Podcasts about mtp

It's Sunday...and that means it's time for your weekly dose of MTP! This week, it's Roy's pick...and he's leading the charge back into "so bad, it's good" horror with his pick of 2005's Return of the Living Dead: Necropolis"! What did your co-hosts think of this movie? Did Kevin survive Roy's pick? Find out now on episode 182 of The Midnight Terrors Podcast!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

It's Sunday...and that means it's time for your weekly dose of MTP! This week on the show, we welcome back our long-time dear friend and horror author, Z.C. Krol! We're getting a little crazy this week with a modern horror movie that somehow manages to maintain a campy 80's element as well! And so, we had to bring in an expert! This week, it's Kevin's pick...and the trio of Kevin, Roy and Zack sit down to discuss 2021's ever perplexing and very fun movie from James Wan...Malignant! What did your co-hosts think of this movie? Find out now on episode 181 of The Midnight Terrors Podcast! SPOILERS EVERYWHERE! If you have never seen this movie...go watch it spoiler free and come back to hear our thoughts! Check out Z.C. Krol's Novella, "Disseverment":Amazon.com: Disseverment: A Horror Story eBook : Krol, Z.C.: Kindle StoreCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

It's Sunday...and that means it's time for your weekly dose of horror talk with MTP! This week, it's the long-awaited return of our good friend, Dan of Dirty Gnome Studios! A while back...he suggested a movie for us to talk about with him...and we're finally sitting down to do so! Tune in as Kevin, Roy, and Dirty Gnome Dan sit down to discuss Dan's pick of 2011's outrageous horror comedy home invasion film...You're Next! What did your co-hosts think of this movie? Is it an underrated gem that you should see? Find out now on episode 180 of The Midnight Terrors Podcast!!!Check out Dan's collectible gnome figures for purchase on his website:Home · Dirty Gnome Studios · Online Store Powered by StorenvyCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Watch the full episode on YouTube:We first covered Baseten last year when DeepSeek mania was at peak hype. Now they have raised a monster $13B round and become one of the new cohort of AI Infra decacorns that are (with Nvidia, Intel, and the semis complex) chief beneficiaries of the Inference Inflection. We return to Baseten at the peak of the 2026 edition of Open Weights debate. Ali has published a viral breakdown of Kimi K3:And since you last saw him, Philip has spoken at AI Engineer and written the definitive book on Inference Engineering spotted all over SF:Three years ago, inference engineering barely existed as a category.Today, it is one of the most critical disciplines in AI. Inference engineering inherently tackles a different question than standard model training: “How do you turn those weights from training into a product that is fast, reliable, and affordable at scale?” Focusing on these creates an entirely new optimization problem.In one recent GLM-5.2 experiment, quantizing more of the model actually preserved its benchmark quality while increasing throughput by 20%, because the errors introduced in different layers could cancel each other out.Inference is no longer just the final step after training. It is becoming its own engineering discipline, with its own research problems, infrastructure, and increasingly specialized roles.In this episode, Baseten's Philip Kiely and Ali Taha join swyx and Vibhu to explain what actually happens after a new open model is released and what it takes to turn “we generated a token” into a fast, reliable, production-ready API.We go deep on cache-aware routing, disaggregated prefill and decode, quantization, speculative decoding, KV-cache movement, model parallelism, GPU kernels, and the race to make frontier models up to 10× faster. Philip and Ali explain why inference optimizations can still produce gains of 20%, 100%, or even 200%; how quantization errors can cancel one another out; why identical weights can behave differently across clusters; and how Baseten grafted a Kimi vision encoder onto GLM-5.2 without changing the underlying language model.The conversation then expands beyond LLMs into NVIDIA Dynamo, mega kernels, Rubin, AI-specific chips, local inference, video generation, diffusion versus autoregressive models, and the enormous compute barrier to generating coherent long-form video. Finally, we explore the convergence of training and inference, continual learning through persistent KV cache, and the emerging loop where models help optimize the infrastructure that runs them.We discuss:* What happens when a 200,000-token request enters an inference system* Cache-aware routing and reusing previously computed KV cache* Why prefill and decode are increasingly handled by different GPUs* When dedicated deployments become cheaper and more reliable than shared APIs* How speculative decoding uses a smaller model to accelerate a larger one* Tool calling, structured outputs, and what LLMs actually do* What it takes to support a new open model on day zero* Grafting Kimi's vision encoder onto GLM-5.2* Retrofitting inefficient model layers with components from other architectures* Why models sometimes collapse into repeating the same token* How hardware, kernels, and race conditions create nondeterministic failures* Preserving model fidelity while making inference faster* How quantization errors can cancel each other out* Why inference optimizations still deliver gains of 20%, 100%, and 200%* How optimized serving can make a model up to 10× faster* NVIDIA Dynamo, KV-aware routing, and distributed model serving* Speculative decoding the speculative decoder* Why local AI is about making models less dumb while data-center AI is about making them less slow* Tensor, expert, and pipeline parallelism across GPUs* Hardware-aware model design, auto-tuning, and the case against mega kernels* Rubin and why inference is becoming a systems problem* Whether modern GPUs are evolving into programmable AI ASICs* Why enormous models like Kimi K3 require GB300-class hardware* Why open-source video generation still trails Veo, Kling, and other closed models* The quadratic attention bottleneck behind long-form AI video* Autoregressive video, real-time generation, and compounding quality drift* Why future video systems may combine autoregressive and diffusion architectures* Training for inference and inference for training* Continuous post-training, deployment, evaluation, and improvement loops* How GLM-5.2 helped optimize the kernels serving GLM-5.2 itself* Why faster networking could unlock dramatically faster decoding* Continual learning, KV-cache compaction, and persistent model memoryShow Notes* How to build a day-0 API for Kimi K3* 22580: From GPT2 to Kimi3, ExplainedPhilip Kiely* LinkedIn: https://www.linkedin.com/in/philipkiely* X: https://x.com/philipkiely* Inference Engineering: https://www.baseten.co/inference-engineering/Ali Taha* LinkedIn: https://www.linkedin.com/in/aliestaha/* X: https://x.com/waterloointernTimestamps00:00:00 Introduction and the 200K-Token Prompt00:03:18 Dedicated Deployments, Speculative Decoding, and Tool Calling00:11:26 Launching Production-Ready Open Models00:19:06 Model Retrofits, Failure Modes, and Nondeterminism00:28:22 Quantization and Canceling Errors00:32:15 The Race to 10× Faster Inference00:40:48 Dynamo, Speculation, and Local vs. Data-Center AI00:50:18 Model Parallelism, Auto-Tuning, and Mega Kernels01:00:55 Rubin, GPUs vs. ASICs, and Custom AI Chips01:10:03 Giant Models and the Limits of GPU Memory01:12:42 AI Video, Quadratic Attention, and Autoregressive Generation01:21:47 Audio, Images, and Diffusion Models01:27:32 Training, Self-Optimizing Models, and Continual Learning01:40:06 Closing ThoughtsTranscriptIntroduction: Baseten, Waterloo Intern, and Inference EngineeringSwyx [00:00:00]: Okay, we're here in the studio with Philip, old friend from Inference Engineering, the book, as well as Baseten and everything that you've done, you and I have done before, as well as Ali. Welcome.Ali [00:00:15]: Pleasure to meet you.Swyx [00:00:15]: Waterloo intern.Ali [00:00:16]: Waterloo intern, always.Swyx [00:00:17]: When did you get “Waterloo intern” as a handle?Ali [00:00:19]: As a handle? Oh.Ali [00:00:20]: I think the rebranding happened mid-March. When I saw it was open, I was like, “I have to take it. Up for grabs.”Philip [00:00:26]: The problem is that Ali is really good at his job and is not gonna be an intern much longer.Philip [00:00:30]: So we have to figure out who's gonna get the handle.Ali [00:00:33]: Well, I'll pass the torch over to the next intern.Swyx [00:00:34]: Oh, okay. It can be, like, you just pass it to another Waterloo grad.Ali [00:00:37]: To another Waterloo intern. No, bruh.Philip [00:00:39]: Yeah.Ali [00:00:39]: Intern.Swyx [00:00:40]: Intern, yeah.Ali [00:00:40]: And no.Philip [00:00:41]: You gotta get an intern from Waterloo.Ali [00:00:42]: Yeah, I've gotta get an intern from Waterloo.Swyx [00:00:44]: Right.Ali [00:00:44]: But they have to follow the path.Swyx [00:00:45]: Oh, it could, but it could come from Baseten, so it's like whoever Baseten gets from Waterloo.Ali [00:00:48]: Right.Swyx [00:00:49]: Has the title of Waterloo.Ali [00:00:50]: It stays in the ecosystem.Philip [00:00:51]: Exactly.Ali [00:00:52]: Halfway through the internship, you either get it or you're out.Philip [00:00:55]: You should also do, like, a big graduation ceremony where you change the handle.Ali [00:00:59]: Just say it.Philip [00:00:59]: For everybody.Swyx [00:01:00]: You guys are good at ceremonies, clearly. We had a nice launch of the book, very successful. But before we get into all that, I wanna start off with a fun question for you. Okay, you're an expert inference engineer. What happens when I send a long query, say two hundred thousand tokens into Baseten's inference? What's the process of query through GPU model routing, balancing, all that? What is all the stuff that we don't think about?Long Context Requests, KV Cache, and Cache-Aware RoutingPhilip [00:01:26]: With a long query specifically, the first thing that I'm gonna ask is, “Have you sent me this query before, or at least part of it?” and I really hope you have, because it's gonna be a lot easier for me and a lot cheaper for you. So the first thing that we're gonna look at is some cache-aware routing, where we're going to see, we probably have a number of instances, a number of replicas up serving whatever model you're hitting. We want to send this one to something with, number one, available prefill workers, and number two, ideally some cached input already there so that we can skip prefill on at least part of these two hundred thousand tokens. If you're doing two hundred thousand tokens, it's probably coding or a multi-turn agent or something where you would expect to have that cached. If you don't, we're gonna have to send it to a prefill worker. We've at least on certain models disaggregated prefill and decode, so you're going to have one set of GPUs that's solely going to process the input, create the KV cache, and get you your first token, and then that's going to be passed over to a separate set of GPUs, which is going to run decode. We're going to iteratively make those tokens. We're probably going to have some speculator model in front of that. I'm going to assume that you're doing coding, and because of that, our speculator model, which assumes you're doing coding, is gonna have a high draft token acceptance rate. If I'm wrong and you're asking me to summarize every Harry Potter book, it's gonna be slower. And then we stream that output to you and account for it, charge you, a couple of pennies and say, “Hey, would you like to send another one?”Swyx [00:03:04]: Except Baseten doesn't charge by pennies.Philip [00:03:07]: Well, yeah, we charge. I'm assuming that we're talking about the public model APIs. If you are setting up a dedicated deployment, then yeah, it's not pennies.Public APIs vs. Dedicated DeploymentsSwyx [00:03:18]: Yeah, one of the key differentiators when I was talking with Baseten initially was that people who want very high volume just need to rent by the box, ‘cause then it's up to you to figure out how to saturate the box.Ali [00:03:31]: And more often than not, it's, like, way cheaper if you're pushing, like, millions of tokens per hour, if you just pay per hour instead of pay per token.Philip [00:03:37]: Yeah, they do. I think that we've increasingly seen a lot of demand for the pay per token APIs, just because everyone wants to try open models, and then once they find a use case that's really sticky, then they move over to dedicated.Swyx [00:03:51]: Is there a best practice on when it's time to swap over?Philip [00:03:54]: Couple reasons. Yeah, reliability, that's a big one, right?Ali [00:03:57]: Like, if they have a very specific use case, they want you to train something specifically for them, like they want their own spec dec, for instance, for their own traffic.Swyx [00:04:04]: Spec dec is speculative decoding.Speculative Decoding and Custom SpeculatorsAli [00:04:05]: Speculative decoding, yeah.Swyx [00:04:07]: You have to explain.Ali [00:04:07]: Sorry. Like, speculative decoding is like, if you have a huge model, right? And so the model is going to be generating one token at a time every single turn, every single forward pass. So we attach, like, this little, like, parasite, like this layer that goes on top of the model, and this model just has to predict. It does three very fast autoregressive forward passes, and it will predict, like, three certain tokens, and then you do one forward stage over the entire original model in order to see if those predictions were correct or not, and then you accept them or you reject them. Now, this draft model is traffic specific, so if you, like, Philip said, if you're summarizing Harry Potter books, I can train exclusively that draft model on Harry Potter books, and I can guarantee you that I'm gonna accept the three tokens every single time. And so with that case, I increase your decode speed. I wouldn't be able to provide this to you if you're a shared endpointSwyx [00:04:53]: YeahAli [00:04:53]: ‘cause I have no idea if you're doing Harry Potter, if you're doing coding, if you're doing English. We don't know. Also, there was a thing in the book that mentioned that if they really cared about a specific threshold, chapter four, I think. Do you remember that?Philip [00:05:06]: Yeah. The things that you can do is you can set a specific, like, batch sizing, a specific, like, parallelism strategy if you're trying to optimize for, like, throughput versus latency. You can. Maybe a NVFP4 quant doesn't pass your benchmarks and you wanna run a model at higher precision, you could do that. There's just a bunch of reasons why you might wanna have your own endpoint and the biggest one, of course, just being, like, you don't have to deal with someone else doing a hundred million tokens of benchmarking traffic at the endpoint when you happen to be trying to serve your users.Swyx [00:05:40]: Yeah. I think one thing that is. That is a classic journey. Like, it's people is asking the, what happens when you type Google into the browser. Tool calling, is that just, you're generating JSON or is there more complication beyond that?Tool Calling, JSON, and Structured OutputsAli [00:05:58]: Certain customers that we have, they have their own post-trained models, and so they demand a tool calling that's not just, like parse a file or go find the weather. It's something that's very specific and you have to do post-training on this. And if the post-training on the model is not good or if the quantization after the post-training to get the inference to be fast, the model will struggle reading the JSON file and reading the tool calling. But it doesn't require its own like sandbox. It's not like it's going to use that tool calling to like escape a sandbox or like it doesn't have to be contained. It can just be a normal dedicated deployment. The challenge with tool calling more and more seems to be that the companies want certain tool calling which is a very sensitive thing to train. And because you're dealing with all of the JSON outputs, if it doesn't like close the end of the request in a very certain manner, you end up with a model that did the tool calling and like the thinking and so as a result of that, it didn't see the result and just hallucinated the result as it decoded. That seems to be the most challenging thing with tool calling, not really the sandboxes model.Philip [00:06:56]: Yeah, that's a challenge on the training side and then on the inference side, there's work that you can do to scope the possible output. So we