump-signing-car-industry-eo-ending-burdensome-biden-regulations-n2196786 https://twitter.com/townhallcom/status/1996325044260085885?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1996325044260085885%7Ctwgr%5E31d2137f89d95733dae2dffe91ce3f002937e448%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fredstate.com%2Fbeccalower%2F2025%2F12%2F03%2Fwatch-trump-signing-car-industry-eo-ending-burdensome-biden-regulations-n2196786 Source: redstate.com https://twitter.com/WallStreetMav/status/1996587083012378947?s=20 https://twitter.com/profstonge/status/1996583802559136057?s=20 https://twitter.com/unusual_whales/status/1996582009460732232?s=20   to a seasonally adjusted 191,000, the lowest level since September 2022, with expectations of 220,000. https://twitter.com/WallStreetMav/status/1996431168883900552?s=20   fourth consecutive month-over-month decline. Apartment rents are down 1.1% from November 2024 and have fallen 5.2% from their 2022 peak. https://twitter.com/JDVance/status/1996413457164566966?s=20 https://twitter.com/KobeissiLetter/status/1996635885232693482?s=20   will provide TRILLIONS of dollars to kids if invested until age 20+. https://twitter.com/MJTruthUltra/status/1996602968750047572?s=20   ELIMINATE the INCOME TAX with tariffs, just like President William McKinley did in 1897 — and President Trump is following in his exact footsteps. The Dingley Tariff Act of 1897, which was the centerpiece of President William McKinley’s domestic economic policy. He believed in protectionism (using high tariffs to protect American industries and workers from foreign competition). Just like President Trump. Between 1897–1901 the United States had just become the wealthiest and highest-output economy on EARTH, surpassing the United Kingdom as the world's #1 manufacturing nation and #1 economy in total GDP By 1900 U.S. industrial production was roughly equal to that of Britain, Germany, and France COMBINED. Taxation without representation is unconstitutional and ILLEGAL. We, The People are now finally coming to grips with the illusions that have been installed around us our entire lives. Who is the FED? What is the FED? How did the FED come to fruition? What is the Titanic? Who was on the Titanic? What year? Those illusions are now dissolving. The “Golden Age of America” truly is among us. There is a plan to restore our country. Gods plan is playing out right now. Political/Rights https://twitter.com/libsoftiktok/status/1995921829316149445?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1995921829316149445%7Ctwgr%5Ede940a2e201b3b2cbe0a9536949880a88f90b3b2%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fwww.thegatewaypundit.com%2F2025%2F12%2Fillegal-immigrant-rapist-who-walks-free-sweetheart-plea%2F https://twitter.com/DHSgov/status/1996301518543159560?s=20  California will let him roam free even though he's been arrested for FELONY hit-and-run. Hector Balderas-Aheelor has been previously deported FOUR times and committed a felony when he illegally entered for a fifth time. This violent criminal must be deported ASAP. Assistant U.S. Attorney Blasts Los Angeles County After It Votes to Ban ICE Agents from Wearing Masks Los Angeles County's far-left Board of Supervisors voted Tuesday to advance an unprecedented ordinance that would ban law enforcement officers, including federal immigration agents, from wearing masks while working in unincorporated areas of the county. The proposal passed 4-0, with only Supervisor Kathryn Barger abstaining. A final vote is scheduled for next week, and the ordinance would take effect in January 2026, according to the LA Times. Legal experts say federal immigration agents would not be required to follow a county mask ban. The county's top lawyer, Dawyn Harrison, has said she suspects the federal government will likely argue that the county law violates the Constitution, which states that federal law takes precedence over conflicting local statutes.  https://twitter.com/USAttyEssayli/status/1996040656733814854?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1996040656733814854%7Ctwgr%5E71ecedc2cdf5f5bfec40c6e10a4a461e8ffa25fd%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fwww.thegatewaypundit.com%2F2025%2F12%2Fassistant-u-s-attorney-blasts-los-angeles-county%2F   by the media and local politicians. We will not expose our brave men and women to personal attacks by allowing agitators to dox them and their families through facial recognition tools. Source: thegatewaypundit.com WA Democrat Rep Wants New Law to Tie ICE’s Hands, Force Americans to Pay Illegal Aliens’ Legal Bills  https://twitter.com/RepJayapal/status/1996294756876325063?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1996294756876325063%7Ctwgr%5E44a6de522e274487c6684ae078f8a1aa15daf059%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fredstate.com%2Fkatie-jerkovich%2F2025%2F12%2F03%2Fwashington-rep-wants-to-make-americans-pay-for-illegal-aliens-defense-n2196788 https://twitter.com/RedWave_Press/status/1996317691536138470?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1996317691536138470%7Ctwgr%5E44a6de522e274487c6684ae078f8a1aa15daf059%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fredstate.com%2Fkatie-jerkovich%2F2025%2F12%2F03%2Fwashington-rep-wants-to-make-americans-pay-for-illegal-aliens-defense-n2196788   dignity, justice, oversight, and accountability to the detention system by repealing mandatory detention, prohibiting the detention of families and children in family detention, phasing out the use of private detention facilities and jails, and requiring DHS to establish civil detention standards.” “The bill creates a presumption of release and imposes a higher burden of proof to detain primary caregivers and vulnerable populations. The bill also mandates the DHS Inspector General to conduct unannounced inspections and requires DHS to admit Members of Congress to detention facilities for unannounced inspections.” What a joke! Source: redstate.com https://twitter.com/bx_on_x/status/1996037478914892112?s=20   “Moist Nigerian”, 26, of Albuquerque, New Mexico Rumaldo Valdez aka “Duck”, 22, of Honolulu, Hawaii David Brilhante aka “CS:GO”, 28, of San Diego, California Camden Rodriguez aka “oHare”, 22, of Longmont, Colorado DOGE https://twitter.com/TheSCIF/status/1996302141309296710?s=20  to the Trump administration. They were taken to court and ordered by a judge to hand over the files and they still refused. They encrypted all the USAID files and internal communications because it would expose the paper trail, the players, and everything happening today. Every single NGO, shell company, and person, including government workers, involved in the active color revolution that’s taking place right now on U.S. soil and the Trump administration and U.S. taxpayers are paying for all of it. To this day the files and internal communications are still encrypted. Tons of other evidence was shredded and destroyed. Geopolitical https://twitter.com/disclosetv/status/1996534889282408841?s=20 https://twitter.com/amuse/status/1996347018621354148?s=20   key figure in the Cartel de los Soles, tells President Trump that Venezuela has weaponized cocaine, exported criminal gangs like Tren de Aragua into the U.S., and allowed FARC, ELN, Hezbollah and Cuban intelligence to operate freely. He alleges decades of espionage inside U.S. installations, Russian tapping proposals, and Smartmatic election manipulation tools exported abroad. Carvajal says the Biden–Harris border collapse allowed Venezuelan operatives to enter the U.S. and asserts Trump's hard-line policies were not only correct, but necessary for American national security. https://twitter.com/TheSCIF/status/1996472423751774516?s=20   This is nothing new to anyone paying attention. Now, it’s just verified by an inside source. Do you really believe everything happening in and around Venezuela is just about drug trafficking and a countries resources like oil? Remember, there are layers to everything. If we do not solidify and fix our elections, there is no point in anything we do because we will lose anyway. Now you understand there is more to this than meets the eye. “My name is Hugo Carvajal Barrios. For many years, I was a high-ranking member of the Venezuelan regime… …Smartmatic was born as an electoral tool of the Venezuelan regime…I know this because I placed the head of IT of the National Electoral Council (CNE) in his position, and he reported directly to me. The Smartmatic system can be altered-this is a fact. This technology was later exported abroad, including to the United States. Regime operatives maintain relationships with election officials and voting-machine companies inside your country…” Please read the documents below in full. https://twitter.com/disclosetv/status/1996243190051967395?s=20 War/Peace https://twitter.com/amuse/status/1996331133437677821?s=20  March 2025 sharing of Yemen strike information over Signal violated departmental rules but was not illegal due to his declassification authority. Sources now say Sen. Mark Kelly disclosed classified elements of that IG report to the Wall Street Journal while Democratic officials were publicly criticizing Hegseth. The reported leak raises questions about the treatment of sensitive oversight documents and the boundary between political messaging and classified material. https://twitter.com/SeanParnellUSA/status/1996361901870313541?s=20 https://twitter.com/PeteHegseth/status/1996368824397094925?s=20 https://twitter.com/sentdefender/status/1996455529644790133?s=20   declaration of war or specific authorization for use of military force,” in regard to potential hostilities – without congressional approval – against Venezuela. https://twitter.com/MarioNawfal/status/1996537466086461599?s=20   Barack Obama built. The “war on narco-terrorists” that Secretary Hegseth is bragging about is being run on a legal foundation poured years before Trump ever touched the Oval Office desk. Obama normalized the presidential kill list. He campaigned in 2008 promising rule of law, transparency, and restraint. He delivered a tenfold increase in drone strikes; a White House “Terror Tuesday” meeting where officials literally flipped through PowerPoint slides picking who lived or died; and a legal doctrine allowing the president to kill U.S. citizens without trial, notice, or judicial review. That doctrine – once unthinkable – is now standard operating procedure. Obama's Director of National Intelligence openly admitted in 2010 that the administration was targeting Americans based on vague criteria like whether a citizen was “involved” in a group “trying to attack us.” That’s not evidence… that’s vibes. When civil-liberties groups sued to force the government to explain the legal basis, Obama declared the entire matter a state secret. Meaning: the president could now kill you, and no court was allowed to ask why. The judiciary rubber-stamped it. Both parties embraced it. And the public largely applauded it. Then came the Awlaki killings. Obama ordered the drone assassination of: •⁠ ⁠Anwar al-Awlaki, an American cleric •⁠ ⁠Samir Khan, an American citizen standing next to him •⁠ ⁠Awlaki's 16-year-old son, killed two weeks later at a café The White House smeared the kid as a “21-year-old terrorist.” The next week a birth certificate proved he was a Colorado teenager with zero ties to extremism. The administration shrugged. Obama famously told aides: “Turns out I'm really good at killing people.” And Washington – media, political class, and voters – rewarded him. Fast-forward to 2025. Trump isn't inventing anything new in Venezuela. He's using the exact precedents Obama left behind: Unreviewable executive kill authority, expanded definitions of “enemy combatant,” secret memos. No congressional oversight, no geographic limits, zero court supervision. Obama created the kill switch. Trump just slammed it. If you cheered the drone program when your guy was doing it, you already endorsed what’s happening now. If you let the government redefine due process as “whatever we decide in secret,” you already consented to the next president weaponizing that power. And if you normalize extrajudicial force abroad, eventually it comes home. Obama built the architecture. Trump moved in. And Venezuela is learning what happens when a precedent meets a president with fewer brakes  EU corruption scandal could take down von der Leyen – Politico A corruption probe into former EU foreign policy chief Federica Mogherini has thrown European Commission President Ursula von der Leyen's position into jeopardy, with opponents preparing to turn the affair into a fresh push to remove her, Politico reported on Wednesday, citing officials in the bloc. Mogherini, who served as the EU's top diplomat from 2014 to 2019 and is now rector of the College of Europe, was detained on Tuesday. She was formally accused by the European Public Prosecutor's Office of procurement fraud, corruption, conflict of interest, and breaches of professional secrecy over an EU-funded diplomatic academy program. In the wake of the scandal, von der Leyen “is facing the starkest challenge to the EU's accountability in a generation,” with her rivals renewing calls for a new no-confidence vote, Politico reported. source:rt.com NATO Members Commit More Than $1 Billion to Purchase U.S. Weapons for Ukraine NATO officials from Canada, Germany, the Netherlands, Norway, and Poland pledged hundreds of millions more in U.S.