Podcasts about Spacetime

Curious to explore the wonders of quantum physics?Dive into this fascinating world and explore what makes up subatomic particles, how exactly entanglement works, and learn about the new theories going around the science community gossip mill; like parallel universes, the holographic principle, and the multiverse.“Quantum Design” starts with a quick and lively tour through the history of quantum physics and how we got to where we are today, and continues to unravel the mysteries of Spacetime, Schrödinger's Equation and Wave Mechanics, the famous Double-Slit experiments, and the interesting nature of quantum spin. “Quantum Design” also delves into the future of quantum mechanics, shedding light on the groundbreaking applications emerging from new technologies such as quantum computers and cryptography.In this book, you'll find:- A Quick Tour through Quantum History- Understanding Subatomic Particles- Schrödinger and Wave Mechanics- The Double-Slit Experiment- Photons and Entanglement- Spacetime and Spin- Proton Stability and New Explorations- New Theories on Gravity and Spacetime- Theories on the Holographic Principle, the Multiverse, Parallel Universes, and more- Quantum Applications in Technology- And MORE…Written in an engaging and accessible style, "Quantum Design" is the perfect companion for anyone eager to grasp the core principles of quantum physics.Whether you're a novice or a curious enthusiast, this book offers a fun and informative exploration of the mysterious and wonderful world of quantum physics!Katie Norris was is originally from the suburbs of Maryland where her pastimes included climbing trees, singing and dancing, and asking a lot of questions. In school she enjoyed attending her classes, but she also had an insatiable passion for learning new skills beyond what was required. At an early age, she began creating a habit of self-study on different topics of interest. At eight years old, Katie taught herself to play the piano and began writing songs. She continued in music and theater all through high school and college, all while maintaining honors in her other academic subjects. When she wasn't playing music or acting on stage, Katie was reading a book or visiting her friends in the science department, always curious about their new projects. She graduated with a Bachelor's in Theater and Music from Frostburg State University, Maryland, and then moved to New York City to start her career in the theater world. Katie continued with her self-study practices between auditions and performance gigs, and learned a handful of random other skills, one of which was teaching herself Japanese, initially on a dare.After a few years in the NYC, Katie moved to Los Angeles where she continues her career in the entertainment industry. And her pursuits of science and self-study on all the topics that interest her have not slowed down. After locating to L.A., Katie dove more deeply into the world of physics and quantum physics. And for a time, she even held a position teaching after-school programs on physics and robotics for elementary and middle school kids. Her passion for knowledge on quantum physics has been ever-growing and evolving.https://katienorris.org/Become a supporter of this podcast: https://www.spreaker.com/podcast/earth-ancients--2790919/support.

SpaceTime Series 29 Episode 103 *Warnings that some Earth microbes could contaminate the Moon NASA says some of Earth's microbes hitching a ride to space with human explorers could survive in the shaded nooks and crannies of the Moon's South Pole contaminating the otherwise pristine environment. *China's first reusable rocket China has finally landed a reusable rocket in one piece. *Russia develops a nuclear rocket motor Russia has developed a prototype nuclear plasma engine which could cut journey times to Mars from 6 months to just 30 days. *September Skywatch The September equinox, and the constellations Capricorn, Pegasus, and Cygnus are among the highlights of the September night skies on SkyWatch.   Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally   

SpaceTime Series 29 Episode 102 *SpaceX's Starship arrives at Christmas Island After some 24 days at sea, a SpaceX recovery team has successfully towed its test flight 13 Starship into the remote Australian Indian Ocean territory of Christmas Island.  The unplanned operation was hastily put together after the spacecraft unexpectedly held together following its splashdown off the Western Australian coast. *How old really is the universe? Astronomers have come up with a new way to work out the age of the universe finding it to be 13.6 billion years old – some 200 million years younger than previously thought. *Einstein's general relativity theory still stands More than 100 years after Albert Einstein published his iconic general theory of relativity, it remains science's best understanding of gravity and the universe on the cosmic scale. *The Science Report A new study shows that carrying a little extra weight could be beneficial as you get older. People lacking human connections more likely to turn to AI chatbots for companionship. The Southern Ocean and parts of the Pacific may be the best places to fertilise to remove CO2. How dogs have learnt to understand human facial expressions. Alex on Tech: The Galaxy passport sized Z Fold 8 tops the sales figures.