published this at this point close to two years ago, the solution to this problem which is you make a state machine and you use that to constrain the output to a specific format. So this is the structured output problem. If you remember backSwyx [00:07:27]: Yeah, the specific grammar is,Philip [00:07:29]: Yeah, exactlySwyx [00:07:30]: GML had this thing.Philip [00:07:31]: Yeah. So it's like the old-school “make sure this is only JSON”, return only JSON orSwyx [00:07:38]: YeahPhilip [00:07:38]: Grandma's gonna die type of prompts.Swyx [00:07:39]: Is it BNF grammar? At some point OpenAI had released a thing that was like, yeah, if you want to constrain your output, write BNF grammar, back as NOR.Philip [00:07:47]: In our inference system, it's just a specified output format. And you get the guarantee that your output's gonna be structured along that format. And so applying that to tool calls can like help cut down on. You can still call the wrong tool or call no tool. It doesn't solve the certainty problem but it at least solves the output structuring problemSwyx [00:08:10]: YeahPhilip [00:08:10]: Within tool calls.Swyx [00:08:12]: And MCP is just another form of tool, right.Philip [00:08:14]: Yeah, exactly.Swyx [00:08:15]: As far as there's no special thing there.Philip [00:08:16]: The thing I'm always like explaining to people is the LLM is not capable of doing anything. It's only capable of making suggestions of what to do and then if those suggestions are formatted in a certain way and applied to a system that knows what to do with them, then an action occurs.Swyx [00:08:32]: Yeah. Part of the fun stuff is, this is solved outside of tool calling too. Like in an agent loop if the output is not correct or you're right, like reasoning, tool calling was done in the reasoning trace, just be like, “Oh, I don't know what to do. Let me just try again.” And it might get there after a few tries. And on your point of training, sometimes this is harder in smaller models, so you don't have the same exact quality outputAli [00:08:56]: Right.Swyx [00:08:57]: When you just swap from a big model, right?Ali [00:08:59]: Yeah. I will say that, before, I think we need to go back to inference engineering proper.Ali [00:09:04]: But, I had expected that something would replace JSON because it's hard to stream JSON ‘cause JSON must be complete and you must have open and close brackets and everything. So it's hard to parse something or validate something while it's being streamed. So people invented all sorts of things that are like, I forget the name of some of these alternatives, but it's something like TOML, something like YAML. But JSON seems to be dominant still.Philip [00:09:30]: The JSON outputs aren't that long, right? Like you could have a long-- ‘cause tool calls also contain the arguments in them and perhaps for a certain tool you might pass like a very long argument. But my impression of the median tool call is that it's a relatively small number of tokens, right? So I would expect that speculators are generally fairly good at something as formatted as JSON. And so you would have like a pretty fast decode step there and that the streaming wouldn't be as valuable, but maybe I'm wrong about that.Ali [00:10:02]: I think you're also bounded by the software or that the model is gonna integrate with if the software is built with JSON for the tool calls or if the company that you'- if your customer says that this is how our software works and our tools are interfaced with JSON, you can ask them to like, change their software and say like, “Yeah, this is gonna be better for the model.” but like with the right training shouldn't be that much of a difference. Also more profitable if it outputs more tokens probably.Swyx [00:10:25]: Depends on your business model.Swyx [00:10:27]: It really depends. But I will say that, as a writer with like experience a lot with generated output, I do try to move from text to JSON text which is very long JSON, right? Like there's paragraphs in every field because I'm trying to structure it, right?Philip [00:10:44]: Right.Swyx [00:10:44]: I want you to first make factual statements, then make opinions then make bullet point summaries, have dates, have entity references have your sources for references, all these things. Anyway, so these are things that like I think people who really experiment with structural output have to really care about. But, let's, let's recurse up the stack a little bit. Before we started recording, you mentioned something really cool, which is that there's a lot of engineering that-- inference engineering that goes on when a new model provider releases a new model, right? So let's call it GLM-5.2, Kimi K3. I had previously assumed, especially if it's like, well, GLM 5 to 5.1 to GLM-5.2, like that you've supported them before. Is it that much work?What It Takes to Support a New Open ModelAli [00:11:26]: It's a lot of work.Swyx [00:11:28]: Yeah. Okay. So like, a lot of people, all you guys, right whenever a new model launch like, people rush to say like, “Oh, Hugging Face supports this, Fireworks supports this, Spacetime supports this,” and I'm like, “Yeah, of course we support it.” But what goes into that? What goes intoPhilip [00:11:40]: I think it's more than just support it too, right? It benefits the consumer a lot. Like I think it was with Kimi K2.5 or GLM-5.2 the latest, there was an inference war, right? X provider is at 90 tokens a second. The next day we're at 150. The nextSwyx [00:11:55]: I kinda kicked that off with the GLM-5.2.Swyx [00:11:58]: I wrote a Twitter article about. It got like half a million views,Ali [00:12:02]: Based on being numberSwyx [00:12:03]: YeahAli [00:12:04]: Or it's for something else.Swyx [00:12:05]: Yeah. Which,Ali [00:12:06]: Oh my GodSwyx [00:12:07]: Which then got everyone really excited about, hey, how can we, bend tracks a little bit further and,Philip [00:12:14]: There's a difference between support the model, as in I can make a token out of this model, and support a model, as in I have a production-ready API from this model.Philip [00:12:26]: Getting to the point of I can make a token out of this model is not that hard because generally the, open source inference engines, vLLM, SGLang of the world oftentimes even receive weights ahead of time, maintainers do, or the people making the model merge PRs to ensure support. So you generally can, just get it working on the standard open source stack without too much pain in most cases. The challenge is, every inference company is gonna have own proprietary stack. Some open source components, some in-house stuff. And for any arbitrary model, there's going to be some new stuff. Sometimes you get lucky, like K, two five to two six was, like, pretty similar.Quantization, Speculators, and Production ReadinessAli [00:13:16]: Yeah. It was pure continued post-trainingPhilip [00:13:18]: YeahAli [00:13:18]: If I remember correctly.Philip [00:13:19]: Even in those cases, there's still stuff you have to do. You have to redo the quantization work. You're taking the model from. Generally, these models are not released in NVFP4, and we want them to be in NVFP4 for maximum Blackwell compatibility. So we have to perform that quantization, and, calibrate the quantization to make sure that we're not causing any regression in the model's intelligence. And then we also have to train the speculator, as we've talked about. Generally, we have. We have ZDR, zero data retention on our model APIs, so we don't know exactly the traffic that people are sending us, but we know what's popular. We know that coding use cases are popular. We know that agents, agentic use cases are popular. So we can get public data sets that are representative of that traffic and train general speculators. Now, with speculators today, you need to train the speculator using the base model itself because you're getting hidden states out of the model from running inference on these specific prompts, and that is the training data you use to create the speculator. So there's that process which you need the real model weights for. And then there's of course just the process of, standing up all the infrastructure behind it, loading all this stuff, testing it. And then when there's a new model with a newer architecture, I think that, like, the DeepSeek models tend to be the most challenging as they have, like, the most novel architectural stuff going on, model after model. But every new model has something. Kimi K2 had. Oh, sorry, GLM-5.2 hadAli [00:14:53]: Sparse attention.Philip [00:14:54]: Yeah,Ali [00:14:54]: YeahPhilip [00:14:54]: the DSA.Ali [00:14:55]: Right. Which is brought from DeepSeek.Philip [00:14:57]: Yeah. AndAli [00:14:59]: So you can copy-paste then?Philip [00:15:01]: It kindAli [00:15:01]: I don't know how this works.Philip [00:15:02]: So, like we had to, like, build support for that into our runtime. And you're right, like it is really interesting the way that all of these open source labs borrow from each other. For example, like GLM-5.2 doesn't have vision. So something that, Haley, a guy on our team, if we could take a look at this, he, like, grafted the Kimi vision encoder onto GLM-5.2.Retrofitting Vision into GLM-5.2Ali [00:15:27]: We'll be training the projector.Philip [00:15:28]: Exactly. So if you think about, like, the encoder, there's the encoder, which is the part that looks at the image and turns it into latent information, and then there's the projector which likeAli [00:15:38]: You can say latent space. It's okay.Philip [00:15:41]: And then there's the projector that maps it onto, the model itself, and then there's the model weights. You don't wanna mess with the model weights because you run a chance of making the model dumber at something else for the purpose of giving it vision. So instead, Haley started with just a projector, which is only a handful of millions of parameters.Ali [00:16:02]: That would be, yeah.Philip [00:16:02]: Yeah.Ali [00:16:03]: Can you show the training one?Ali [00:16:04]: Like the way it groksPhilip [00:16:05]: YeahAli [00:16:06]: Very interesting.Philip [00:16:06]: And maybeAli [00:16:07]: That right therePhilip [00:16:07]: Maybe Ali, you should take it from here. You've got a betterAli [00:16:10]: Ooh, double the sandPhilip [00:16:11]: Understanding of this than I do.Ali [00:16:11]: Yeah. You can see, like, he. The way he trained this is really cool. At the beginning, he was training it using just like, “Here's a picture of a mountain. Can you describe what's in this mountain?” And that caused it just like the first, learning walls. Like here you can see this all we're trying to teach it is to translate the encoded. Like it's already taken the encoder from Kimi K. It's taken the image. It'Philip [00:16:31]: Yeah. FrozenAli [00:16:31]: FrozenPhilip [00:16:32]: With adapter.Ali [00:16:32]: Exactly.Philip [00:16:33]: Yeah.Ali [00:16:33]: So the brain is frozen and the eyes are frozen. It's just we're tryingPhilip [00:16:37]: AlignAli [00:16:38]: Interconnect between the eye and the brain, right? So the projector. And so you take the tokens and then he's like, “Oh, can you describe what's in this image?” And he's like, “Oh, it's a mountain,” or it's a person or it's a human, whatever the case is. But that didn't cause complete understanding. So he changed it such that every image was associated with a data set of questions. Like, does this image have a white male? Does this image have birds in the top corner? Does this image have a scientist in it? All of that stuff. And it would have to answer questions correctly. And using not just training on describing an image, but being able to answer question, another question, answer over time. Like you can see the grokking, which is like genuinely insane, that retrofitting vision into a large LLM can learn to that extent. And even for images that it doesn't perform well on, for instance, if you ask it a picture of like Stephen Hawking, “Who is this?” Maybe it doesn't get it, but it will say something like, “This is Albert Einstein.” Like it still understandsPhilip [00:17:25]: Close enoughAli [00:17:26]: That this is a scientist who is a man who has, some significant achievements, all that stuff. So that's like really cool.Philip [00:17:32]: Yeah. So, we've covered Hao Tian before, who the author of the LLaVA paper that did this, a while ago. And I think that's very foundational work for anyone who hasn't done vision work before.Ali [00:17:41]: Same with the CLIP and MetaCLIP, where you go from just captioning to building out questionsPhilip [00:17:47]: RightAli [00:17:47]: Off the image and how much better you can get performance.Philip [00:17:50]: Right. Right. Right. Yeah. But what's, what's so exciting about this is if you look at a model like this. Now, this is a little bit more of a research project. It's not. It got to 56% on MMLU Pro, I think. So not quite frontier. But if you're running this model, you haven't suffered any loss on your GLM-5.2 quality. If you don't have an image, it'll just behave exactly the way it used to. And ultimatelyAli [00:18:14]: Which in the inference code you literally do not include the other part, right?Philip [00:18:18]: Yeah. You would just skip the encoder if you don't have an image input.Ali [00:18:22]: Okay.Philip [00:18:22]: Just confirming.Philip [00:18:23]: YeahAli [00:18:23]: Does it affect a lot on the overall inference side? Like you're not adding much, you're adding a very small vision encoder. These are typically likePhilip [00:18:30]: They're super fineAli [00:18:31]: Less than a billion parameters, right?Philip [00:18:32]: Yeah. It's, - There's a little bit less standardization among vision encodersSwyx [00:18:37]: YeahPhilip [00:18:37]: So the support matrix can be a little bit, sparser. But overall, yeah, it's a pretty, it's a pretty minor component of the overall system. And ultimately what you get out of the system is all of a sudden you have Kimi Vision, GLM weights, and DeepSeek attention all in one model.Open Source Model Grafting and Franken-MergesPhilip [00:18:56]: And that's, I think, a lot of the power and beauty of open source, is that you can take all of these different components and combine them together into a system that's better than anyoneSwyx [00:19:05]: YeahPhilip [00:19:05]: Can be individually.Swyx [00:19:06]: People used to say that you would also do Franken-merges where you would take likePhilip [00:19:10]: YeahSwyx [00:19:10]: Layers from each model.Swyx [00:19:11]: Does anyone do that anymore?Ali [00:19:13]: Well, to your point previously when you were mentioning like, the work that goes into supporting a model when it first comes out, like GLM-5.2 or MiniMax M3 or whatever the case is. Sometimes you do have to like, you do have to switch out some things. Like, for instance, the MiniMax M3 head uses full attention, and with full attention you end up with this like insane bottleneck in spec dec ‘cause you're doing auto-regressive token generation for three tokens, and you're doing this like N squared over all of the tokens that are in your sequence. Your KV cache is like very large because it's not sparse, it's not top K. So we find it better to like, okay, we're gonna replace this, we're gonna replace this layer with a layer from another model that's using like GQA, for instance. And then just with the right training, you can get it to have the same acceptance rate. So it is very possible to retrofit layers from other models and very much needed. If a layer is like inefficient, the training just becomes the challenge, like how do you ensure that you train it properly? Which again to your earlier point is like the mesh between training and inference. As in like you need very good training in order to do fast inference. That's like, I feel like more and more becoming true.Swyx [00:20:21]: Yeah. Anything else on the support side when you say like get it to fully production ready?Loop Detection, Race Conditions, and Non-DeterminismPhilip [00:20:26]: Yeah. I think that there's also a question of just, we can test a model to a pretty extensive degree, but we're trying to get it out quickly and then you see a bunch of other people test it and you get interesting results. There was an issue with, GLM briefly where we had some like mode collapses where it would just output the same token over and over again for certain prompts on certain temperatures. Like once you expose an endpoint to the real world, there's going to be, so many more varieties of things given to it that you're able to, discover and patch things. So it's not just a, day zero process, it's then like for the first week, for the first month, if a model remains popular, like how do you both fix bugs and then continue to push the envelope on performance?Ali [00:21:21]: What do you mean you don't want your model outputting S?Swyx [00:21:24]: Is there loop detection on that stuff, by the way? It still happens like quite a lot, which is surprising.Ali [00:21:30]: We have like we, in our endpoint, like if a model was to output the same token like four plus times, we just cut the generation. We say like, “Oh, sorry, this-- Like try again,” or like we will reprocess the request. ‘Cause we know then, like if it, like if, yeah, it's four times the same token, it's probably collapsed.Swyx [00:21:45]: Yeah. Is there a way to opt out in case I really want that?Ali [00:21:48]: You want that?Ali [00:21:50]: I think there's a way that we have to handle it. I'm not exactly certain, but I feel like in certain models, like when they output something like you can imagine, like a table for instance, and so they want, they wanna draw like 12 dashes and 12 dashes. Yeah, I think there's a way for that to happen. I think we only do it on certain tokens. Like we exclude certain special characters.Swyx [00:22:07]: Yeah.Ali [00:22:07]: So we only do it on like certain like S is the most common almost. GLM-5.2Swyx [00:22:11]: OhAli [00:22:11]: And I think it was DSV 4 as well. Like you'd just have like looping issues where like you literallySwyx [00:22:17]: ItAli [00:22:17]: Just have like S.Swyx [00:22:18]: Yeah. Is there a special, something special about S? No, just randomlyAli [00:22:21]: It just seems to be the one token involved.Swyx [00:22:23]: Yeah. And it'Philip [00:22:24]: Is thereSwyx [00:22:24]: And it's only temperature 0Ali [00:22:27]: NoSwyx [00:22:27]: Even at other temperaturesAli [00:22:27]: Even at like 0.9 or whatever, it will still, it will still collapse.Swyx [00:22:30]: That's weird, right?Ali [00:22:30]: It's, it is an inference problem to be honest, like a software problem. Like oftentimes, the image you run will-- like NVIDIA will release an image for instance, and if we will upstream the changes from their latest TensorRT-LLM image into our stack, we'll find that it fixes it. Or oftentimes this will only happen in an inference engine that you're using like SGLang. But if you were to switch to vLLM, that isn't the case. So it seems to be like an extremely like deterministic software issue and not really a model issue. It's not like a weights problem. Like I'- we'll say like, “Oh, it's a problem with the quant. We did PTQ wrong,” right? But that isn't, that doesn't make sense because the same weights used with a different inference engine does not repeat the problem. And sometimes it's, the kernels that are being used in the backend have like these very subtle sometimes race conditions, where if you were to use this model hosted on one cluster, you will never get this problem.Swyx [00:23:19]: Oh my God.Ali [00:23:19]: But if you host it on a different cluster, you will. And the reason is the KV cache transfer from a node to node in that one cluster is using a slower interconnect than the node to node in another cluster. So that exposes the race, whereas in another cluster it doesn't. So then you end up just like, okay, this model is not gonna be hosted on this cluster. We're gonna host it on, another cluster because that cluster exposed that problem. But then it ends up with like, okay, is it the software? Is it the model weights or is it the hardware?Swyx [00:23:42]: There is a thing about this with temperature 0 still not being deterministic, right?Ali [00:23:46]: Right.Swyx [00:23:46]: Mostly because of hardware. Even at temperature 0 same model, you won't always get the same output.Swyx [00:23:52]: Even-- But I'm surprised by the race condition one because, I thought PyTorch was a graph that like guarantees that you at least, execute things in the right order.Ali [00:24:02]: Well, yeah, true. Like I'm not, I'm not saying that there is. Like well, you have things like PTL optimizations where like you can start a kernel before the end of the previous kernel, and that's like ‘cause you want to do that because there'sSwyx [00:24:12]: It's like pipeliningAli [00:24:12]: Expense. Exactly.Swyx [00:24:13]: Yeah.Ali [00:24:13]: But it'- But you don't do it cleanly. Like you overlap a little bit of the execution. No, it is very possible that the kernel itself, like that one block that is supposed to be running in this instance of time, that kernel itself has a race condition. For instance, like a missing barrier. Like often if you're designing a kernel and you want it to make it to be very fast, if you don't test it extensively, you'll, you'll have certain threads access data points from registers before they've been written to by other threadsSwyx [00:24:36]: YeahAli [00:24:36]: For example, because like your barrier is wrong or your synchronization was wrong. But yeah, like the testing itself is very difficult in those like, andSwyx [00:24:42]: And there's no like borrow checkerAli [00:24:45]: What does that mean?Swyx [00:24:46]: Like Rust. Like the. If you're trying to have like memory safety It sounds like a comparable problem.Ali [00:24:52]: Well, yes, but you're working in CUDA, right, NVIDIA GPUs. Like- You just need a higher level language like modular Maybe that's what modular is supposed to do. I don't know.Quantization Quality and Vendor FidelityVibhu [00:25:00]: How do you see keeping quality of the model? So you talked about all these steps of, okay, you gotta do quantization, train your own speculative decoderAli [00:25:07]: RightVibhu [00:25:07]: Run on different hardware. Looking at other model providers, okay, you kicked off a inference speed race on the consumer end. What goes into keeping quality the same across them, right? Sure, you can run benchmarksAli [00:25:22]: YeahVibhu [00:25:22]: But, like, how do you determine how much quantization are there standards? What goes intoPhilip [00:25:27]: There's a few things on quality. Most inference optimizations are lossless. KV caching, for example. You are just recomputing or preventing recomputing the same values. Speculation, of course, if a draft token is wrong, it gets rejected. The main lossy optimization is quantization. And that really comes down to, number one, data format, number two, which parts of the model you choose to quantize, which layers, and number three, like doing a lot of calibration on the quantized weights, to ensure that you're preserving all the outliers. There's other tricks that you can do, though. A big one is long context, ‘cause one thing you asked at, right at the beginning is, “Oh, what's gonna happen if I send a 200,000 token request in?” So with a long input sequence, you need to, store a lot more information. You need to process a lot more tokens. And so even if a model has a context of a certain length, you might, as an inference provider, choose to build an API with a shorter context length, and of course a full length one as well. Because if someone doesn't need the full million token context, for example, you can get them better performance. I don't know if that's exactly like quality of the model. The way that I think about quality is to what degree are we faithfully serving the original model? If you think of a golden implementation of a model that performs exactly the way the model is designed to perform, I think of quality as how close are we getting to that, 100% fidelity of the model.Philip [00:27:13]: You can also, of course, think about quality from the training side and how do you push yourself past 100%. But when I think about purely inference optimizations, it's getting faster while staying as close to that 100% fidelity mark as possible. And certainly our standard internally is that, like you should not be able to tell the difference between our API and a, official API. I think Kimi in particular does a good job of vendor benchmarking hereAli [00:27:41]: YesPhilip [00:27:41]: Where they haveAli [00:27:42]: They released an actual vendor benchmark.Philip [00:27:43]: Exactly, yeah.Ali [00:27:44]: ‘Cause they accused, some people, Amazon? There was some provider that was not doing very well on Kimi's benchmark.Philip [00:27:50]: Yeah.Philip [00:27:51]: So, with Reflect we probablyVibhu [00:27:52]: This was a long time ago, right?Philip [00:27:54]: No.Ali [00:27:54]: Yeah, like threeVibhu [00:27:55]: They alsoAli [00:27:55]: Four, five months agoVibhu [00:27:57]: This also happened with, I don't remember which model, but they pulled out quite a few, and then they started a whole chart about this. It might have beenPhilip [00:28:03]: Kimi Vendor Verifier.Ali [00:28:04]: Yeah.Philip [00:28:05]: Yeah.Ali [00:28:05]: Yeah, ‘cause you, ‘cause you'd be pissed, right? Like if you'Philip [00:28:07]: Yeah.Ali [00:28:07]: If like if I'm a consumer and I'm using like Amazon's endpoint for instance, and I've used Kimi and I'm like, “Oh my God, like this is bad,” I'm not gonna say, “Oh, Amazon quantized the model in a bad way.” I'm gonna say, “Oh, Kimi sucks.” Right?Philip [00:28:17]: Yeah.Ali [00:28:17]: So it seems like that makes sense.Philip [00:28:19]: Yeah, they care. They care.Vibhu [00:28:21]: Justifiably.Ali [00:28:21]: Yeah, justifiably.Vibhu [00:28:22]: This is probably a stupid question, but just checking, has anything improved from main quantization?Philip [00:28:28]: Yeah.Vibhu [00:28:28]: Like, is quantization always strictly worse?Ali [00:28:30]: Well technicallyVibhu [00:28:32]: NoAli [00:28:32]: It's a lossy. QuantizationPhilip [00:28:33]: YeahAli [00:28:33]: Is a lossy, it's a lossy implementation.Philip [00:28:36]: Speed improvesVibhu [00:28:36]: Speed improves.Ali [00:28:37]: It the number, likeVibhu [00:28:38]: No, I' always look for inverse scaling laws.Philip [00:28:40]: Yeah.Ali [00:28:40]: Yeah.Vibhu [00:28:40]: This is something I learned from Noam Brown, where like things that normally act in one direction sometimes do.Philip [00:28:45]: Well, technically when you run a benchmark, because these models are deterministic, sometimes your,Ali [00:28:52]: YeahPhilip [00:28:52]: NVFP4 quant is like, two basis points higher than yourAli [00:28:56]: No, it's noise. It's noise.Philip [00:28:57]: Yeah, exactly. I'm like, yeah, it's, it's within. That's why I always say within margin of error.Philip [00:29:01]: And I stopped saying that because everyone assumes that what is, well, within some margin of error, we're barely inside of that to the worst, so we're saying. But yeah, sometimes it's just like, gives you a higher output score. But like Ali said, that's noise. To my knowledge, you're not necessarily making the results better. You're just trying to, again, like keep your fidelity as close to 100% to the original model.Layer Selection, KL Divergence, and Better QuantizationAli [00:29:27]: There is, to your point, research that we did on MP. I don't know if you are able to pullPhilip [00:29:31]: YeahAli [00:29:32]: A tweet we did. One of our research interns, Joshua, I think it's a tweet on how we have 20% better quantized GLM-5.2 than NVIDIA. Essentially what we found throughout like this month research is, okay, quantization is a lossy. It's. You're compressing the data from, occupying 16 bits to occupying, four bits, for instance. And so you're losing some information, and you're trying to minimize that. And so when I say that I'm gonna quantize the model, my job becomes how do I find the layers that I can quantize, and how to find the layers to not. For instance, with image models, I don't quantize modulation layers, and I don't quantize out projections because those two are. Like out projection is what you see as the user. Modulation is what the model sees or understands. Right, exactly. And so to his paper, do you have the. It doesn't have the. Yeah. It's a long paper. I don't know if I can findVibhu [00:30:25]: If there's a part to search or it's probably in the thread.Ali [00:30:28]: It's probably in the thread.Vibhu [00:30:29]: Yeah.Ali [00:30:29]: But the long and the short is it is very possible that quantizing more of the model makes the results. Like if I have a model that I quantize layers one, five, and 10, and another model where I only quantize layers one and It is possible that the model in which I quantized more information is going to perform better because the quantization errors have canceled out. And so what Joshua showed in his mathematical proof where he had like a verifier in, is that you can predict which layers are going to have quantization errors that will cancel out with each other, and you choose to quantize those layers. And so the result of doing this mathematical quantization is you end up with a model that's 20% more quantized than another provider, so you get 20% more throughput of it because there's more layers than running an NVFP4, and your quality is better than that other quant because the layers that you chose to quantize have their errors cancel out, like one layer skewed to the right one layer skewed to the left, one layer skewed to the right. Your final logits distribution is more similar to the original distribution of the model, so you have better fidelity. And so the way we proved this was with KL divergence. So instead of just scoring on the benchmarks, we scored the KL divergence between the logit distribution of the quantized model and the logit distribution of the original full precision model, and we showed that with this technique we get. If your probability distribution on the logits which token it wants to select is more of the same as the original model, you're probably gonna end up staying true to the original model. So yeah, so it seems like previously before this, it seemed like the industry was, well, the more you quantize, the worse it's gonna be, ‘cause the more loss you introduce. That's not exactly, not necessarily true. So yeah, doesn't improve it, but can cancel out.Philip [00:31:57]: I think it might be this, but reminds me a good bit about pruning where you can prune off certain layers.Philip [00:32:03]: But very interesting. Didn't know this was a whole paper you guys put out.Ali [00:32:06]: It's. Fun fact, it was originally 72 pages, this paper, and then we decidedPhilip [00:32:11]: WowAli [00:32:11]: We can't tell. We couldn't release it. So it's now 45.Swyx [00:32:15]: Still 39 pages, so very substantive. We talked about evals and all these things and, like what's possible in terms of speedup? Like it's like probably like the numberInference Speedups and BenchmarkingSwyx [00:32:25]: Thing that people do wanna care about, and it's something that you wrote about in your post. Like official API is 70 tokens per second, and you push it up to 90. Is that like a normal thing?Philip [00:32:36]: So what's cool about working in inference, the reason that I think inference is going to be a useful place to do engineering for a long time, is that if you look at highly optimized domains like, say, finance, if you're in finance, you measure how much better you got in basis points. It's like, “Oh, I got five basis points better, like twentieth of 1% better,” that's huge news because everything is so optimized. When we publish optimizations, it's 20%, it's 100% it's 200%. So there's still probably like a lot further to go, honestly. Like you'll, you'll know that inference is pretty much solved when researchers start publishing about how they got 1% faster at something.Swyx [00:33:19]: Which by the way, because I am from the finance background, in the ‘70s, that was the margin at the time. When you did quantitative finance research, you would findAli [00:33:27]: And like 20%, tens of percent.Swyx [00:33:29]: That's. Yes.Philip [00:33:29]: Yeah.Swyx [00:33:30]: And now it'Philip [00:33:31]: Tiny fractionsSwyx [00:33:32]: For those people interested, look up Andrew Lo's paper. He had a really interesting illustration of quant, stat arb, distribution, narrowing down from like those kinds of 20% differences in the ‘70s, down to nothing today, which is very cool.Philip [00:33:48]: Exactly, and we're at the beginning of the same type of thing. Now benchmarking is hard. I think anyone will tell you that, and benchmarking provider speeds is hard because there's so many variables that go into it. What hardware are you using? How much load do you have on the system? What's the exact nature of the prompts and input and output sequence lengths? All that stuff. But overall, when you start stacking these improvements, you're looking at multiples. You can look at it. The most common form, of course, is TPS, tokens per second, which is bad naming by us in the industry, ‘cause there's two tokens per second. There's tokens per second, the throughput number, and the latency number.Ali [00:34:31]: TTMT, yeah.Philip [00:34:32]: Like total tokens per second out of the, out of the GPU as a throughput number. Most people only care about tokens per second as the latency number, which we should call ITL, intertoken latency, but we don't.Philip [00:34:44]: Anyway, so you can imagine a standard API without many optimizations for a 1 trillion parameter model operating somewhere in the 30 to 50 tokens per second range for reasonable traffic profile. And we generally see the goal of, pushing to 10X that. But, not necessarily day zero, but by stacking enough optimizations, if you have, say like four optimizations, each of which doubles performance. Or sorry, three optimizations, each of which doubles performance, then you stack that up, that's an 8X gain. That's the order of magnitude that we're working with in this space. We're trying to make things substantially faster, not just go from like 70 to 90.Swyx [00:35:38]: Are you saying you've. You have done that?Philip [00:35:40]: So let's say you have as a reasonable baseline, 30 or 40 tokens per second. You can achieve 10X that. So like on GLM-5.2, if you run it unquantized, perhaps on H100s even, and you're just using an off-the-shelf inference engine with no particular optimizations, no speculator, nothing extra around like KV routing, no disaggregation, you're, you're probably, yeah, looking at that like 30 to 40. You think that's like a reasonable baseline?Swyx [00:36:12]: Right. Right.Philip [00:36:12]: To get to something like 10X, there's a lot of trade-offs that you're making. If we're running at more like a 300, 400 tokens per second range, you are using the best hardware possible. You have a optimized speculator. You have done all of your quantization work. You are Seeing a pretty high cache hit rate. You are running with a reasonably small batch size and a parallelism configuration that is tuned for latency versus throughput, but it is possible. So the spreads that you see if you, like, go on artificial analysis or you go on OpenRouter and you look at, the worst provider to the best provider, oftentimes can hit that range. 