-made weapons under the Prioritised Ukraine Requirements List (PURL) scheme, Ukrainian media reported. Amid Russian gains on the battlefield, they insist Ukraine must be armed “to keep the fight going.” The alliance offered these commitments while being accused of attempting to sabotage peace talks. Canada, Germany, Norway, the Netherlands, Poland, and the U.K. all announced new contributions, pushing the total put toward U.S.-made weapons bound for Ukraine toward $5 billion this year alone. These newly allocated funds come at a time when many European governments, increasingly unpopular at home, are cutting domestic programs and warning of budget shortfalls. Source: thegatewaypundit.com Leaked Transcript of EU Conference Call with Zelenskyy Highlights Fear of Trump Securing Peace for Ukraine Ukraine President Volodymyr Zelenskyy was in Paris talking to Emmanuel Macron last weekend while Rubio, Witkoff and Kushner were meeting with Ukraine officials in Florida.  This telephone call is reported to have taken place the following day, on Monday. French President Emmanuel Macron, German Chancellor Friedrich Merz, NATO Secretary-General Mark Rutte, Finnish President Alexander Stubb, Ukrainian President Volodymyr Zelenskyy together with European Commission President Ursula von der Leyen, Polish Prime Minister Donald Tusk, Italian Prime Minister Giorgia Meloni, Danish Prime Minister Mette Frederiksen, Norwegian Prime Minister Jonas Gahr Støre and European Council President António Costa held a conference call. A transcript of the call was leaked to Spiegel, a German news outlet. According to the narrative the assembled group of EU leaders were discussing how the Trump administration was going to betray Ukraine in order to get a peace deal with Russia. The gist of the narrative sounds accurate, though some of the EU leaders are denying the specifics of the wording used. The EU is very worried President Trump may formulate a peace agreement then present the final terms to Ukraine without the EU being involved in the construct of the details. The EU is opposed to any peaceful end to the conflict, because the EU and NATO have positioned their collective economies to only benefit if the military spending continues; they are backstopping their spending with the confiscated Russian assets.   Source: theconservativetreehouse.com   Medical/False Flags https://twitter.com/disclosetv/status/1996567491334803864?s=20 [DS] Agenda https://twitter.com/WallStreetApes/status/1996550058809188740?s=20   “Ilhan and this group, they spent a lot of money” “So, the people that work for Ilhan (0mar) are actually counting the ballots, counting the vote?” “They (Ilhan Omar’s campaign staffers) become a manager, in the precinct too.” Somalians who don't speak English are also told out how to vote “They walk with you to the booth and then they vote. Oh, vote this guy, vote this guy, vote this guy. Vote – even if you speak English.” They are on camera in this video paying people $200-$800 per vote and telling them how to fill out the ballots https://twitter.com/libsoftiktok/status/1996613642113183985?s=20 https://twitter.com/amuse/status/1996600165377945957?s=20  Investigators and policy researchers are pointing to overlapping relationships between Rep. Ilhan Omar and individuals later charged or convicted in Minnesota's billion-dollar pandemic-meal fraud. Omar's 2018 victory party was held at Safari Restaurant, co-owned by Salim Ahmed Said, now found guilty of pocketing more than $12 million in fraudulent reimbursements after reportedly serving “phantom” meals. A member of Omar's campaign staff has also been convicted in the same broader scheme. Critics note Omar publicly praised the program that later enabled the fraud, and that she maintained ties with several of the participants. Prosecutors say Minnesota's COVID-era meal programs operated with almost no verification, allowing widespread abuse through Feeding Our Future and affiliated entities. Democrat Senator Mark Warner Accused of Calling for a MILITARY COUP Against President Trump After Saying the Military Should “Save Us from This President” (VIDEO) Sen. Mark Warner appears on MSNBC's “Morning Joe” as he delivers the controversial remarks suggesting the U.S. military may need to “save us from this President.” Democrat Senator Mark Warner (D-VA) is now facing accusations of openly hinting at a military coup after suggesting on national television that the U.S. armed forces may need to “save us from this President.” https://twitter.com/gentrywgevers/status/1996413726979928245?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1996413726979928245%7Ctwgr%5Ee826188d2a4172a057b36af16e52c8fb9ad3e1a2%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fwww.thegatewaypundit.com%2F2025%2F12%2Fdemocrat-senator-mark-warner-accused-calling-military-coup%2F Source: thegatewaypundit.com President Trump's Plan https://twitter.com/Rightanglenews/status/1996576953248211046?s=20 BREAKING: January 6 Pipe Bomb Suspect Identified  The January 6 Pipe Bomb suspect has been identified. The FBI arrested a man named Brian Cole in connection with the January 6 pipe bombs on Thursday morning. “Brian Cole is the person the FBI has in custody and whom they believe left the pipe bombs in DC on 1/5/2021, according to two senior law enforcement officials briefed on the matter,” NBC News reported. Brian Cole will appear in court on Thursday. Source: thegatewaypundit.com https://twitter.com/TonySeruga/status/1996628056853958759?s=20   Falls Church, Northern Virginia, because he gave the suspect a ride. Home Depot credit card transactions and CCTV video evidence also link the suspect to the purchase of many of the materials used to make the ‘pipe bombs’. The Big Question Why did Steven M. D’Antuono, Assistant Director in Charge of Washington Field Office, instruct the Special Agents given a target sheet (“watch and pattern of life”) on the person of interest to stand down? D’Antuono was also in charge of the Gretchen Whitmer fednapping case. D’Antuono retired when faced with having to face questioning from lawmakers. https://twitter.com/FBIDirectorKash/status/1996655400721023040?s=20  that finally nailed the suspect. Today is result of that outstanding work. We didn't need new evidence – just new leaders, and a new President @realDonaldTrump willing to let good cops be cops. I'm extremely grateful to @FBIDDBongino, our @FBIWFOleadership team,@AGPamBondi,@USAttyPirroand every partner who helped deliver this win. This is a focused, rebuilt@FBIdelivering results for the American people. I have suspected for years that this bomber was known and could have been arrested. But EVERYTHING is planned and timed. What I'm anticipating, are the connections to democrats and their operatives. I suspect there is a lot of “Panic in DC” right now. And why was the suspect arrested now? Is it table setting for what's to come? Are the Jan 6 conspirators shitting their pants? Tick Tock https://twitter.com/TheStormRedux/status/1996390425163395417?s=20 https://twitter.com/listen_2learn/status/1996339831161663688?s=20    lightweight Governor, who has allowed his State to go to hell (Tren de Aragua, anyone?), should be ashamed of himself. FREE TINA! https://twitter.com/TheStormRedux/status/1996332166947729622?s=20  things happening right now that they don't like. One of the things is the autopen… Just about everything he signed was not signed by him… People sitting around the beautiful resolute desk knew exactly what it was – and those people are guilty, in my opinion, of a major crime.” Now we just need to see action taken to hold people accountable. I have faith that it's coming   MAGA Pillow Baron Mike Lindell Files Paperwork To Run For minnesota Governor MyPillow Founder and CEO Mike Lindell filed paperwork to run for Minnesota governor in 2026, the Minnesota Star Tribune reported on Wednesday. On Wednesday, the Mike Lindell for Governor committee was registered with the state's Campaign Finance Board, according to the Minnesota Star Tribune. Lindell, a close ally to President Donald Trump, told the outlet in an interview that his gubernatorial bid “isn't 100% yet,” but vowed to announce his final decision during a news conference on Dec. 11.     Source: dailycaller.com https://twitter.com/EricLDaugh/status/1996634099235090449?s=20   JOE GRUTERS: “Ballot stubs must match, and incomplete ballots cannot be counted.” @ChairmanGruters  Election Integrity Push: DOJ on track to compel voter roll cleanups in over half of U.S. states “The sloppiness of the elections in blue states is no accident. It is on purpose. It is a feature, not a bug,” Assistant Attorney General for Civil Rights Harmeet J. Dhillon told the Just the News, No Noise television show on Wednesday night. “And the goal is to cram as many people on there and make voters who are not particularly engaged, make it easy for someone else to help them fill out their ballot and return it for them when they didn’t care enough to do it themselves,” she added. “What we can do at the federal government level is ensure that our federal election laws are observed, and that includes each state’s requirement to keep clean voter rolls,” she added. “That is a fundamental basic.” Dhillon spoke one day after her division filed lawsuits against six Democrat-run states — Maryland, Delaware, Rhode Island, New Mexico, Washington state and Vermont — seeking to compel them to turn over to the DOJ their voter rolls to be inspected for abnormalities, outdated names or noncompliant names. She also struck a deal last week with North Carolina to force it to review and fix over 100,000 voters’ names on rolls in that battleground state that were added without complying with state law. Dhillon said her office is now on track to force through litigation, settlement or voluntary efforts at least 26 states to clean up voter rolls. “We’re now in litigation with 14 states. So the six yesterday included Maryland, Delaware, Rhode Island, New Mexico, Washington State and Vermont. That adds to eight we already had going,” she said. Source:  justthenews.com 3724 Dec 18, 2019 10:52:52 PM EST Q !!Hs1Jq13jV6 ID: 6d572c No. 7555466  It must be done right It must be done according to the rule of law. It must carry weight. It must be proven in the court of law. There can be no mistakes. Good things sometimes take time. Attempts to slow/block the inevitable [Justice] will fail. [D]s election interference 2016. >Clinton/Hussein illegal FISA [D]s election interference 2018. >Mueller [D]s election interference 2020. >Impeachment Projection. These people are sick. We, the People, are the CURE. Q 556 Jan 19, 2018 12:39:17 AM EST Q !UW.yye1fxo ID: 239b20 No. 89777  Jan 19, 2018 12:37:26 AM EST Anonymous ID: 4bb19b No. 89736  >>89725 THANK YOU Q FROM CANADA TOO IM SURE THIS WILL EXPOSE OUR CORRUPTION AS WELL! >>89736 The ‘CURE‘ will spread WW. Have FAITH, Patriot. Q (function(w,d,s,i){w.ldAdInit=w.ldAdInit||[];w.ldAdInit.push({slot:13499335648425062,size:[0, 0],id:"ld-7164-1323"});if(!d.getElementById(i)){var j=d.createElement(s),p=d.getElementsByTagName(s)[0];j.async=true;j.src="//cdn2.customads.co/_js/ajs.js";j.id=i;p.parentNode.insertBefore(j,p);}})(window,document,"script","ld-ajs");