SpaceTime Series 29 Episode 101 *NASA scraps its Swift rescue mission NASA has been forced to abandon its planned rescue mission to save the Swift Gamma Ray Space Telescope. *The Milky Way's first big galactic collision Astronomers have discovered what was probably the first major galactic collision involving our home galaxy the Milky Way cannibalising a dwarf galaxy less than two billion years after the big bang. *Understanding the stuff atoms are made of New observations by the world's largest atom smasher have revealed some of the basic fundamental physics of the universe, challenging long-standing theories of how gluons behave inside atomic nuclei. *The Science Report Australia records its first confirmed case of H5 N1 bird flu in a mammal on the nation's main land. How bushfires are polluting water supplies with arsenic. Successful Phase 3 trials of a new Australian developed personalized mRNA cancer vaccine. The growing impact of plastic pollution on Lord Howe Island's Sable Shearwaters rookery. Over 90% of Blue Mountains rainforests wrongly catalogued as eucalypt forests. The Skeptics guide to the ghost of the Crying Tree pub.Our guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

What happens when time stops behaving the way it should? Tonight, on Tales of the Night, we will hear accounts from people who claim to have lost hours, lived through moments that seem to repeat, and experienced situations for which—even years later—they still have no explanation. These are stories of space-time glitches and inexplicable experiences that force us to question whether our perception of reality is as reliable as we believe. Hosted by Simplecast, an AdsWizz company. See https://pcm.adswizz.com for information about our collection and use of personal data for advertising.

SpaceTime Series 29 Episode 100 The ingredients for planets formed far earlier than thought A new study suggests that the basic ingredients needed to form planets - and possibly life - may have existed just a hundred million years after the big bang -- far earlier in the history of the Universe than previously thought. Hunting for the hypothetical axion For years scientists have been hunting for a hypothetical elementary particle called the AXION which could lead to new physics unlocking some of the greatest mysteries of the universe. Now the German Research Foundation has given a green light to the construction of a new six million Euro experiment called the BabyIAXO – a first step in the search for the elusive and as yet undetected particle. NASA's Swift rescue mission back on track Katalyst Space has successfully uploaded a new flight software update designed to return its Link spacecraft into a more stable configuration.  Link is on a rescue mission to save NASA's Swift gamma ray space telescope from crashing back to Earth. The Science Report A new potential cure for baldness. Study shows people are most ticklish on the neck, armpits, belly, and the soles of their feet. Scientists have discovered a fossil of the earliest known siphuncle-bearing cephalopod. Study shows moving a pet cat from an outdoor lifestyle to living indoors is easier than you think. Skeptics guide to chasing the Moon.   Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

SpaceTime Series 29 Episode 98 Three supermassive black holes discovered in a single galaxy for the first time Astronomers have for the first time, discovered a galaxy near the dawn of time containing three monster black holes. A new White House advisory body on UFOs The pentagon has just released its latest batch of X-files looking at unexplained phenomena and strange happenings.  The release follows the appointment of Harvard University astrophysicist Avi Loeb as chair of the new White House scientific panel on UFOs. Was Venus once just like Earth A new study suggests that the hellish inhospitable world of Venus could have once been far more Earth like. The Science Report Western Europe has just experienced its hottest June and July on record. A new study says if you want to live longer then take the stairs. Scientists still don't know how long most fungi can live - or even how to define their age. A dog's ability to understand human language is likely a product of both nature and nurture. Skeptics guide to the missing nuclear scientists. Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

SpaceTime Series 29 Episode 97 The world watches as a SpaceX rocket crashes into the Moon A disused SpaceX Falcon 9 upper stage has crashed into the Moon providing astronomers with useful data on the effects of impacts hitting the lunar surface. The radiation eating mutant fungus from Chernobyl Scientists say there's a black fungus in the highly radioactive ruins of the Chernobyl power station that's now evolved to feed on radiation. NASA's Habitable Worlds Observatory Scientists have designed two revolutionary new computer chips to process data from a new space telescope called the Habitable Worlds Observatory that will search the skies for Earth like planets and for signs of life on those worlds. The Science Report A strong correlation found between dementia death rates and living in the suburbs. Improving pedestrians' thermal comfort through tree, shrub, and grass plantings along city streets. The exotic origin story of the Australian platypus. New Zealand's endangered yellow-eyed penguin found to be three genetically distinct subspecies. Skeptics guide to high tech water diviners.   This Week's Guests Professor Robert Ward from the Australian National University   Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