10X is of course very aggressive. It's oftentimes maybe more of a four to six times improvement. But that's the performance that makes us really excited, is when we can get these huge gains, not just go from 70 to 90 tokens.Stacking Optimizations: NVFP4, Speculation, and DisaggregationAli [00:37:19]: It's also, like, hardware dependent. Like, ifPhilip [00:37:20]: YeahAli [00:37:20]: If you have a thing where you're serving it on just, like, a node of H100s and then you throw, like, you shard the model across, like, four nodes of B200s. Like, you can definitely increase the speed with just throwing more hardware at it. Like, normalizing for the same exact hardware and the same number of GPUs.Philip [00:37:35]: Yeah. Then you're looking at, like, a two to 4X improvementAli [00:37:38]: Right. RightPhilip [00:37:38]: Depending on the inference optimizations. So yeah, it's. Some of it's, what's the call, and some of it's who's the driver.Vibhu [00:37:46]: If you break down the two to 4X, say the example is run GLM-5.2Ali [00:37:51]: YeahVibhu [00:37:51]: On B200sAli [00:37:53]: YeahVibhu [00:37:53]: Single node, right? What's, like, the cost trade-off for effort to get, like, the last bit of juice out versus what should people just think of, right?Ali [00:38:01]: Spectre quantization. Yeah.Vibhu [00:38:03]: Spectre quantization.Ali [00:38:04]: That's, that's, that's like 95%. LikeVibhu [00:38:06]: And how far does that get you? And how easy is that for the average person to do? So say right I wanna throw the weights of GLM-5.2 on a node of B200s, how easy is it to find speculative decoder- decoder model or already quantized model? How much work goes into it?Philip [00:38:23]: If you're doing it up front, it's quite a lot of work. If you're doing it today, there's going to be people who have published things that you can just, you can just grab some NVFP4 weights. You can grab a speculator. Yeah, if we're thinking about, like, what are the 2Xs we're stacking, going from, BF16 to NVFP4 is, it's not quite a 2X, right? It's like. I think it's about, like, 30 to 40%, from 16 to 8, and then another 30 to 40% multiplied from, 8 to 4. So that doesn't quite get you a 2X, but, like, roughly a 2X. Speculator, roughly a 2X. Disagg on top of that if you're able to get enough hardware and put enough traffic through it, another roughly a 2X. And then you add in some, double-digit percent increase from having just a better runtime with, the latest kernels and stuff behind it. And that's how it stacks up.Ali [00:39:21]: YeahPhilip [00:39:21]: So building each of those, like, building the, quantized weights is, for someone who really knows what they're doing, hours to days of work. Building the speculator, again, like, hours to days of work. And the, disagg setup, hours to days. Well okay, but like once you haveAli [00:39:39]: Once set up. Once set up. YeahPhilip [00:39:40]: Yeah, getting disagg working for the first time, I'm saying, of course, is very difficult.Philip [00:39:44]: The marginal implementationAli [00:39:48]: Like, if you're just grabbing, like if you are a person, like just a normal consumer who has access to, like, a node of B200s and you're wondering, “How can I just host it myself?” You don't need to quantize the model yourself. There's always gonna be, like, an open source quantized checkpoint. NVIDIA's gonna push one out if no one else does. You. Usually, the providers will have their own spec dec that they've trained as well. You don't need to train your own spec dec. You can just use that as well.Philip [00:40:09]: Yeah. Like, GLM-5.2 has its own MTP.Ali [00:40:13]: Right. Right.Vibhu [00:40:14]: What's multi token prediction?Philip [00:40:15]: Yes.Ali [00:40:16]: I'm justVibhu [00:40:16]: Can you explain that?Ali [00:40:16]: I'm just an expert.Ali [00:40:18]: I can do it for you in case I get it wrong?Vibhu [00:40:20]: No.Vibhu [00:40:21]: Yeah, you should correct if we're wrong, but their multi-token prediction can be used for self-speculative decoding.Ali [00:40:27]: I'm not sure. I'm not gonna correct that.Vibhu [00:40:28]: Okay. I'm semi-confident in thatAli [00:40:30]: Okay. YeahVibhu [00:40:30]: But someone can check. But it's useful to paint the story of, okay, not just the average person, but say a company wants to switch from serverless inference I wanna throw this up on. I wanna rent some GPUs, throw it up. These are the steps you take to do significantly faster than just put it behind vLLM.Ali [00:40:48]: Right.Vibhu [00:40:49]: I was waiting for a mention of Dynamo.Vibhu [00:40:51]: I feel like, that's supposed to be the baseline that you measure against.Dynamo, KV Routing, and Disaggregation ToolkitsPhilip [00:40:55]: I would think of Dynamo as less of a box system and more of a toolkit for building with. So when we talk about doing aware routing, when we talk about doing KV offloading, when we talk about doing, PD disaggregation, Dynamo fundamentally is. By the way, Dynamo is an open source library from NVIDIA.Ali [00:41:17]: We've done a pod with KylePhilip [00:41:18]: OkayAli [00:41:19]: Kyle Cranin.Philip [00:41:19]: Cool. So then your listeners know then that it supports all the different inference frameworks. And it is multi hardware, which is interesting.Ali [00:41:28]: But it's just a router, it's not like an optimizer layer.Philip [00:41:30]: Yeah. All it does, like, what Dynamo is good at, it is a library for moving information around your cluster, around your hardware. So if you have, KV cache on one place and you need it to be somewhere else, Dynamo coordinates NIXL for you to move that around.Philip [00:41:49]: That doesn't mean that, like, out of the box, you just say, “Pip install Dynamo,” and then you get, like, a massive performance speed up. It's more of a developer toolkit.Ali [00:42:01]: Yeah. I would have said it would. It comes with a set of defaults that you can then swap out.Philip [00:42:06]: It does. If the industry at large, I think, was, like, rolling out all of these deployments, standard, then I think it would be, like, a credible baseline. But, we've got to, we've got to benchmark against, like, what we're seeing in the wild.Speculative Decoding Methods: Medusa, EAGLE, n-Gram, and Spec-SpecVibhu [00:42:23]: I did wanna talk a little bit more about PD disagg, because that is probably, like, number three after quantized and speculative decoding. In your book though, I was just gonna pull out the book.Philip [00:42:31]: Yeah.Vibhu [00:42:32]: Like section 522 on Medusa, 523 on EAGLEPhilip [00:42:35]: YeahVibhu [00:42:36]: 524 on gram.Philip [00:42:37]: It's 55, would be disaggregationAli [00:42:42]: Yeah. Well, no, I just wanted to dwell a little bitPhilip [00:42:44]: YeahAli [00:42:44]: The other. Like, so what do you choose to include? What do you choose to not to include? Because there was all these other techniques.Philip [00:42:51]: Yeah.Ali [00:42:51]: Are these still relevant? Because I think they came out, like, a year and a half ago maybe.Vibhu [00:42:55]: Medusa is quite old.Philip [00:42:56]: Yeah, Medusa's old.Ali [00:42:58]: It was old.Vibhu [00:42:58]: But is it in the book as a good, here'sPhilip [00:43:01]: BaselineVibhu [00:43:01]: Baseline vanilla understand it?Philip [00:43:02]: Like you should know this.Vibhu [00:43:03]: Like I read the paper, I'm like, “ it makes so much sense.”Philip [00:43:05]: Yeah.Philip [00:43:05]: So with the book, I had a couple goals. One was to give people just a working vocabulary for the space as a whole, and the other was to give them some intuition about how each of these techniques works. As I mentioned in my AI Engineer talk, which is the first public addendum to this, the speculation space has moved much faster than everything else. So yeah, even at the time that I wrote the book Medusa, I very much included as a way for people to understand how the space evolved rather than what the most modern technique is. And now of course, there's DFlash, dSpark. There's, there's newer techniques even than EAGLE, although EAGLE is still very commonly used.Ali [00:43:51]: SpecSpecta.Philip [00:43:52]: Yes. Speculative decoding.Vibhu [00:43:54]: What canAli [00:43:56]: Oh, it's a paper by Tri Dao and it's like, it's doing speculative decodingVibhu [00:44:00]: HuhAli [00:44:01]: For the speculative decoder.Philip [00:44:02]: Oh, in spec- oh my God.Ali [00:44:02]: It's literally just an another. It's like, yeah, that's the most simple way to explain it, and it seems like he got trivial speed ups there. But it seems that the complexity with training, it's almost like in our mind at least, it's almost as complex as training GANs. Like it's like a very delicate balance and oftentimes you, it's just but yeah, it's literally speculative decoding on speculative decoding.Vibhu [00:44:21]: Speculative.Ali [00:44:22]: Yeah. We saw this paper.Vibhu [00:44:24]: It's interesting, right?Ali [00:44:24]: Yeah.Vibhu [00:44:24]: I wouldn't even expect it to be very particular to train, I wouldAli [00:44:29]: Right.Vibhu [00:44:29]: The naive part of me is like, okay, train speculative decoder.Ali [00:44:32]: But like, and it makes sense, like the whole idea of speculative decoding is you. It's like, it's like almost like the iPhone auto predict version but for a normal model, right? Like you're just, you're just, generating three tokens and you're like, okay, I'll do prefill on them. And so you save those three turns for your original model. Now your speculative decoder is doing three turns of auto regression, so why not just have an even smaller model?Ali [00:44:53]: The other question there is what are the size of speculators? So say forPhilip [00:44:58]: Right. It's like a billion parameters.Ali [00:45:01]: Like for MiniMax, it's. Yeah. It's like one layer. It's like one 60th of the original model usually.Philip [00:45:06]: Yeah. I think we should do a paper when we get back to the office.Philip [00:45:10]: SpeculativeAli [00:45:11]: SpeculativePhilip [00:45:11]: Decoding.Ali [00:45:13]: No, it's, it does seem like how, when do you stop? But then it also seems like if you're able to train spec-spec decode for instance, right? Like if you're able to have a small model that is accurately predicts what the intermediate speculator is gonna predict, that is able to predict what the original target model's gonna predict, then why not just use that smallest model directly, right?Vibhu [00:45:34]: Yeah. This isAli [00:45:35]: Like it seems likeVibhu [00:45:35]: Adjacent to the routing problem.Ali [00:45:36]: Right.Vibhu [00:45:36]: Yeah.Ali [00:45:36]: Right.Philip [00:45:37]: The thing with speculators is one of the practical constraints on using them is that you do have to run a small model on the same hardware that you're running the big model on. There is a orchestration and resource competition problem inherent in that, and that is one of the constraints on speculation in general, is that draft tokens cost resources to create and cost software complexity to manage. And so if you have like infinitely recursive speculators, you add in quite a bit of that complexity on the actual implementation within the inference engine as well, not just in the training process.Vibhu [00:46:17]: I was gonna say, I would wonder if you could do similar, like distillation and pruning of, it's the same thing, it's just a model. Can we not just distill a lot of the weights, quantize the speculator, out of my domain? The question that also comes up is, this is all for big server workloads, right? How much of this applies to, say I have this MacBook, I wanna run Gemma really efficiently. Similar problems, not the same?Local AI vs. Data Center InferencePhilip [00:46:45]: Pretty different. I talked to Selo, about this on his podcast a couple weeks ago. The difference between inference engineering for the data center and for production workloads versus inference engineering for local AI, is that we start with fundamentally like different constraints and different goals. With local AI, it's how do I fit this model onto my hardware and then make it less dumb? And with data center influence, it's how do I load this model and then make it less slow? And we care about less dumb, and they care about less slow. But the local AI inference engineering ecosystem, I think has a lot for us to learn from in the data center space. They are experts in various forms of quantization, including dynamic quantization that we just don't touch, in the pruning, in the distillation, in the, layer removal. There'Ali [00:47:42]: Layer removal matters less.Philip [00:47:43]: Yeah. There'Ali [00:47:44]: No one loves pruning really.Philip [00:47:45]: Yeah. Well, but the, but they doVibhu [00:47:46]: Which is surprising, right? But that's, that's a whole different thingPhilip [00:47:48]: Just to fit something on the laptop.Ali [00:47:50]: Right.Philip [00:47:50]: So yeah, it's a, it's an interesting, it's an interesting space. Not necessarily that like their techniques make sense for us to do in the data center, because we have different resources and different goals, but more that the process as well as the openness of that field is something to, admire.Ali [00:48:12]: Yeah. Like to your point, like, certain optimizations that would. Like for instance, Turbo Quantum Sharper, like it made such huge hype on that and we did like a whole deep dive on Twitter and like said, what is it? How does it work? Why is it good or not? And it took off and it was implemented on local devices because your memory bandwidth is so slow on like a MacBook, for instance. But try putting the same thing on like an NVIDIA GPU on a B200 Turbo quant would not be. Like, it would not be used. Like, NVIDIA - Like, NVIDIA made it clear that this is not a good optimization, and we've seen it firsthand where the overhead of doing dequantization, quantization of, in the kernel itself with turbo quant kernel, each end is much slower than the time that you save from doing the bandwidth. ‘Cause on the B200s, you have like 3.5 terabytes per second. You don't need decrease the storage that much. You don't need to do, FP4 KV cache. You don't need to use a requant. There's, there's, there's better optimizations to be made. But on Edge devices, it's extremely important, it's extremely useful. So, seems to be, like, different optimizations there, but then they're all uniquely combined with like all you wanna quantize the model, you wanna do speculative decoding, like certain common prefixes with bothPhilip [00:49:18]: Principles.Ali [00:49:19]: Yeah, exactly. Exactly. Exactly.Philip [00:49:20]: They also do a lot of work on, model parallelism, especially over, heterogeneous topology, where you have, some sparks and they are wired together with, Ethernet, DGX sparks.Ali [00:49:35]: Yeah, this is the Exo Labs guys.Philip [00:49:36]: Yeah. You have, a nu