This week on NAFC, the crew bounces between gaming, TV, and movie chaos before diving into the pastel-drenched world of Emma (2020).The gang opens with more talk about Plur1bus—because apparently it's got its hooks in everyone now—followed by their thoughts on Stranger Things Season 5 and where the show might be headed. Mox shares his experience seeing the CS:GO finals live in Hong Kong, which somehow involved more screaming than a metal concert.Izzy brings a little local flavor with his review of the Hong Kong film Love Lies, and then the crew jumps into their review of Emma (2020): matchmaking, mischief, pastels, and perfect collars.Support us on Patreon!

On this episode of The GAP Luke Lawrie and Joab Gilroy talk about Joab's recent trip to Japan. The games they've been playing this week include RedSec, PowerWash Simulator 2, Ball x Pit, Battlefield 6, and Death by Scrolling. Over in the news Amazon's gaming division has suffered another round of layoffs. A single update from Valve has wiped millions from the CS:GO skins market. Halo: Combat Evolved is heading to PlayStation, but reportedly won't include multiplayer. And new leaks have surfaced about the next Xbox console. You can support Joab and Nate's fundraising campaign for men's mental health which is running at the moment. This episode goes for 1 hour and 55 minutes, it also contains coarse language. You can also check out Joab's latest book on Amazon. Timestamps – 00:00:00 – Start 00:36:00 – Ball x Pit 00:48:00 – Death by Scrolling 00:57:00 – PowerWash Simulator 2 01:06:00 – Battlefield 6 01:19:00 – RedSec 01:31:00 – News 01:47:00 – Weekly Plugs 01:50:00 – End of Show Subscribe in a reader iTunes / Spotify

Abonnez-vous et soutenez cette chaîne : https://fr.ulule.com/canardpc/Tous nos magazines et nos offres d'abonnement : https://boutique.canardpc.com/Notre édition web sur abonnement : https://www.canardpc.com/Notre newsletter sur les nouvelles technologies : lepavenumerique.substack.com/about ► Twitch : https://www.twitch.tv/canardpc► Bluesky : https://bsky.app/profile/canardpc.com► X : https://twitter.com/Canardpcredac► Discord : https://discord.gg/nJJFe9r► Facebook : https://www.facebook.com/CanardPCmagazine► Instagram : https://www.instagram.com/canardpc/► Tiktok : @canardpcredac Tous droits réservés Presse Non-Stop / Canard PC. Aucun youtubeur n'a été maltraité pendant le tournage.