SpaceTime Series 29 Episode 95 Spectacular vortices discovered on the Sun's surface Astronomers have for the first time observed swirling whirlpool like vortices deep in the Sun's photosphere – its visible surface. The event horizon of merging black holes seen for the first time Astronomers have for the first time picked up gravitational waves from the event horizon during the actual merger of two black holes. Starship recovery efforts underway off Western Australia SpaceX are towing their Test 13 Starship prototype back to shore after it survived its Indian Ocean soft splash down off the Western Australian coast. The Science Report Scientists develop a new blood test to help test for signs of Alzheimer's. E-scooter riders found to be 3.5 times more likely than motorcyclists to get traumatic brain injuries. Reshaping sciences understanding of mammal evolution during the age of dinosaurs. A new study shows how wireless charging for drones could be achieved using lasers. Skeptics guide to claims Artemis II was a hoax. This Week's Guests Professor Robert Ward from the Australian National University   Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

SpaceTime Series 29 Episode 94 Black holes are a hundred times more powerful than we thought Astronomers have been forced to rewrite their textbooks on the power of supermassive black holes after new observations have shown that these cosmic monsters can generate enough energy to impact galaxies hundreds of thousands of light years away. The dramatic collision the flipped our Milky Way galaxy A new study suggests that our home galaxy the Milky Way may have undergone a dramatic change in orientation following a head on collision with another galaxy. A successful return to Earth for the crew of the Soyuz M-28 The Roscosmos Soyuz M-28 spacecraft has returned safely to Earth bringing home the expedition 74 crew who have spent the past 8 months aboard the International Space Station. August SkyWatch The red supergiant Antares, Barnard's star is the second nearest star system to the Sun, and the annual Perseids meteor shower are among the highlights of the August night skies on SkyWatch   This Week's Guests NASA WB-57 pilot Tom Parent NASA WB-57 Sensor Equipment Operator Cary Klemm Nation Wide Eclipse Ballooning Project Principal Investigator Dr Angela Des Jardins from Montana State University. Juno Principal Investigator Scott Bolton Southwest Research Institute in San Antonio Texas Juno Launch Director Omar Baez NASA Juno Mission Manager John Calvert NASA   Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

We play a very interesting clip about the nature of reality. Turns out it might be entanglement all the way down. Enjoy.godseyeviewbook@gmail.comJoin the discord at: https://discord.gg/PRwpPBNY4

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

Sponsor Link:This episode of SpaceTime is brought to you with the support of NordVPN. Now more than ever, you need a reliable VPN service to protect you online. Governments, bad actors and who knows are all trying to access your details online. Get protected by the best - NordVPN. Check out our special deal by visiting nordvpn.com/stuartgary and help support the show at the same time. Win/win!SpaceTime Series 29 Episode 91 The near-Earth asteroid that's actually a comet Astronomers have confirmed that the near Earth Asteroid 1998 SH2 is actually a comet further blurring the lines between the two types of celestial bodies. Psyche's gravity assist Mars Flyby NASA's Psyche spacecraft has undertaken a successful gravity assist flyby of the red planet Mars to help slingshot the probe to its ultimate target – the strange metallic asteroid 16-Psyche. China expanding its space station As NASA looks at winding down and de-orbiting the International Space Station in 2030, Beijing has announced new plans to double the size of its Tiangong Space Station. The Science Report New clues about how the body's nervous system senses warm and cool temperatures. Alarming pesticide levels found in the food of endangered Western Australian Carnaby's cockatoos. Scientists have discovered evidence of an ancient species of snake that lived underground. Study warns working from home may not have a significant impact on your carbon footprint. Skeptics guide to multiple Ohio Bigfoot sighting.     Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian SkepticsBecome a supporter of this podcast: https://www.spreaker.com/podcast/spacetime-with-stuart-gary--2458531/support.