SURPRISE!! Schedules and stars aligned...and you get a new episode of MTP after all! :) And...it's a BIG one! A few months ago, a little horror movie made for 750,000 dollars came to theaters and took the world of horror by storm! It blew audiences away and made a killing at the box office! It's being viewed as one of the best horror movies of the last few years! Now...Kevin and Roy are here to give their thoughts on the new movie that took us all by surprise this year...Obsession! What did Kevin and Roy think of this movie? Find out now on episode 179 of The Midnight Terrors Podcast!!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

The JournalFeed podcast for the week of July 27-31, 2026.These are summaries from just 2 of the 5 articles we cover every week! For access to more, please visit JournalFeed.org for details about becoming a member.Monday's Spoon Feed:In this single-center study, advanced imaging utilization increased significantly following the deployment of a Provider-in-Triage (PIT) model. In addition, the rate of negative CT scans in patients with abdominal pain was higher in PIT-exposed patients.Tuesday's Spoon Feed:In this retrospective study, children with reliable physical exam and normal plain T- or L-spine x-rays had 1 missed injury out of 126 children, and 100% had no operative injury.Wednesday's Spoon Feed:A clinical decision support alert shortened the time to stress-dose hydrocortisone administration for patients with known adrenal insufficiency in the emergency department.Thursday's Spoon Feed:FFP-first MTP may have a slight mortality benefit, but this is not a practice-changing study.Friday's Spoon Feed:A structured approach to cervical spine clearance using clinical decision rules, modern CT, and select use of MRI and x-ray in adult and pediatric patients allows us to avoid prolonged collar use, reduce ED crowding, and limit adverse collar effects for patients.