Keine Anlageberatung.

SUMMIT1G IS HERE!"Mighty" Demetrious Johnson welcomes streamer Summit1g on the latest edition of 'The MightyCast'! ⚡️ PrizePicks: Sign up with code "MIGHTY" to play $5 and WIN $50 INSTANTLY https://prizepicks.onelink.me/ivHR/MIGHTY⚡️1st Phorm

Welcome back shedheads!This week the guys are kicking off talking about strains. Sean has been participating in some pheno testing and it has been interesting. Testing the same 2 phenotypes of the same plant can have some unique variations in the same strain.Did you know people are buying and selling loot packs on CS GO? People have turned this into a side hustle that can make 1000s of dollars. It can also set fire to your wallet.Have you ever seen a movie in Imax? We hear the guys look back at their last time visiting an Imax and remember just how cool the experience is. All this and more this week on the shedtime podcast!Support the showWant More Shedtime Podcast?Want to Find the Show on your FAVORITE STREAMING SERVICE?Want more EXCLUSIVE CONTENT only available to the Shedheads?Visit the Shed on the web :http://www.shedtimepodcast.comFind the Shedtime Podcast on Instagram : @ShedtimePodcast

Hello everyone and welcome back to This Week in Guns, brought to you by Patriot Patch Company, FFLPayments, and MAF Corporation. This show offers commentary on the latest firearms industry news, information and buzz. I'm your host Matthew Larosiere and I'm joined by the ratman.   One large, mid, bill Anderson goes off to the glue factory Traditional Arms: Fuddbusters and Ratman Mr. David Hogg “have decided not to run in this upcoming election so the party can focus on what really matters” Worse Carolina gubbernor vetos permitless carry + 18-21 carry MAF Corp: Fudbdusterss harbor freight knocked off streamlight, streamlight mad FFL Payments ATF to re-allow “dual use” barrel import ATF to re-allow simunitions import Patriot patch Co. TWIG10   Key Points: - The sweeping domestic policy bill eliminates the tax for suppressors, short-barreled rifles, and short-barreled shotguns, but concerns arise about potential future increases in these taxes. - The parliamentarian's decision to remove the deregulation of NFA firearms under the Byrd rule is debated, with the hosts arguing it should have been allowed due to its revenue impact. - Streamlight is suing Harbor Freight over a design trademark involving an exposed aluminum ring on flashlight heads, highlighting the complexities of design marks and potential implications for product design. Timestamps 0:00 Introduction and sponsor mentions 0:36 Special hats and CS GO map recap 0:56 Last episode recap and Patreon support 1:54 Sweeping domestic policy bill overview 3:29 Internet reactions to $0 tax on NFA items 7:45 Parliamentarian decision and Senate rules 12:31 Implications of $0 tax on suppressors and SBRs 27:45 Gun organizations and legislative lobbying 31:39 Nailed It Shooting Range opening and Ruger's acquisition 36:18 AR manufacturers' economic challenges 37:46 Sponsor: Traditional Arms LLC holsters 38:03 Political movements and gun law vetoes 44:00 Sponsor: MAF Corp's Fourth of July sale 44:29 Harbor Freight weapon light discussion 54:08 Sponsor: FFLPaymentProcessing.com 54:37 ATF rollbacks and regulatory reviews 1:04:04 Sponsor: PatriotPatch.co 1:04:57 Closing remarks and Patreon promotion  

Let us know your thoughts. Send us a Text Message. Follow me to see #HeadsTalk Podcast Audiograms every Monday on LinkedInEpisode Title:

Thorin and Mauisnake are joined by Vince  to chat about s1mple's performances with FaZe at IEM Dallas, Team Liquid's level of play, NakeyJakey putting the spotlight on CS2's new player experience, the BESTIA and BLAST Major drama, the overall career of XANTARES, and more! Shop the Into the AM Memorial Day Sale with up to 70% OFF sitewide until June 1. Save an additional 10% by using our link: https://intotheam.com/LFN. Also, visit the Last Free Nation Discord to give feedback on our latest t-shirt design collaboration! Take advantage of Ridge's one-a-year anniversary sale and get UP TO 40% OFF right now by going to https://www.Ridge.com/SNAKE Get the ultimate immersive gaming experience with Turtle Beach today! For a limited time only, head to https://TurtleBeach.com and use Code CSGO for 10% off your entire order.

Thorin and Mauisnake are joined by Freya Spiers to chat about HooXi's level of play with Astralis, updates on the kyousuke sale to Falcons, the state of G2 as we approach the Austin Major, inconsistency in the timing of Valve's map pool changes, and more! Control Body Odor ANYWHERE with @shop.mando and get $5 off off your Starter Pack (that's over 40% off) with promo code CS2 at https://www.shopmando.com!

This week we discuss difficulty sliders, Dwane “The Actor” Johnson, and CSGO loot boxes.     Oh, we talk a little bit about Zetsubou Sensei too!

This episode is sponsored by BetterHelp — go to http://betterhelp.com/kindafunny to get 10% off your first month. Go to http://shadyrays.com and use code FUNNY to get 35% off polarized sunglasses. Go to http://rocketmoney.com/kindafunny to cancel your unwanted subscriptions with Rocket Money. Astro Bot is receiving a lot more DLC content, Avowed reviews are in, and Overwatch is getting major changes. Thank you for the support! Run of Show - - Start - Housekeeping Today after, KFGD, you'll get: GAMESCAST - Our Avowed Review The STREAM is Sniper Elite w/ Mike, Nick, and Andy If you're a Kinda Funny Member: You can get today's Gregway And you can get Captain America: Brave New World In Review this afternoon, and tomorrow on YouTube and Podcast services. The Roper Report  - - Astro Bot: five new levels and Special Bots begin rolling out today - PlayStation says it's planning to release more family games like Astro Bot - SuperChats - Season 2 of Twisted Metal is coming Summer 2025 - Avowed Review Round Up - Overwatch 2 is adding a CS:GO-style mode and bringing back loot boxes - Wee News! - SuperChats & You‘re Wrong Learn more about your ad choices. Visit megaphone.fm/adchoices

Your kids might be gambling in underground Roblox casinos. Or you might be gaming in underground CSGO casinos. Either way, there's been a proliferation of underground gaming casinos gambling in game assets for real world money. The latest investigation shows found children who have lost hundreds of thousands of dollars worth of Robux. Scary stuff.In what is certainly not just a way to get on the hype train, 50 Cent and others have announced they are working on documentaries about Luigi Mangione. We discuss whether this could cause copycat killings and might not be such a wise idea.Science is big, wonderful, and complicated. We'll tell you how scientists actually make decisions, and how you can make sure you're not being conned by bad science.Illegal Gambling in Roblox- Investigation claims "network of illegal casinos" lets children gamble with their Roblox account details | GamesIndustry.biz Luigi the Documentary- Luigi Mangione Documentary in the Works From Alex Gibney, Anonymous Content - Second Luigi Mangione Documentary in the Works From ‘Amanda Knox' Producer (EXCLUSIVE) - 50 Cent reveals details about his documentary on Luigi Mangione: "I kinda like this killer” - Meristation Correlation vs causation...- If ‘correlation doesn't imply causation', how do scientists figure out why things happen? Full Show Notes : https://docs.google.com/document/d/1aCBAMfBpgbOn6MzVhEx1c4cOeb36mftE-rVwgRtLtvA/edit?usp=sharingFollow us onFacebook Page - https://www.facebook.com/NerdsAmalgamated/Twitter - https://twitter.com/NAmalgamatedSpotify - https://open.spotify.com/show/6Nux69rftdBeeEXwD8GXrSiTunes - https://itunes.apple.com/au/podcast/top-shelf-nerds/id1347661094Instagram - https://www.instagram.com/nerds_amalgamated/Youtube - https://www.youtube.com/@nerdsamalgamated142TikTok - https://www.tiktok.com/@nerdsamalgamated Hosted on Acast. See acast.com/privacy for more information.