SpaceTime Series 29 Episode 90 Discovery of an atmosphere on a planet orbiting a red dwarf Astronomers think they may have detected an atmosphere on a super-Earth exo-planet orbiting in the habitable zone of a nearby red dwarf star. How mega-dust storms and the Sun affect Martian weather A new study has found that solar storms hitting Mars at the same time as planetary wide dust storms could cause major changes in Martian weather. Approval granted for the world's first orbital space reflector array America's Federal Communications Commission has approved plans for the first demonstration flight of Reflect Orbital's Eärendil-1 satellite which will test the idea of reflecting sunlight back to Earth on demand. The Science Report Family socioeconomic status more important than parent smoking or drinking for kid's future health. Ancient 3000-year-old wooden beams dating to the first Jewish temple discovered in Jerusalem. How Rottnest Island Salt Lake bacteria could make biodegradable plastic. The artery-on-a-chip technology that replicates a patient's blood vessel structure and flow dynamics. Alex on Tech AI getting more Skynet dangerous. Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

SpaceTime Series 29 Episode 89 A lucky 13th test flight for the world's biggest rocket The world's largest and most powerful rocket has successfully launched on its latest test flight bringing it a step closer to NASA's Artemis program's aim of returning humans to the lunar surface, and eventually on to Mars and beyond. Scientists identify the asteroid that killed the dinosaurs A new study has determined that the asteroid which slammed into the Earth 66 million years ago wiping out over 75 percent of all life on the planet including all the non-avian dinosaurs was most likely a rare CO type carbonaceous chondrite. SpaceX rocket about to crash into the Moon The spent upper stage of a SpaceX Falcon 9 rocket is about to slam into the visible side of the Moon, providing scientists with a new method of pinpointing exactly where objects strike the Moon, data that could help prepare for future lunar seismic experiments. The Science Report Warnings that a third of all heart related deaths due to ultra processed foods. New data suggests fewer people in Australia are now smoking. Scientists have identified high levels of vehicle tyre dust on local Brisbane balconies. Iran continuing its nuclear weapons program in a complex deep below Pickaxe Mountain. Skeptics guide to fake news click bait. Our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

SpaceTime Series 29 Episode 88 The hellish planets orbiting a nearby star Astronomers have painted the most detailed portrait yet of the planetary system orbiting Barnard's Star – the Sun's closest stellar neighbour after Alpha Centauri, just under six light-years away. Australia's role in India's manned space program India's manned space program is moving ahead with the commissioning of a new space tracking station on Australia's Cocos (Keeling) Islands. The latest mission to the International Space Station The Russian Soyuz MS-29 spacecraft has successfully docked with the International Space Station three hours and two orbits after launching from the Baikonur Cosmodrome in the central Asian republic of Kazakhstan. The Science Report A new study has found a link between health risks and electromagnetic fields. Wild koala given a chlamydia vaccine implant to help them survive the ever spreading disease. A quadriplegic now able to feed himself using his own hand, thanks to microchip brain implants. Study shows the last five decades of pop music has slowly shifted towards being more self-focused. Skeptics guide to the Pentagon's latest UFO file release.     Our Guests This Week: Mars Perseverance Rover Deputy Project Scientist Katy Stack Morgan NASA JPL Mars Perseverance Rover project Scientist Ken Farley NASA JPL Mars Perseverance Rover Deputy Project Manager Matt Wallace NASA JPL Mars Perseverance Rover Chief Engineer Adam Steltzner NASA JPL Mars Perseverance Rover Mobility Team Member Farah Alibay NASA JPL     And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