Série Som do Céu 2026: Vozes e Canções - Paulinho: união de fé e risos Apresentação: André Castilho - Diretor de Produção de Conteúdo da RTM Brasil Convidado: Paulo César de Sousa, mais conhecido como Paulinho Palhaço. Pastor e fundador do Ministério Terra dos Palhaços - MTP.Instagram: @paulinhompc O riso pode ser uma ponte para a alma? No Som do Céu, Paulinho Palhaço prova que sim! De hospitais a presídios, esse comunicador criativo transforma o humor em uma ferramenta poderosa de fé e cuidado. Descubra como o nariz vermelho abre portas para conversas profundas, levando esperança onde há dor e espiritualidade onde o riso parece impossível. Entre histórias emocionantes e reflexões sobre os limites da arte na igreja, Paulinho nos convida a enxergar o humor como um presente de Deus para restaurar vidas. Emocione-se com o poder transformador de um sorriso! See omnystudio.com/listener for privacy information.

It's been a hot minute since we've had a musical guest on the show for an interview!! This week, we are extremely excited to welcome some new friends of ours to the show! Kevin discovered this band through a Facebook post in a music group...and he's become a huge fan of them ever since! Now...they are making their debut on MTP! This week, please join us in welcoming to the show...horror rock band from South Florida...The Scream Queens! Kevin sits down with vocalist, Berlin, and Bassist, Shoey, to discuss the roots of the band, the influence of the horror genre on their music, where the band is headed in the future, and so much horror talk! Thank you to The Scream Queens for joining the show this week! Everybody be sure to go check out their album, "Midnight at the Graveyard" on streaming and support the band by purchasing the album!All rights to the song used in this episode, "Midnight at the Graveyard", go exclusively to The Scream Queens and its members. We do not own the rights to the song used.Check out The Scream Queens:The Scream QueensCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

And finally, for your third new episode of MTP this week...Kevin and Roy are back with another special guest...horror author, Holly Knightley! Holly is celebrating the release of her new book, Cadbury House, and is here to talk about some Stephen King horror with us! This week, tune in as Kevin, Roy, and Holly sit down to discuss Holly's pick of 2004's Stephen King based movie, Secret Window! What did your co-hosts think of this movie? Find out now on episode 178 of The Midnight Terrors Podcast!Check out Holly's new book on Amazon:Amazon.com: Cadbury House: A Riveting Haunted House Mystery Thriller (Knight Time Houses): 9781958761793: Knightley, Holly: BooksCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

No new episode of MTP next week due to schedule craziness...so here's several new episodes THIS weekend!! First up, it's the return of the Spooky Bestie Series as Kevin and Jules return with a new movie pick! This time, Kevin has chosen one of the movies that serves as the reason that he is a fan of horror...and he is ecstatic to discuss it with his Spooky Bestie, Jules! This week...Kevin and Jules sit down to discuss the 2004 American adaptation of Ju-On...The Grudge! Check out Charleston's Absent Friends:Charleston's Absent Friends | CAFCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Midnight Terrors is back with an all-new episode! And...it's the introduction of a brand new guest and friend of the podcast! We are ecstatic to welcome to MTP for the first time...independent horror author, Alex Grass! In celebration of Alex's new book, Infernal Tramps: Tales of Weird Horror, being released on July 15th...Kevin, Roy, and Alex all sit down as a trio to discuss Alex's movie pick! To coincide with Alex's weird horror tales...we tackle Alex's choice of weird movie...2014's Tusk! What did your co-hosts think of this movie? Find out now on episode 176 of The Midnight Terrors Podcast!Amazon.com: Infernal Tramps: Tales of Weird Terror eBook : Grass, Alex: Kindle StoreCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

It's Sunday...which means it's time for a weekly dose of Midnight Terrors! MTP is back with an all-new episode featuring our very good friend and extremely talented horror author...Viggy Parr Hampton! To coincide with the upcoming release of Viggy's new book, "Ripped Up The Middle In Two", on July 29th...Kevin, Roy, and Viggy sit down to tackle some dark fairy tale talk! Per a recommendation from Kevin...the trio sits down to discuss 1995's Rumplestiltskin! What did your co-hosts think of this outrageous movie? Find out now on episode 175 of The Midnight Terrors Podcast!Pre-order Viggy's new book on Amazon:Amazon.com: Ripped Up the Middle in Two: A Horror Novel eBook : Hampton, Viggy Parr: Kindle StoreCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Midnight Terrors is back with our weekly episode! It's 4th of July weekend...so we decided to get a little festive with this week's episode! As it's 4th of July weekend AND the 4 year anniversary weekend of Midnight Terrors being a podcast...Kevin and Roy decided to sit down to discuss the ever so outrageous dark comedy...Uncle Sam from 1996! What did your co-hosts think of this movie? Find out now on episode 174 of The Midnight Terrors Podcast!Thank you all for 4 years of MTP! 4 down...many to go!! Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Midnight Terrors is back with an all-new episode! This week, it's your monthly dose of The Spooky Bestie Series! To top off our epic celebration of Pride Month, Jules and Kevin are back with an all-new installment of The Spooky Besties Series to discuss a highly anticipated movie pick...Fear Street Part One: 1994! What did your co-hosts think of this movie? Find out now on episode 173 of The Midnight Terrors Podcast!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | LinktreeCheck out Charleston's Absent Friends:Charleston's Absent Friends | CAF

Midnight Terrors is back with our weekly episode! This week, it circles back to Kevin's pick with 2006's The Gravedancers! After discovering a copy of this movie at a horror convention and seeing it 7 years ago, Kevin thought it was about time to revisit it and introduce it to Roy! What did your co-hosts think of this movie? Find out now on episode 172 of The Midnight Terrors Podcast!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

For the next new episode of the day...it's the return of our good friend, Eric Cota! Back in January, Kevin and Eric ventured out to the theater to see none other than 28 Years Later: The Bone Temple! It seemed like this movie was a critically well-received box office bomb. However, it seems to have found a second life on streaming! This week, Kevin, Roy, and Eric all sat down to discuss this incredibly unique entry into the zombie subgenre! Do they recommend you check out Bone Temple? Find out now on episode 170 of The Midnight Terrors Podcast!!Check out Eric's production company, Video Film Productions, on Instagram:VideoFilm Productions LLC (@videofilmproductions) • Instagram photos and videosCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

For our last episode of the week, it's a highly anticipated one! A little while back, we were lucky enough to cross paths with Charleston local filmmaker, Daunte Brown, who is hard at work on an upcoming horror movie called The Boonkey Trail! To celebrate the upcoming film, we have had the pleasure of sitting down with our new friend Daunte to discuss a movie pick of his as well as discuss the production of The Boonkey Trail! Tune in as Daunte, Kevin, and Roy sit down to discuss Daunte's pick of 2001's Japanese horror movie, Pulse! What did your co-hosts think of this movie? Find out now and tune into our interview with Daunte about all things The Boonkey Trail! Thank you to Daunte for joining us and be sure to check out the instagram for The Boonkey Trail!Check out The Boonkey Trail on Instagram:(@theboonkeytrail) • Instagram photos and videosCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Midnight Terrors is back with a ton of new episodes today! First up, tune in as your dynamic duo, Kevin and Roy, sit down to discuss some Stephen King via Kevin's pick of the 2024 film adaptation of Salem's Lot! What did your co-hosts think of this version of Stephen King's classic Vampire story? Find out now on episode 168 of The Midnight Terrors Podcast!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Second new episode of the day is from your dynamic duo once again! Tune into this next episode of the show as Kevin and Roy sit down to discuss 2016's sleeper hit from Fede Alvarez...Don't Breathe! Does this movie hold up 10 years later?? Find out now on episode 169 of The Midnight Terrors Podcast!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Next up...2 live episodes recorded from Frothy Beard! Last weekend, MTP celebrated Pride with a live event at Frothy Beard Brewing Company in West Ashley! For the first live episode in front of a great audience...MTP presented the first round of a new program that we call...MTPride Amplified! As it is currently Pride Month...we called upon our audience to send in their favorite picks of Queer Representation in Horror! We then read the results anonymously to our audience to spread some love to Queer Representation in the horror genre! Your co-hosts, Kevin and Jules, also sat down to discuss their individual list of 5 movie picks that they each watched for Pride! Thank you to everyone who participated in MTPride Amplified Year 1! Our microphone is your microphone! Tune into this great episode from our live podcast event, MTPride: Volume 2, from Frothy Beard Brewing Company!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Trump storms out of MTP, Zig stars on CNN, Scott Pelley's kamikaze mission, college football outrage and more.Become a supporter of this podcast: https://www.spreaker.com/podcast/the-death-of-journalism--5691723/support.

"Jack Riccardi talks about CA elections being just the latest to strain, Trump walks on MTP, creepy white male dems and GI Joe Pelley."

For our second all-new episode this week...the Spooky Bestie Series is back this month with our next installment! This month, Jules and Kevin team up to discuss Kevin's pick of a movie that is long overdue on MTP...1988's Elvira: Mistress of the Dark! What did your co-hosts think of this movie? Find out now on episode 167 of The Midnight Terrors Podcast!!Grab your tickets to Charleston's Absent Friends' production of 5 Lesbians Eating A Quiche:Tickets to Piccolo Spoletto 5 Les:PS 26 - 5 Lesbians Eating a Quiche

Midnight Terrors is back with two all-new episodes this weekend! First up, tune in as Kevin and Roy sit down for a discussion of Roy's pick of 2010's Legion! Continuing with the 2010's horror that they touched on with Devil...Legion felt like a logical next step! Does this movie hold up 16 years later? Find out now on episode 166 of The Midnight Terrors Podcast!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Last weekend, Midnight Terrors was back at AtomaCon for a weekend of live events and convention fun! During our time there, we linked up with our good buddy, The Carolina Storyteller, for some talk about The Conjuring 2 and the true story behind it! For part 1, check out The Carolina Storyteller Podcast for our discussion about the true events behind the film! May we present to you part 2 of our crossover with The Carolina Storyteller in which compare the movie, The Conjuring 2, to its true events! Check out part 1 of this crossover with The Carolina Storyteller:https://open.spotify.com/episode/0mF4mUfVvXfBbST8ZNW6RQ?si=4d64602aff6e4ae4Check out The Carolina Storyteller on Patreon:The Carolina Storyteller — Creator of Valmont, Red Coats, Through The Flames and more. | PatreonCheck out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Finally, the third new episode of the week is Kevin's pick of 2010's M. Night Shyamalan written masterpiece...Devil! Check out Roy's book on Amazon:Books for the Broken: An Anthology of Horror: Honeybrook, R. Jacob, Honeybrook, R. Jacob: 9798248717354: Amazon.com: BooksCheck out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

For our second episode of this week, it's a collective choice of a movie from the Nightmare on Elm Street franchise! Tune in as Kevin and Roy team up to discuss the ever crazy and bizarre movie that is Freddy's Dead: The Final Nightmare!Check out Roy's book on Amazon:Books for the Broken: An Anthology of Horror: Honeybrook, R. Jacob, Honeybrook, R. Jacob: 9798248717354: Amazon.com: BooksCheck out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Midnight Terrors is back with 3 all-new episodes this week! First up, Roy decided to torture both himself and Kevin with this random Tubi pick of 2010's...Alice in Murderland! Yep, you read that title right! Is there any way this movie can be good? Is it an entertaining watch? Find out now on episode 163 of The Midnight Terrors Podcast!Check out Roy's book on Amazon:Books for the Broken: An Anthology of Horror: Honeybrook, R. Jacob, Honeybrook, R. Jacob: 9798248717354: Amazon.com: BooksCheck out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast Official: TikTok, Instagram, Facebook | Linktree

Mihai Dascalu has been to 99 countries Hey now, I am your host, Ric Gazarian. In this episode, I had the pleasure of speaking with Mihai Dascalu. I met him at the ETF in Bangkok in 2024, and I am excited he is coming back this October. Mihai and his family left all he knew in Romania to start anew in the US. He and his family went on an epic family journey for over a year. We touch on this and his desire to Chase 193. I would like to thank everyone for their support of Counting Countries, especially my Patrons. You know them, you love them! Bisa "fully nomadic" Myles, Ted Nims, Adam "one-away" Hickman, Steph "Phuket" Rowe, Simen Flotvik Mathisen, Ed Hotchkiss, Barry Hoffner, Philippe "BC" Izedian, Gin Liutkeviciute, Sunir Joshi, Carole Southam, Sonia Zimmermann, Justine, Per Flisberg, Jorge Serpa, Sam Williams, Scott Day, Peter Fenger, Mihai Dascalu, Ryan Knott, Zipping Around The World Podcast, and Shawn McDonough for supporting this podcast. You can support this podcast by going to Patreon.com/CountingCountries. My patrons will hear the entire conversation with Mihai.. Please remember the next Extraordinary Travel Festival will be on October 22-25 in 2026. You can join the event and use the code BANGKOK. Excited to announce a new addition to the ETF, Mike Boisvert known as the Skate Nomad. He plans on skateboarding to every country in the world … what won me over was watching him skateboard with the Skater Girls of Ethiopia. Check out his reels!. Consider joining our Instagram and Facebook groups and signing up for the ETF newsletter. Any questions, please let me know. Some other community notes and updates. I will be speaking at the TCC meeting in Hanoi in May, about … my take-aways from 150 in-depth conversations with the world's most traveled people. 193 Masters looking to connect with each other consider joining a UNESCO tour in Portugal in early June, if so reach out to MTP's Justyna. And also in June is the awesome NomadMania trip to Brazil. I went to their Fergana event and it was quite awesome. And on July 12, come see me in NYC to meet Barry Hoffner, the author of Belonging To The World, for a reading RSVP exploringed@gmail.com I was in Bangkok while Mihai was in New York for this recording. Please listen in and enjoy. Thank you to my Patrons - you rock!! … Bisa Myles, Ted Nims, Adam Hickman, Steph Rowe, Simen Flotvik Mathisen, Ed Hotchkiss, Barry Hoffner, Philippe Izedian, Gin Liutkeviciute, Sunir Joshi, Carole Southam, Sonia Zimmermann, Justine, Per Flisberg, Jorge Serpa, Sam Williams, Scott Day, Peter Fenger, Mihai Dascalu, Ryan Knott, Zipping Around The World Podcast, and Shawn McDonough. Be the first on your block to sport official Counting Countries apparel! And now you can listen to Counting Countries on Spotify! And Alexa! Subscribe on Apple Podcasts today! And write a review! More about Mihai Dascalu Counting Countries: Instagram Facebook Books And check out Thor Pedersen: The Impossible Journey (Amazon US Kindle (affiliate)): https://amzn.to/46pRuDi Other book options: Thor Pedersen | Instagram, Facebook, TikTok | Linktree And Barry Hoffner: Belonging To The World (affiliate) About Counting Countries Counting Countries is the only podcast to bring you the stories from the dedicated few who've spent their lives on the singular quest of traveling to every country in the world. Less people have traveled to every country in the world than have been to outer space. Theme music for this podcast is Demeter's Dance, written, performed, and provided by Mundi. About GlobalGaz Ric Gazarian is the host of Counting Countries. He is the author of three books: Hit The Road: India, 7000 KM To Go, and Photos From Chernobyl. He is the producer of two travel documentaries: Hit The Road: India and Hit The Road: Cambodia. Ric is also on his own quest to visit every country in the world. You can see where he has traveled so far and keep up with his journey at GlobalGaz.com How Many Countries Are There? Well… that depends on who you ask! The United Nations states that there are 193 member states. The British Foreign and Commonwealth office states that there are 226 countries and territories. The Traveler's Century Club states that there are 329 sovereign nations, territories, enclaves, and islands. The Nomad Mania divides the world into 1301 regions. The Most Traveled Person states that there are 1500 unique parts of the world. SISO says there are 3,978 places in the world. And the video that explains it all! Me? My goal is the 193 countries that are recognized by the UN, but I am sure I will visit some other places along the way. An analysis of these lists and who is the best traveled by Kolja Spori. Disclaimer: There are affiliates in this post.Sherri Donovan Counting Countries