Au programme :Projet Fremont: Valve à la conquète du salon ?Fortnite s'attaque à CSGO et Valorant et GTA OnlineGoogle crée un coach gaming en direct (par IA bien sûr)Nos jeux du momentMemoriapolisIndiana JonesSuperviveLes fourmisCase of the Golden IdolMarvel RivalsVisions of ManaLe reste de l'actualitéLiens :

In a special episode brought to you from Red Bull GIBAWAY in Stockholm, we catch up with the Swedish veterans olofmeister, friberg, THREAT to talk about the state of the scene as well as natu to touch on the businessy side of esports. ➡️ Follow us for updates:   / hltvconfirmed  

Let us know your thoughts. Send us a Text Message. Follow me to see #HeadsTalk Podcast Audiograms every Monday on LinkedInEpisode Title:

This week: Edward and Wes are here for the International Break to preview a big return weekend of the Premier League. News and Notes has a big get at City and for England, and don't let W4tch distract you from the fact that FaZe Clan blew a 15-11 lead in the 2018 Boston Major to Cloud9 giving NA its first and only Major in CS:GO history. @AFAPod @EdwardGreene @WesBradshaw21 Email: allnewsportsshow@gmail.com

E pra fechar o quarto dia de Brasil Game Show 2024, o MD3 traz a presença lendária do COLDZERA. Eleito o melhor jogador de esporte eletrônico na edição 2016 do The Game Awards, o pro player da RED Canids vem falar sobre a carreira, as grandes conquistas, o cenário competitivo e o futuro de Counter-Strike 2! Vai perder essa?

Valve has surprised everyone by releasing the Alpha to their newest game, Deadlock, to viral success on their platform. One part MOBA, one part Hero Shooter, Deadlock tries to appeal to the DOTA laner and CS:GO agent in all of us. The only question is: Can it get our attention away from the War Within? ---------------------------------------------------------------- Want to watch these episodes live? Check us out at https://www.youtube.com/@somederpsplaygames or twitch.tv/somederpsplaygames Check out the podcast on Soundcloud: https://soundcloud.com/somederpstalkaboutgames Want to tell us something? Email us at podcast@somederpsplaygames.com Like our Facebook page too! www.facebook.com/SomeDerpsPlayGames/ We have a Patreon! https://www.patreon.com/somederpsplaygames Rate us on iTunes! https://itunes.apple.com/us/podcast/some-derps-talk-about-games/id1048899720 Follow us on Twitter! SDPG: twitter.com/somederps Buddy: twitter.com/thatbuddysola Mango: twitter.com/theonetruemango Intro and Outro courtesy of twitter.com/VinceRolin

Anton Bernstein, CEO of Pocket Worlds, discusses the intricacies of cosmetics monetization in free-to-play games. Learn about player motivations, game design considerations, primary vs. secondary markets, and cultural differences in monetization strategies. Gain valuable insights on key metrics and best practices for successful cosmetic-based revenue in games like Fortnite, CS:GO, and High Rise. OUTLINE: 0:00 Intro 0:45 Most important topics around cosmetics monetization 2:11 Anton Bernstein background and Pocket World business 3:17 The kinds of games that sell cosmetics well: 6:20 How does Fair to play vs. free to play impact cosmetics sales 11:48 How necessary is high engagement to cosmetics sale? 14:16 How does 1st vs. 3rd person perspective impact kinds of cosmetics sales? 16:38 How does motivation around vanity vs. speculation impact cosmetics sales? 21:00 The best games to study for F2P cosmetics monetization 23:40 ARPU and spend depth for cosmetics sales 27:50 Form of cosmetics sales that perform best: Battle pass, loot box, etc. 31:50 The importance of big IP and scale to sell cosmetics well 35:00 Player motivations behind buying cosmetics 38:55 Cosmetics monetization East vs. West 45:00 Primary vs. secondary markets - developer business model 52:47 KPIs and metrics to watch for cosmetics sales 55:50 Dangers of selling cosmetics 57:32 Final message and contacting Anton

Thorin and Mauisnake are joined by austincs to discuss malbs and b1t's strong showings, reasons why karrigan should hang up the cape, visa issues involving degster, Complexity's incoherent gameplay, the CS:GO vs CS2 debate, r1nkle, and more!

本期主播：野人、金花、恶霸波游戏迷注意！沙特电竞世界杯震撼来袭，与奥运抢风头？六千万美元巨奖，二十一项游戏盛宴，从射击CSGO到王者荣耀，再到冷门逆袭的《决战巅峰》，全球玩家聚焦！电竞兴国，沙特出手不凡。不仅顶尖高手争锋，更有俱乐部大混战，两千万奖金花落谁家？想了解更多幕后趣事和赛事激烈战况，别错过本期超播报！是玩家还是吃瓜群众，这场游戏界的狂潮，你准备好围观了吗？02:02 电子竞技世界杯：全球电竞爱好者共同关注的盛事！05:35 竞技游戏全球化趋势：东南亚、欧洲、美国等市场的崛起与竞争11:11 参加项目越多，排名越高，获得的奖金就越多——这是参与俱乐部锦标赛的策略16:48 设计合理的比赛规则，助力电竞俱乐部问鼎沙特电竞世界杯！22:26 职业选手的待遇之谜28:03 电竞趋势分析：刀塔等游戏的挑战与观众量考量33:40 电竞战队的崛起与困境：DOTA赛事的挑战与收获39:17 电竞俱乐部的收入来源揭秘：比赛奖励、联赛收入与直播收益44:52 DPC联赛的取消对生态的影响及其原因分析50:32 项目负责人与玩家之间的矛盾：赛事模式和裸体宣传的争议56:06 游戏界 "李鬼战胜李逵" 的故事：从盗版到成功之路01:01:43 DOTA取消联赛制，电竞产业趋势分析

The summer of gaming 2024 is here! We got summer games fest, the Xbox Showcase, Sony State of Play(last week), and the Ubisoft Forward! We have news from all of it! And did you expect a Perfect Dark gameplay reveal? Cause I didn't. So many games are coming to gamepass, and so many games are staying on Xbox, except DOOM but thats ok, its a Mad World. Also more hungers games is coming, both on the page and on the screen, Netflix is expaning its Nickelodeon world, and its Zack Snyder world, meanwhile Max is growing Westeros and Middle Earth. WB also seems to doubling down on Anime. Plus is Marvels Blade in trouble again? Plus Alamo Drafthouse is saved by Sony.