SpaceTime Series 29 Episode 87 Ancient impacts display the red planet's past history A new study has shown how ancient Martian rocks exposed on the rim of Jezero Crater are preserving a 3.9-billion-year-old “weather report” from the solar system's most dynamic era. Understanding of the Moon's regolith for human habitation Scientists have mapped out the thickness of the Moon's regolith, the layer of loose dust and rock that covers the entire lunar surface finding it averages about six meters in the highlands and about four meters in the maria. Strange goings on Pluto's Charon A new study has found that the rotation of Pluto's binary partner Charon is slowing down. The Science Report Heart disease linked to nine specific molecules produced in your gut biota. A billion more people now face at least one day of extreme heat stress compared to the 1970s. Claims consuming an avocado daily, lowers the risk of heart disease in people with obesity. Just like humans, monkeys appear to dislike AI avatars that look like them, but are just a bit wrong. Alex on Tech: APPLE IOS-27 beta publicly released.Our Guests This Week: Mars Perseverance Rover Deputy Project Scientist Katy Stack Morgan NASA JPL Mars Perseverance Rover project Scientist Ken Farley NASA JPL Mars Perseverance Rover Deputy Project Manager Matt Wallace NASA JPL Mars Perseverance Rover Chief Engineer Adam Steltzner NASA JPL Mars Perseverance Rover Mobility Team Member Farah Alibay NASA JPL     And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

It may be the best book subtitle we’ve encountered in the Planetary Society book club: “Particles, Poetry and the Cosmic Dream Boogie.” Theoretical physicist and black feminist science theorist Chanda Prescod-Weinstein has followed her acclaimed first book, “The Disordered Cosmos,” with this new, equally celebrated, deeply personal romp across the Universe, “The Edge of Space-Time”. Swinging from quantum mechanics to hip-hop, and from galaxies to Alice in Wonderland, she guides us through the looking glass to what we know and would love to know about the Universe. It all but guaranteed a mind-bending conversation with host Mat Kaplan that begins with Chanda’s concern for the state of scientific research in the United States, and ends among the stars. Discover more at: https://www.planetary.org/planetary-radio/book-club-chandra-prescod-weinsteinSee omnystudio.com/listener for privacy information.

SpaceTime Series 29 Episode 85 Detecting nuclear weapons in space New modelling says small cube-sat sized spacecraft not much bigger than a fridge could be used to detect the presence of nuclear weapons on satellites in space. NASA's rescue mission to save the Swift space telescope reaches orbit NASA's Swift re-boost rescue mission is finally on its way to intercept the gamma ray space telescope and try to save it from its fiery fate. Stopping astronauts from wanting to kill each other on long space trips It can be tough enough getting on with family members at home or the people you have to work with – but imagine what it will be like if you're stuck with a small group of people on a multi-year mission to Mars. The Science Report A new study indicates that people who speak more than one language seem to have younger brains. Warnings that two thirds of Australian adults and a quarter of kids are now overweight or obese. Half of all fatal electric scooter accidents in Sweden are caused by drunk riders. Great news for chocolate lovers with 4 new species discovered. Skeptics guide to the case of the missing peacock.   Our Guests This Week: Professor Chris Kirkland from Curtin University NASA planetary geologist Cynthia Phillips Professor Dorothy Carter from Michigan State University   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

SpaceTime Series 29 Episode 84 Japan swoops past a cosmic snowman shaped asteroid Japan's Hayabusa2 spacecraft has just swooped past a tiny asteroid in deep space that's shaped like a snow man. Revealing the secrets of the ice moon Europa A new study has shown that the Jovian Ice Moon Europa is reflecting radio signals in a strange unexpected way. What happens to Earth when the Sun dies A new study has provided astronomers with a glimpse of the future we're likely to face when our Sun dies in around seven billion years from now. The Science Report Study's show early risers have more nutritious diets than people who prefer staying up late. Australia shown to be home to a large chunk of the world's seagrasses, but is losing them rapidly. A new study claims Meta's AI smart glasses could help people with low or no vision. Data centres found to have a far bigger carbon footprint than previously thought. Alex on Tech: Samsung Galaxy Z fold 8.Become a supporter of this podcast: https://www.spreaker.com/podcast/spacetime-with-stuart-gary--2458531/support.