For our second new episode of MTP today...it's Kevin's pick...and it's a wild card! After scrolling through Tubi to find a movie that the hosts haven't seen before...the boys have landed on Kevin's pick of 1985's Creature! What did your co-hosts think of this movie? Is this just an Alien rip-off? Or is there more than meets the eye with this movie? Find out now on episode 161 of The Midnight Terrors Podcast!Check out Roy's new book on Amazon:Books for the Broken: An Anthology of Horror: Honeybrook, R. Jacob, Honeybrook, R. Jacob: 9798248717354: Amazon.com: BooksCheck out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

About a month ago, MTP ventured out to Savannah, GA to do a live show with our new friends at Debellation Brewing Co.! For this show, Kevin and Jules got to team up with MTP's great friend, Viggy Parr Hampton, to talk about back luck in horror for St. Patty's Day! Our first live episode was a tier maker episode in which we discuss which horror characters have the worst luck in horror! Thank you to Debellation Brewing Co. for having us and thank you to Viggy for joining us for the show!

It's the return of the Spooky Bestie Series on MTP! This week, MTP co-host (and President of Charleston's Absent Friends) Jules is back on the show for an episode of The Spooky Bestie Series alongside Kevin! And the Spooky Besties are making a grand return with the selection/discussion of the Scary Movie franchise via the original movie that started it all! With Scary Movie 6 coming out very soon, your co-hosts sit down to discuss the movie that started it all and give their thoughts on where the franchise may be heading next!Placing a quick Trigger Warning here as this movie is from 2000 and contains some very extreme/offensive humor.Tune in to episode 162 of The Midnight Terrors Podcast as Kevin and Jules discuss 2000's Scary Movie!Check out Charleston's Absent Friends' production of "5 Lesbians Eating a Quiche" at Piccolo Spoletto: PS 26 - 5 Lesbians Eating a Quiche

Midnight Terrors is back with a few all-new episodes today! First up, we circle back to Roy's pick! You all know how much Roy loves zombies...so we're back with a zombie discussion via 2007's Diary of the Dead! What did your co-hosts think of this movie? Is this an underrated zombie entry? Or one to skip? Find out now on episode 160 of The Midnight Terrors Podcast!Roy's new book, "Books For The Broken: A Horror Anthology" is available now on Amazon:Books for the Broken: An Anthology of Horror: Honeybrook, R. Jacob, Honeybrook, R. Jacob: 9798248717354: Amazon.com: BooksCheck out Roy's Articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

Midnight Terrors is back with an all-new episode!! This week, it's Kevin's pick and he decided to take it back to the year 2007 with the outrageous horror comedy...Murder Party! What did Kevin and Roy think of this movie? Find out now on episode 159 of The Midnight Terrors Podcast!Also, today's episode is in celebration of Roy's new book which just came out today..."Books For The Broken"! All of Roy's stories up to this point in his career are now available in one excellent collection...physical and digital copies! The link to buy Roy's new book is below! Be sure to check out "Books For The Broken" and show Roy a ton of love! Congratulations to Roy from his MTP Family!!Pick up Roy's new book here:Amazon.com: Books for the Broken: A Horror Anthology eBook : Honeybrook, R. Jacob: BooksCheck out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

Midnight Terrors is joined this week by a very good friend of the show who is making his MTP debut! Kevin, Jules, and other members of the MTP Family have been lucky enough to work with this gentleman on an upcoming short film project which will debut in Charleston soon! In the meantime, he is here on the podcast for the first time! Please join us in welcoming our good friend, filmmaker Eric Cota, to the show! For his first episode with us, Eric sits down with Kevin to rank all 7 Scream movies! Each co-host gets to create their own list of the Scream movies...from their favorite to their least favorite! Which movies ended up where? Find out now on episode 158 of The Midnight Terrors Podcast!Check out Eric's film company on Instagram:VideoFilm Productions LLC (@videofilmproductions) • Instagram photos and videosCheck out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

For our second new episode this week...it's another follow-up episode to a previous episode of ours! On episode 150, Midnight Terrors teamed up with our good friend, Viggy Parr Hampton, to discuss House on Haunted Hill (1999)! As a follow-up episode, we welcome Viggy back to discuss a movie that was mentioned on that episode...2001's From Hell! What did your co-hosts think of this movie? Find out now on episode 157 of The Midnight Terrors Podcast!Ps, yes our live show with Viggy (as mentioned in this episode) has already passed...but stay tuned for more events with Viggy to come in Savannah!Check out Viggy's new book:A Veritable Household Pet: A Horror Novel - Kindle edition by Hampton, Viggy Parr. Literature & Fiction Kindle eBooks @ Amazon.com.Check out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

Midnight Terrors is back with some all-new episodes for you all! After a brief break of recording episodes and a live show at a new location in Savannah, MTP is here to drop some episodes recorded over the last month! First up, it's Roy's pick once again! And this episode serves as a follow-up episode to one of our previous episodes! Following up on our Subservience episode featuring Guy Quintero...Roy sits down this week with Kevin to discuss his pick of 2021's Till Death! What did your co-hosts think of this movie? Find out now on episode 156 of The Midnight Terrors Podcast!Check out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

Are you living fake freedom? One of my biggest realizations with productivity expert Dr Ann Tsung was that - many of us are stuck in stage 3 of the time freedom cycle and treading water in fake time freedom. The question then is - how do we escape it and start living a life of more freedom and more flow and what are the habits, the values, the mindset and the tools we can employ to become time transcenders. Ann Tsung MD is a NASA Flight Surgeon, Critical Care/Emergency Medicine physician, peak performance/productivity coach at Productivity MD, real estate investor, podcast show host of Productivity MD, and a mother of a 19-month-old. With a unique fusion of medical precision and entrepreneurial success, she guides top physician entrepreneurs to unlock over 10 hours a week, master their mindset, scale their businesses, and achieve unparalleled freedom and leadership excellence.Contact:Website - https://www.productivitymd.com/Instagram - @anntsungmdJoin us as we explore:The performance masterclass - why it's critical to know our baseline data, liquid biopsies, DEXA, Vo2Max and more.True freedom, freedom's relationship to time and the 5 stages of time freedom, the flow cycle and how to life a high flow state life.Fake time freedom, and the biggest productivity mistake.The mindset and psychology of freedom, why we all need to know MTP and “engineering out distraction”.Habits and practices for improved performance, self-awareness and self-discovery.MentionsLink - 5 Stages to Time Freedom, https://www.skool.com/time-and-energy-mastery-3331/aboutApp - Freedom, https://www.freedom.toSupport the showFollow Steve's socials: Instagram | LinkedIn | YouTube | Facebook | Twitter | TikTokSupport the show on Patreon:As much as we love doing it, there are costs involved and any contribution will allow us to keep going and keep finding the best guests in the world to share their health expertise with you. I'd be grateful and feel so blessed by your support: https://www.patreon.com/MadeToThriveShowSend me a WhatsApp to +27 64 871 0308. Disclaimer: Please see the link for our disclaimer policy for all of our content: https://madetothrive.co.za/terms-and-conditions-and-privacy-policy/

In this episode of HSS Presents, Dr. Matt Conti sits down with Dr. A. Holly Johnson to discuss the rapid progress and clinical benefits of minimally invasive foot and ankle surgery. The conversation explores how percutaneous techniques utilizing specialized burrs lead to smaller incisions, reduced soft tissue damage, and significantly faster recovery times for patients. Dr. Johnson candidly shares her own learning curve and highlights the most impactful minimally invasive procedures in her practice, from calcaneal osteotomies to bunion corrections and first MTP fusions. Balancing enthusiasm with caution, the experts also debate the complexities of the MIS Lapidus and lesser toe procedures, offering practical advice for surgeons looking to incorporate these game-changing techniques into their own practice.

Midnight Terrors is back with our weekly episode...and it's about a movie that's currently available to see in theaters! During MTP's trip to Ohio...Kevin and Roy got to head to the theaters to watch a movie with Reviewed To Death, Creepy Kroly and some of our attendees from our Seventh Son Brewing live show! The group sat down to watch the new horror movie that is in theaters now...Whistle! And now, Kevin and Roy sit down to discuss the movie in depth! SPOILERS EVERYWHERE!! Watch the movie and then give this a listen to hear what your co-hosts think of the movie! What did they think? Is this a worthy entry into the genre this year? Find out now on episode 154 of The Midnight Terrors Podcast!Check out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | LinktreeVote for MTP for Best Local Podcast in Charleston:Best of Charleston 2026

This past weekend, MTP embarked on our tour of Columbus, Ohio live shows! Our second show of the tour was at Seventh Son Brewing in Columbus! The theme for this second show? SLASHERS!! This live event was extra awesome because not only did we Kevin, Roy and Zack get together...but MTP was also joined by the remaining members of our podcast squad...Marcus and Luke of Reviewed To Death! Your 5 co-hosts sat down to do a live show centered around slasher movies in front of the Columbus audience! And man...did this crowd get vocal!! Tune into our first live episode from Seventh Son Brewing as we sit down to play a game of One Must Go with the slasher franchises...aka eliminating one movie from each franchise! Thank you to Seventh Son Brewing for having us! And thank you to our amazing audience who helped make this show extra awesome! See you again soon, Columbus!