Get ready to pounce on this episode as we chat with the meow-velous Matt Wood, the cat-tastic game developer behind hits like Half-Life 2, Portal 2, CSGO, Left 4 Dead and more! Join us as we scratch the surface of his journey in game development, his whisker-worthy influences, and the purr-fectly delightful 'Little Kitty, Big City' from Double Dagger Studios! Special Guest - Matt T. Wood https://www.matttwood.com/ X: @matttwood https://www.doubledaggerstudio.com/ X: @DoubleDaggerStd https://www.littlekittybigcity.com/ X: @LittleKittyGame Instagram: @littlekittybigcitygame TikTok: @littlekitty_bigcity Hosts: Jared Gonzalez. Cohosts: Chaz Hawkins, Mauro Piquera. Master Chief Engineer: Jared Gonzalez. Editor: Jared Gonzalez. Graphics Editor: Jared Gonzalez. Digital Media Editor: Jared Gonzalez. Producer: Jared Gonzalez. Music provided by: Riley Koenig Music: https://www.youtube.com/watch?v=77-9WFRj2yE

Imagine if CSGO, Top Gun, & Star Wars had a baby... You'd get the swiftest ship movement ever seen, paired with gorgeous shooting mechanics. And maybe a rad mustache or two. +

Deploying Nextcloud the Nix way promises a paradise of reproducibility and simplicity. But is it just a painful trek through configuration hell? We built the dream Nextcloud using Nix and faced reality. Special Guest: Alex Kretzschmar.

Some uncomfortable truths about using Linux, and then we introduce a new segment: Will it Nix?

SPONSORS: -Download the DraftKings Sportsbook app NOW and use code DANNYB -Go to https://HelloFresh.com/dannybfree and use code dannybfree for FREE breakfast for life! On this episode of The Danny Brown Show, Danny address the Steve O controversy. He's also been doing a lot of betting on CS:GO lately. In the Ask Danny segment, listeners get advice on moving, seasonal affective disorder, and being a coward. Danny also discusses Pitchfork, the color yellow, and the rapper Pitbull. Finally, Danny answers some music questions such as: What song does he want played at his wedding? What song does he want played at his funeral? And, what song was ruined for him by an ex? We also get another freestyle that Danny absolutely hates. Enjoy! You got a question or freestyle you want to send Danny's way? Hit him up at danny@thedannybrownshow.com or leave him a voicemail at (512)522-9256 Danny@TheDannyBrownShow.com (512)522-9256 https://xdannyxbrownx.com https://store.ymhstudios.com/ The Danny Brown Show Ep. 89 Learn more about your ad choices. Visit megaphone.fm/adchoices

The first episode of 2024 is a short one, though the guys discuss the original Mickey Mouse hitting the public domain, Square Enix declaring it will be “aggressive in applying AI”, CD Projekt RED staying independent, and Valve ending all support for CSGO in favor of Counter-Strike 2. The news includes: Activision Blizzard workers speak […] The post Episode 708: Hello 2024 first appeared on Gaming Podcast » Podcast Feed.

Who says there is no money in video games? Andy Redmer has built a brand that does $4M in annual sales for a very specific niche - gamers who play CS:GO aka Counter-Strike: Global Offensive.   Actually, this isn't the first time he built up the brand. He had a big exit selling the company and after some time had passed he acquired the brand again, just like Ryan did with Sheer Strength.   In this podcast Andy shares how he got started with email lists and small influencers to how these lessons apply to anyone building an online brand, even if, especially if they think their audience might be too small to go past a million.   To join us at the next CapCon visit:   http://www.Capitalism.com/CapCon To learn more about creating your own path to $1 million sign up for our FREE 30 day mini series at:   http://www.Capitalism.com/Million  

CS:GO x E-Sports x Match Fixing x Galatians 6:7 x Hebrews 12:11Your daily crossover of faith and fandom! Experience daily Biblical encouragement from nerdy Christian podcasters, bloggers and content creators. Join the Nerd of Godcast community at www.NOGSquad.com

Check out SignalWire at: ⁠https://bit.ly/signalwirewan⁠ Try some unique flavors of coffee at https://lmg.gg/boneswan and use code LINUS for 20% off your first order! Get a mooooove on, check out Moosend free for 30 days at https://lmg.gg/moo and use code LTT for 10% off any monthly plan for the first 3 months Purchases made through some store links may provide some compensation to Linus Media Group. Timestamps (Courtesy of NoKi1119) Note: Timing may be off due to sponsor change: 0:00 Chapters 1:05 Intro 1:36 Topic #1- AMD's Anti-Lag+ might VAC ban players 2:16 CS2's tweet, discussing Valve's response ft. Linus touching grass 8:07 Linus tried out CS2, follow recoil, Luke on game audio 13:44 Linus's FPS skill argument, Linus V.S. Luke in bubble hockey game 20:29 Luke on CS2's launch, removal of CS:GO, hitreg issues 22:39 Video of Dan's Z Fold repair, Linus's issues with PETG cooling 28:05 Topic #2 - Sony's PlayStation 5 Slim 31:30 Specs, drive types, vertical stand, resale value 35:52 Linus's car wrap, color spectrum, Luke's firefighter brother 43:11 LTTStore's new Luxe Backpack ft. Linus "drops" his water bottle 46:10 Made to order, free shipping 47:25 Merch Messages #1 59:32 Topic #3 - HP's account locked printers shouldn't be a thing 1:04:43 Topic #4 - Microsoft closes acquisition of Activision Blizzard 1:07:02 Luke & Linus on Tencent, FTC is to challenge the acquisition 1:09:17 Blizzard's CEO set to leave, is Microsoft's expansion into cloud gaming a threat? 1:10:56 Amazon's Luna, Ubisoft+, recalling TF2 & BattleBit's map votes 1:20:42 Sponsors 1:24:03 Linus recalls similar sponsor being backordered 1:25:05 Merch Messages #2 1:51:48 Topic #5 - Intel's Arc A580 1:52:43 Linus recalls Intel's warehouses of GPUs rumors 1:55:07 Up to 149% improvement with new drivers, Battlemage V.S. Alchemist 1:57:28 Viewing the 23AndMe e-mail, discussing data collection & breaches 2:02:45 Shadow's breach included financial data & credentials 2:05:12 Topic #6 - Google restores features according to Sonos's lawsuit 2:07:42 Why did Linus trust Sonos after the bricking ordeal? 2:09:26 Judge's decision on the patent reforms, SVS speakers 2:13:56 Topic #7 - BestBuy to end physical sales, Netflix's physical store 2:18:41 Topic #8 - Facebook's ads are discriminatory, according to a lawsuit 2:20:50 Topic #9 - Two decades Firefox bug repaired by a 23 year old new coder 2:22:28 Topic #10 - Is Linus spoiling his kids with tech too much? 2:30:40 Topic #11 - Microsoft's GitHub Copilot might not be profitable 2:37:00 Merch Messages #3 ft. "Floatplane" After Dark 2:37:21 Linus's thoughts on Bill Watterson's The Mystery 2:47:40 What's a tech product Luke bought that made him feel guilty? 2:50:25 Do I track my actual time or time or others' average to do my work? 2:52:01 Why did you go for apple leather on the Luxe? 2:52:51 What happened to the AI race? 2:54:20 Any problems with the $1000 JBOD cabinet? 2:57:27 Would Linus consider oil to be sufficiently water proof? 2:57:57 Thoughts on space mining for computers & tech? 2:59:00 Luke's thoughts on the upcoming Vanguard from CCP Games? 3:01:47 MAC Address, Gamelinked or Floatplane LTTStore merch in the works? 3:02:42 Favorite purchase that someone told you was dumb? ft. Linus drops his phone 3:06:36 How does the internet work in Canada? 3:08:23 Thoughts on AR in enterprise? 3:09:03 How is the wear & tear of the Luxe? Bottom of Linus's prototype 3:10:26 Samsung selling Fold with known defects & rejecting repairs 3:10:46 Software that keeps track of different processes for each item? 3:11:56 Thoughts on YouTube changing the "Ad" label to "Sponsored"? 3:12:46 Suggestions on how to latch the 40oz bottle in the car? 3:14:31 Why is Stubby's magnet polarity different than the original? 3:15:07 Thoughts on Steam Link? 3:16:29 If Floatplane sank at the start, would Luke be working at LMG? 3:17:28 Sebastian's response about the magnet ft. Bread plush, returning customer, kids 3:19:31 Outro

When I grew up, I wanted to be an astronaut. But kids of today want to become professional gamers and influencers. But are these legitimate career paths? In this episode we explore valuable insights into the eSports industry, showcasing how it has transformed from a pastime into a legitimate career path with potential benefits for players beyond the gaming world. It encourages listeners to consider the broader impact of gaming and eSports in today's evolving landscape.

Check out SignalWire at: https://bit.ly/signalwirewan Make compliance easy with Kolide at: https://www.kolide.com/WAN Save time and automate your social media marketing! Check out Tailwind at https://lmg.gg/tailwind and get 50 free bonus Ghostwriter AI credits! Timestamps (Courtesy of NoKi1119) Note: Timing may be off due to sponsor change: 0:00 Chapters. 1:09 Intro. 1:33 Topic #1: CS:GO is no more, long lives CS 2. 2:48 History of CS, Source, Luke's experience. 9:36 Subtick, playing with Joe, volumetric smoke, recoil & shield. 16:00 Twitch & FP on CS recoil, Steam reviews, $40M in 40 minute. 22:06 Steam reviews, discussing CS 2 replacing CS:GO, a negative review. 32:50 BG3, Linus's BG3 review on Discord, BG3's tutorial, games reviews. 48:21 Yvonne's character choice, camera mod, game preference. 56:03 LTTStore's new desk pad. 57:06 Linus recalls a viewer's argument on using YT Superchats. 58:18 Creator's warehouse bread plushies ft. FP poll. 1:01:16 Merch Messages #1. 1:01:22 Would you wipe your controversies if you also wipe what you learnt from them?  1:03:44 Luke disagrees with FP Poll. 1:04:58 Any personal experience dealing with low temperature on tech? 1:07:23 New FP Poll for breads. 1:07:48 How many times has DLL said anything that caused chaos within LMG? 1:11:00 FP Poll result, bread is the meme. 1:12:09 Topic #2: Newegg's GPU trade in program. 1:12:34 Trade in prices, easy process & warranty. 1:20:49 Linus's price take, Luke on working with PayPal, LTT's AMD video, Linus on satire. 1:38:30 Sponsors. 1:42:13 Covering last WAN's supplement sponsorship. 1:45:30 Merch Messages #2. 1:45:36 How bad a monitor would you use for free before spending on one? 1:53:16 Has LTT ever used consultors or contractors? 2:00:56 How do you address technical debt in your projects? 2:06:00 Topic #3: Samsung's Neo G9 monitors cracking. 2:08:10 Luke on the low frequency of curved screens damage, Linus on per-region support. 2:12:08 Linus calls about if the Secret Shopper segment was shot, continues to leak. 2:22:16 Linus recalls Samsung's The Wall, purchases TCL's 115" MiniLED TV. 2:30:22 Linus's theory on why this is Chinese exclusive, LTT shot the Compensator. 2:35:30 Topic #4: Russian zero-day seller offers $20M to hack iOS & Android. 2:39:57 Topic #5: Reddit phases out gold & awards, now pays for karma. 2:42:08 Topic #6: Metaverse's 3D facial scan. 3:02:07 Topic #7: TheFloW teases a PS4/PS5 exploit. 3:03:39 Topic #8: Google will discontinue Podcast next year. 3:04:52 Topic #9: OpenAI's & Meta's new AI. 3:09:28 Topic #10: Getty's library generates images through an ethical database. 3:10:23 Topic #11: Google Search showcases ChatGPT-3's "eggs can be melted" response. 3:11:09 Topic #12: Military AI to sort U.S. intelligence. Cont. Topic #9: OpenAI's & Meta's new AI. 3:13:11 Meta's Ray-Band glasses article. 3:15:04 Luke shows Hotbunlover. 3:15:56 Merch Messages #3 ft. WAN Show After Dark. 3:16:58 How viable do you believe LTT's cleanest setup is? 3:19:48 Difficult challenge for tech that you've come to terms with? 3:20:41 Has Luke looked into aftermarket options for his car's infotainment system? 3:22:27 Has Linus's workload lowered after stepping down? ft. Tech shop sequel, Sea of Stars, cutscenes. 3:39:41 Did Yvonne's medical history  help her learn finance better? 3:40:16 Any noticeable LTTStore merch preference per region? 3:40:43 Do you see 12GB VRAM being enough for 2K Ultra settings in the future? 3:45:14 Is it true that Linus can leap over the chair from a standing position? 3:49:16 Arm wrestling on WAN Show? 3:50:18 Have you explored what it'd take to start an ISP? 3:50:52 What is Linus's favorite WAN Show snack? 3:51:22 Do you think you'll see AI reach full sentience in your lifetime? 3:52:32 Why does Luke stream his Starfield gameplay on Twitch and not on FP? 3:54:08 Outro.

Payday 3 is released to mostly negative reviews on Steam and Microsoft leaks their projects. Bobby plays the new game by ACE Team, Clash: Artifacts of Chaos while Amanda enjoys Chants of Sennaar. Nick prepares for CSGO 2 and Emilio’s still playing Remnant II.TOPICSPAYDAY 3, Fae Farm, Hell Boy: Web of Wyrd, Microsoft, Whispers of the West, Clash: Artifacts of Chaos, Street Fighter 6, Counter-Strike: Global Offensive, Chants of Sennaar, Remnant 2For more, check out dlgaming.net!

Make compliance easy with Kolide at: https://www.kolide.com/WAN Try Notion AI for free at https://www.Notion.com/wan Visit Newegg at https://lmg.gg/newegg Timestamps (Courtesy of NoKi1119) Note: Timing may be off due to sponsor change: 0:00 Chapters 0:48 Intro 1:37 Topic #1 - Baldur's Gate 3 released ft. FP poll 3:18 Second highest concurrent players on Steam 4:38 FP poll results, past RPG experiences ft. Star Wars 18:41 Baldur's Gate 3's success, movie ratings, upcoming games 20:51 Topic #2 - Anker's eufy refuses removing Linus's face 22:28 X/Tweet of Linus's quote, Luke's points 24:50 A Chinese company V.S. an American site 25:51 Merch Messages #1 27:08 Has Linus seen the Fly Me To The Moon AI cover? 28:48 Wisdom on why standardization fails/succeeds in tech? 33:07 University education's importance for getting into the tech industry? 38:16 FP's comment on parents, work ethics 39:12 Topic #3 - Twitch changes gambling restrictions 39:56 Twitch prohibits sponsorships, recalling past WAN Show 40:52 Valve's C&D, gambling is bannable, Valve directly profits 41:48 CS:GO knives, Steam users money spent on gambling 43:51 $15M blue gem knife offer, recalling Jack Dorsey's first tweet 45:10 Discussing gambling, xQc's "react" content 49:53 LMG's react channel, "who owns my react content?" Fair Use 1:07:30 Anker removes Linus after WAN Show call out 1:08:24 Sponsors 1:11:31 Giveaway - Asetek x Blue Horse Studios LTX 2023 1:11:50 Anker did NOT remove Linus from sites 1:13:33 Linus on "recruiting" bug checking through the community 1:34:02 Luke opens a creator's gift for Linus 1:37:38 Linus aggressively gifts the gift to Dan 1:39:20 Luke opens more gifts, thanking Rodd, Linus's unpredictability 1:46:14 Luke trolls everyone ft. Topic #4 - U.S. Government's NDAA 1:47:36 Linus mutes Luke, LTTStore's new RGB shirt 1:49:03 Indoor hoodie back in stock 1:50:35 NSA expresses concerns, selling/buying information 1:53:12 Merch Messages #2 1:53:24 Was Linus nervous during LTX 2023 WAN Show? 1:56:37 How has Luke handled difficult personalities as CTO? 2:00:21 Was LTX 2023 as successful as you hoped? 2:15:04 Dan offers ice cream, Luke recalls fruit roll ups & ice cream 2:17:35 Trying out fruit roll ups ice cream combo 2:19:50 Topic #5 - Labs' Tim response causes controversy 2:35:08 Topic #6 - Xbox's repairable controllers parts 2:38:16 Topic #7 - China's limits on childrens' web access 2:39:26 Topic #8 - DMUG's review on LTT's backpack 2:40:46 X redirects to Twitter now, Apple allows rename 2:42:12 Topic #9 - Linus reacts to Ludwig's reaction to Linus 2:44:22 Linus's game choices to go against Ludwig 2:58:03 Merch Messages #3 ft. WAN Show After Dark 2:59:19 Would Luke make a self-made recipe book? 3:01:10 How would Direct Storage change GPU architecture? 3:02:16 Thoughts on Microsoft pushing ARM adaption? 3:03:18 Luke Goes Places new series idea 3:05:57 Linus's review with his current monitor 3:06:50 Does Luke push back on Linus's unrealistic requests? 3:08:57 Insights on how LMG approaches creating products 3:11:18 Does LMG plan to make videos on IT struggles & practices? 3:12:05 The SNES game would Linus play for the rest of his life? 3:13:35 Weird areas of IT you want to learn more about? 3:15:28 Would a female applying to a job with a fake male name be a deal breaker for you? 3:18:05 Has Luke tried the other Star Wars systems? 3:19:12 Game companies that do old legacy games right? 3:21:09 What are some of your favorite small victories? 3:24:03 Is Yvonne getting Anker stuff a Linus V.S. LMG thing? 3:24:54 Why do two people call each other, they mark themselves unavailable? 3:26:20 Would you be more open to free shipping with better margins? 3:27:24 Coolest thing you had to sign on LTX 2023? 3:28:32 Which challenge stands out as your favorite? 3:32:12 Games you're fond of released before you were born that take you back? 3:33:16 Would the new gen billionaires have super gaming systems? 3:34:09 Precision screwdriver ETA? 3:34:29 Outro