Surprise!! Midnight Terrors is putting out this week's episode a little early! :) This week, Kevin and Roy will be doing MTP live and in person together in Columbus, Ohio! In celebration of the events (the first of which is a ghost themed live show), the guys HAD to talk about some ghosts! This week, Kevin and Roy sit down to discuss Kevin's pick of 2011's cult classic ghost movie...Grave Encounters! What did your co-hosts think of this movie? Find out now on episode 153 of The Midnight Terrors Podcast!Check out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

Midnight Terrors is back with another new episode!! This week, we are joined once again by our very good friend and horror author, D.R. Kane! To celebrate the release of D.R.'s new book, we are here to discuss his pick of 2012's deeply unsettling cult classic...Sinister! What did your co-hosts think of this movie? Find out now on episode 152 of The Midnight Terrors Podcast!!Check out D.R. Kane's new book on Amazon:Amazon.com: All That We Destroy: A Paranormal Horror eBook : Kane, D.R.: Kindle StoreCheck out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

We are 150 episodes old!!! Midnight Terrors is here with our 150th episode...and it's a great one! As we celebrate this incredible milestone...we are honored to be joined once again by our very good friend and very talented horror author...Viggy Parr Hampton! A monumental episode such as this deserves a monumental movie to discuss! This week, tune in as we discuss Viggy's choice of movie for episode 150...House on Haunted Hill (1999)! What did your co-hosts think of this movie?? Find out now on episode 150 of The Midnight Terrors Podcast! We also discuss the upcoming release of Viggy's new book..."A Veritable Household Pet" which is set to release on January 28th, 2026! Pre-order link below!!Thank you all for supporting us for 150 episodes!!! Here's to many more years of horror talk together!Pre-order Viggy's new book, "A Veritable Household Pet" here:A Veritable Household Pet - Kindle edition by Hampton, Viggy Parr. Literature & Fiction Kindle eBooks @ Amazon.com.Check out Roy's articles on TBM Horror:Everything Horror!Check out MTP's Linktree:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

It's not just Not Invented Here Syndrome. Some technologies like CSS simply don't scale if you're building some of the largest websites on the planet with thousands of engineers committing to the same code base every day. StyleX is Meta's open-source solution for CSS at scale and allows atomic styling of components while deduplicating definitions for bundle size and exposing a delightfully simple API for developers.  Tune in to learn from Melissa, one of the StyleX maintainers how Open Source has acted as a force multiplier for the project, how interacting with other large companies adopting StyleX has been, and much more! Got feedback? Send it to us on Threads (https://threads.net/@metatechpod), Instagram (https://instagram.com/metatechpod) and don't forget to follow our host Pascal (https://mastodon.social/@passy, https://threads.net/@passy_). Fancy working with us? Check out https://www.metacareers.com/. Links How AI Is Transforming the Adoption of Secure-by-Default Mobile Frameworks: https://engineering.fb.com/2025/12/15/android/how-ai-transforming-secure-by-default-mobile-frameworks-adoption/  StyleX: https://stylexjs.com MTP 67: Measuring Developer Productivity with Diff Authoring Time: https://pca.st/pt4p4tv5  Timestamps Intro and news 0:06 Introduction Melissa 1:47 Why did we build our own styling system? 4:07 StyleX API 5:36 cx vs StyleX 7:37 Component styling and priorities 10:38 How StyleX evolved in the past seven years 15:20 Community influence 19:33 Open Source 24:07 Challenges of OSS 27:02 Managed breaking changes in OSS 29:48 Measuring success for StyleX 32:04 Packaging challenges 34:34 StyleX competition 38:42 Creating the StyleX roadmap 40:24 Outro 43:15

As the 2025 NFL regular season reaches its final weekend, Missing the Point breaks down Week 18 of the NFL season, the finalized playoff picture, and the quarterbacks who defined the year. Hosted by Dave Clarke and Bob Kelly, this episode delivers a full NFL playoff preview along with in-depth 2025 NFL quarterback power rankings based on what actually happened on the field. This is not a preseason projection episode. This is about who earned it. Dave and Bob analyze the top teams heading into the postseason, including the New England Patriots, Chicago Bears, Buffalo Bills, Denver Broncos, Los Angeles Rams, San Francisco 49ers, Seattle Seahawks, Kansas City Chiefs, Baltimore Ravens, and Philadelphia Eagles. They break down matchups, seeding implications, and which teams look built to survive January football. A major focus of the episode is the 2025 NFL quarterback power rankings, where Dave and Bob rank quarterbacks based on full-season performance, not reputation. They evaluate efficiency, consistency, decision-making, durability, supporting cast impact, and how much each quarterback elevated his team. Quarterbacks discussed include Drake Maye of the Patriots, Caleb Williams of the Bears, Josh Allen of the Bills, Lamar Jackson of the Ravens, Brock Purdy of the 49ers, Matthew Stafford of the Rams, Jalen Hurts of the Eagles, Bo Nix of the Broncos, Trevor Lawrence of the Jaguars, and Patrick Mahomes of the Chiefs. Dave explains why some quarterbacks who entered the season with MVP expectations failed to meet the moment, while others quietly established themselves as legitimate franchise leaders. Bob adds context around coaching stability, offensive scheme, and how late-season performance often tells the real story heading into the playoffs. The episode also previews Wild Card Weekend, discussing which teams are dangerous lower seeds, which favorites feel vulnerable, and which quarterbacks are under the most pressure once the postseason begins. The conversation highlights how Week 18 results, recent trends, and injury context shape expectations far more than early-season narratives. Key topics covered in this episode include: NFL Week 18 playoff seeding scenarios 2025 NFL playoff preview and Wild Card matchups Full NFL quarterback power rankings Patriots and Bears postseason expectations Which quarterbacks can win playoff games Teams built for January football Late-season trends that matter in the playoffs If you’re searching for NFL Week 18 analysis, 2025 NFL playoff preview, NFL quarterback rankings, Patriots playoff breakdown, Bears postseason outlook, or who the best quarterbacks in the NFL were this season, this episode delivers detailed, opinion-driven analysis without hot takes or manufactured controversy. Follow Missing the PointWebsite: https://www.mtpshow.comYouTube: https://youtube.com/@MTPPodTwitter (X): https://twitter.com/MTP_podInstagram: https://instagram.com/MTP_podTikTok: https://tiktok.com/@MTP_podFacebook: https://facebook.com/MTPPod

As 2025 comes to an end, MTP is here with our final episode of the year! As we did last year, Kevin and Roy are back to discuss the horror genre across this year! They also come prepared with two top 5 lists for the year...their top 5 favorite episodes of MTP this year AND their top 5 favorite horror films of the year! Tune in for our final episode of 2025 and celebrate the great stuff that's happened in horror this year! THANK YOU ALL for an amazing year!! We can't wait to kick it into higher gear in 2026! Happy New Year from MTP!Check out Roy's new book on Amazon:Amazon.com: Thaddeus Greene's Spooktactular House of Horrors eBook : Honeybrook, R. Jacob, Hoyle, Mary: Kindle StoreCheck out Roy's articles on TBM Horror:Everything Horror!Check out our socials:midnightterrorspodcast | Instagram, Facebook, TikTok | Linktree

Mike Marcangelo and Dave Clarke return with a pivotal late-season episode of Missing the Point as the 2025 NFL season heads into Week 17 and the playoff picture finally comes into full focus. This show centers on two defining Week 16 victories that reshaped expectations across the league, the New England Patriots and the Chicago Bears both delivering comeback wins that solidified their status as legitimate contenders. The episode opens with a deep breakdown of the Bears’ dramatic win over the Green Bay Packers. Mike and Dave discuss Caleb Williams’ continued growth, his ability to respond in high-pressure moments, and why this performance felt like a franchise checkpoint for Chicago. They analyze the late-game execution, DJ Moore’s impact, and Ben Johnson’s offensive philosophy, while debating whether this Bears team is officially ahead of schedule or right on time. From there, the focus shifts to New England’s comeback victory over the Baltimore Ravens. Drake Maye’s performance takes center stage, from navigating early adversity to closing the game with confidence. The guys break down Maye’s decision-making, his chemistry with Stefon Diggs, and why this Patriots team looks more complete than it has at any point since the post-Brady rebuild began. They also examine lingering concerns such as turnovers, pass protection, and whether New England can consistently close out playoff-caliber opponents. Beyond the individual games, this episode zooms out to the broader NFL playoff race. Mike and Dave walk through AFC and NFC seeding scenarios, potential first-round matchups, and which teams are trending up or quietly slipping as January approaches. They debate momentum versus résumé, quarterback ceilings, coaching advantages, and why certain teams feel more dangerous than their records suggest. If you are following the Patriots’ rise, the Bears’ breakthrough season, or the evolving NFL playoff landscape entering Week 17, this episode delivers smart analysis, honest debate, and clear-eyed football conversation when it matters most. Follow Missing the Point:Website: https://www.mtpshow.comYouTube: https://youtube.com/@MTPPodX (Twitter): https://twitter.com/MTP_podInstagram: https://instagram.com/MTP_podTikTok: https://tiktok.com/@MTP_podFacebook: https://facebook.com/MTPPod

Week 16 of the 2025 NFL season is here, and the Missing the Point crew unloads on one of the most chaotic weeks the league has seen in years. Dave, Mike, Bobby, and Rayshawn break down the fallout from a Week 15 that completely reshaped the playoff picture and may have permanently altered multiple franchises. It starts with the injuries. The ACL gods showed no mercy. Patrick Mahomes is done for the season, ending the Chiefs’ playoff hopes in brutal fashion. Micah Parsons goes down, ripping the heart out of Green Bay’s defense. Davante Adams, Devin Neal, and several other key contributors leave games early, forcing contenders to scramble as the season enters its final stretch. The crew debates whether this was bad luck, bad field conditions, or just the inevitable price of a longer and more violent NFL season. Then comes the moment that broke everyone’s brain. Philip Rivers is back. The Colts are handing the offense to a quarterback who retired years ago, and the crew reacts with disbelief, laughter, and real football analysis. Is this desperation or genius. Can Rivers still read defenses. Can he survive a hit. And what does this say about the Colts’ confidence in the rest of their roster. From there, the episode hits the Cowboys’ slow collapse, the Chiefs being mathematically cooked, the Ravens living dangerously on the playoff bubble, and the Panthers, Texans, and Steelers trying to survive December football. The Patriots winning streak came to an end, thanks to a dreadful defenseive preformance in the second half vs the Bills, while the Broncos continue to look like the most complete team in the league. The episode closes with a full breakdown of the Real BK Top 10 Power Rankings heading into Week 16. Texans, Chargers, Jaguars, 49ers, Bears, Seahawks, Patriots, Bills, Rams, and Broncos. Every spot is debated, every rise and fall explained, and no team is safe from criticism. If you want real NFL conversations, sharp humor, and power rankings that actually react to what happens on the field, this episode delivers. Follow Missing the Point:Website: https://www.mtpshow.comYouTube: https://youtube.com/@MTPPodX: https://twitter.com/MTP_podInstagram: https://instagram.com/MTP_podTikTok: https://tiktok.com/@MTP_podFacebook: https://facebook.com/MTPPod

The whole crew is back as Week 15 arrives. Dave, Mike, Bobby, and Rayshawn react to one of the strangest NFL stories in years. The Colts are bringing back Philip Rivers. The crew explains why Indianapolis turned to a quarterback who last played when TikTok was still optional and why this move could change the AFC playoff race. New England stays on top. Ten wins in a row. The Patriots look like a complete team. The crew breaks down Drake Maye’s growth, Vrabel’s command of the roster, and the defense that keeps winning important downs. They also look at how the Patriots match up with the Broncos, Bills, and Rams as the postseason gets closer. Dallas hits another wall. The Lions knock them down again. The crew debates whether the Cowboys have any real chance to sneak into the playoffs or if this season has reached the point of no return. Detroit looks confident, physical, and built for December football. Cross offs expand again as the Jets, Titans, Saints, Raiders, Cardinals, Giants, Commanders, Falcons, and Vikings all fall fully out of the playoff picture. The crew debates which bubble teams are next and which ones still have a small chance to survive. Every team in the Real BK Power Rankings gets a clear breakdown. The Chargers and Jaguars keep pace. The 49ers, Bears, and Packers remain dangerous. The Seahawks, Bills, and Rams climb at the right time. The Broncos keep winning close games. The Patriots stay at number one with a complete roster and real momentum. This episode gives sharp analysis, clear opinions, and quick humor as the league heads into the final stretch of the season. 00:00 The Shocking Return of Philip Rivers to the Indianapolis Colts 08:12 Dallas Cowboys’ Playoff Hopes Dim After Lions Loss 19:20 Uncrossing the Jags and Debating Playoff Bubble Teams 25:13 Bengals, Ravens, and the Future of Tomlin and Harbaugh 39:36 The Houston Texans Emerge as a Surprising Playoff Dark Horse 48:52 Why the Houston Texans Are a Nightmare Matchup for the Patriots 01:01:14 Caleb Williams and Ben Johnson Lead the Chicago Bears’ Resurgence 01:18:25 Jordan Love, Sam Darnold, and the Evolving NFL Quarterback Landscape 01:29:57 AFC Showdown, Patriots vs. Bills and Playoff Implications Stay connected with MTP.YouTube: https://youtube.com/@MTPPodX: https://twitter.com/MTP_podInstagram: https://instagram.com/MTP_podTikTok: https://tiktok.com/@MTP_podFacebook: https://facebook.com/MTPPodWebsite: https://www.mtpshow.com