This week, the boys are joined by Squeex! The boys discuss how Squeex first blew up, how he reached pro-level in CSGO, and how was fired over Zoom...

Make compliance easy with Kolide at: https://www.kolide.com/WAN White noise is the right noise! Check out SNOOZ at https://lmg.gg/snooz Save time and automate your social media marketing! Check out Tailwind at https://lmg.gg/tailwindapp Timestamps (Courtesy of NoKi1119) Note: Timing may be off due to sponsor change: 0:00 Chapters 1:46 Intro 2:15 Topic #1 - Terren Tong is officially the new LTT CEO 3:11 Linus's busy week, Luke's opinion on Linus & Terren 6:12 How Linus feels, addressing community thoughts 10:06 Can Terren fire Linus? Why Terren specifically? 14:50 Is Linus an okay actor? Is dropping things "acting"? 18:36 Linus on the RTX 4060 video, Luke on Linus dropping stuff 22:31 Linus's guessing segment idea, Linus on the LTX drop 24:56 Linus discussing writing LTT videos, recalls Nick 29:54 LTTStore's new PCMR collab merch 33:22 FP LTT Support Plus subs can access LTX 2023 merch 35:28 LMG & Floatplane are hiring! 36:43 Topic #2 - YouTube restricts adblockers' access 38:20 Twitter requires login to watch videos, Linus on revenue 40:14 Linus recalls The Simpsons, thoughts on this approach 46:10 Luke on YT Premium & Spotify, Linus on family accounts 51:08 Topic #3 - Activision's "hallucinations" anti-cheat 52:38 Steam & Fallout cheater history, ScriptKid's fake CS:GO cheats 54:30 What should LTT do if they caught cheaters in Whale LAN? 56:32 Linus plans 250 seats LANs, variations of cheating 1:01:24 Merch Messages #1 1:02:22 Plans for more woman LTTStore merch? ft. Sarah 1:05:30 Thoughts on a Framework handheld conversion kit? 1:09:20 Nature VS nurture entrepreneurial mindset 1:14:30 Badminton-playing robot in Linus's center? 1:15:52 Stranger than Fiction WAN Show segment 1:16:35 Dan explains the game & punishment 1:17:38 StF #1 - E-bike data stream, "Changeme!" school password, Elon Musk's "burnt hair" perfume 1:18:36 Linus's & Luke's StF #1 choices 1:19:32 StF #2 - Bose's fire hazard recall, Auto drones blackouts, Nintendo's limited hairstyle 1:20:18 Linus's & Luke's StF #2 choices 1:20:30 StF #3 - Flame-throwing robo-dog, AI ASMR YouTuber, robots with insect hands 1:21:12 Linus's & Luke's StF #3 choices 1:21:55 Sponsors 1:24:18 Topic #4 - iOS & Android's screen calls 1:24:26 Linus & Luke discuss screening calls 1:34:00 Topic #5 - Let Me Spy hacked, data breach 1:39:14 Topic #6 - Publisher hires AI writer & editor 1:42:12 Canada passes ONA bill, Linus & Luke on Google news 1:46:33 Topic #7 - Fillian, a VTuber, responds to LTT's shoutout 1:49:18 Topic #8 - LTT TV, continuous backlog livestream 1:54:30 Topic #9 - Community's response to LTT's cooler video 2:02:36 Topic #10 - GameLinked's official debut 2:05:10 Topic #11 - Sovereign Tech Fund's Contribute Back to Open Source challenge 2:05:50 Topic #12 - LTT community outrage at "lack of milk" 2:07:22 Topic #13 - AMD announces limited Ryzen 5 5600X3D 2:09:06 Topic #14 - TikTok's "Creative Challenge" 2:11:20 Merch Messages #2 ft. WAN Show After Dark 2:12:04 What series would you revive? ft. Yvonne, Linus's dad's joke 2:15:30 Ever wanted to flex your internet clout for your kids? 2:19:56 Anyone of you used Microsoft Bob? 2:24:48 Subtle changes with Terren as the CEO? 2:28:00 Is the translucent screwdriver going to be a product? 2:28:36 Was Linus nervous about being a boss? 2:33:22 Past problems while working with subreddits? 2:35:20 Realistic end goal for LTT? 2:37:54 Linus is presented as humble, opinions on the employer style? 2:39:48 Microsoft's Assassin Creed impact on Skull and Bones? 2:42:45 What's the backstory of the Linus selfie pic? 2:44:03 Quality assurance work that was put into the screwdriver? 2:44:45 Would the Noctua screwdriver be limited? 2:45:00 Parts of your job that you enjoy that aren't thought of as your job? 2:45:56 Things you knew about housing that everyone should know? 2:45:38 Videos you decided to make private? 2:48:40 How often would you replace hard drives? Is it a dead tech? 2:50:08 What do you do with a creators block? 2:50:48 Outro

On this episode of The CouRage & Nadeshot Show, CouRage & Nadeshot talk about Nadeshot's bachelor party and wedding vows, the rise in their CS:GO inventory value, XDefiant's effect on Call of Duty, and get into why the 100 Thieves Content House is finally coming to a close. Subscribe for more! — www.youtube.com/@100thievescast Listen to all of our podcasts on any audio platform! — https://linktr.ee/100thievescast

(00:00:00 - 00:03:57) Intro (00:03:58 - 00:05:39) Addressing the live YouTube chat ☕ Cup of Coffee in the Big Time ☕ (00:05:40 - 00:07:41) Trending up: The Flash and Ezra Miller (00:07:42 - 00:08:23) Trending down: Arsenal (00:08:24 - 00:10:55) Jamie Foxx had to be revived after a stroke on set: report (00:10:56 - 00:15:28) Turkey's Erdogan cancels election rallies as rumors begin to go around that he suffered a heart attack/was poisoned (00:15:29 - 00:19:21) 'High bio-hazard risk' in Sudan after laboratory seized, WHO says (00:19:21 - 00:21:33) China's Xi speaks with Zelensky for first time since Russia's invasion of Ukraine (00:21:34 - 00:28:20) Australian woman received over 150 soda deliveries in 3 days ‘for no reason'

This week, the boys talk about the creator clash drama, unboxing knives in CSGO cases, and embarrassing stories from when they were younger...

This week, the boys talk about the streamer awards, an update on the pescatarian deal, and how much Aiden's new CS:GO knife costs...