SpaceTime Series 29 Episode 83 Earth's oldest known asteroid impact crater dated in Western Australia Scientists have determined the most precise age yet for the oldest known impact crater on Earth finding it to be some 3.024 billion years old.   A Chinese spacecraft has just reached Earth's second moon China's Tianwen-2 spacecraft has arrived at a temporary second moon orbiting Earth -- although technically the object is a small near Earth asteroid which is orbiting in sync with the Earth around the Sun. Sea floor spreading seen in action for the first time Scientists studying a seafloor site southwest of Australia have for the first time actually captured how molten rock emerges at boundaries between the Earth's tectonic plates. The Science Report A new vaccine with the potential to protect people from developing HIV AIDS. Warnings that losing even a small amount of sleep each night could be linked to weight gain. South Australia's algal bloom found to be the most toxic species of its kind ever studied. China launches a nuclear capable missile across the Pacific in a clear threat to island nations. Skeptics guide to the Westall High School UFO sighting. Our Guests This Week: Professor Chris Kirkland from Curtin University NASA planetary geologist Cynthia Phillips Professor Dorothy Carter from Michigan State University   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

SpaceTime Series 29 Episode 82 China's Mars sample return mission set for 2028 China says its planning to launch a Mars sample return mission in two years bringing back at least 500 grams of Martian regolith by 2031. Is science wrong about the universe The universe should look the same in all directions on the large cosmic scale, but new data based on dark energy observations are suggesting otherwise. An ASSASSN reveals its secrets Astronomers have converted observation of a nova explosion on a distant star into sound waves to better understand the dynamics of the spectacular blast. The Science Report Vitamins A and D linked to better lung function and a slowdown of biological aging. The first ever human bladder-kidney transplant reaches promising six-month milestone. Discovery that some native grasses not only survive and thrive after local wildfires. Scientific confirmation that Female faces are consistently rated as more attractive than males. Skeptics guide to the most popular UFO hotspots.   Our Guests This Week: Professor Tim Johnson from Curtin University   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

Sponsor Link:This episode of SpaceTime is brought to you by Inconi. Take back your privacy and data online with Incogni. Check out our special offer: www.incogni.com/stuartgarySpaceTime Series 29 Episode 81How a passing star redirected comets to the inner solar systemA fascinating new study reveals how a passing star, HD 7977, may have altered the trajectory of comets from the Oort Cloud, sending them cascading into the inner solar system. This event, which occurred approximately 2.47 billion years ago, could still be influencing comet activity today. Researchers used data from the European Space Agency's Gaia mission to refine the distances involved and suggest that the gravitational perturbations from HD 7977 temporarily dominated the generation of new comets.Changing our map of the Milky Way GalaxyAstronomers have redrawn the map of our Milky Way galaxy, moving its outer arms up to 10% further away than previously estimated. This revised picture is based on observations of gamma-ray bursts and the subsequent echoes of X-rays that helped to measure distances within the galaxy. New techniques have allowed for a clearer understanding of the Milky Way's structure, confirming the existence of its four spiral arms.Evidence of vast hidden magma systems inside MarsNew findings suggest that Mars once hosted extensive magmatic systems beneath its surface, despite the absence of plate tectonics. Data from NASA's InSight mission has revealed a previously unidentified boundary layer deep within the Martian crust, indicating complex geological processes that may have allowed the Red Planet to develop a habitable environment. This challenges long-held assumptions about the geological capabilities of rocky planets without tectonic activity.The Science RobertA new study indicates that the mental health of high school peers can significantly affect individual mental health outcomes. Additionally, research finds no link between paracetamol use during pregnancy and the risk of autism or ADHD. A detailed analysis of a fossilised pterosaur wing reveals insights into its diet, while scientists discover new methods to control quantum light sources, bringing us closer to practical quantum technologies.1. How a passing star redirected comets to the inner solar system 2. Changing our map of the Milky Way Galaxy 3. Evidence of vast hidden magma systems inside Mars 4. The Science RobertIf you'd like to support the podcast and gain access to bonus content by becoming a SpaceTime crew member, you can do just that through The Big Bang editions on Patreon, Spotify, and Apple Podcasts. Details on the Support page on our website https://www.bitesz.com/show/spacetime/support/

Is there no single shared "now" for everyone in the universe? Neil deGrasse Tyson and comic co-host Paul Mecurio dive into another grab bag of fan questions about superconducting asteroids, what memories are made of, and if you could ever experience pure unwarped spacetime. NOTE: StarTalk+ Patrons can listen to this entire episode commercial-free here:  https://startalkmedia.com/show/cosmic-queries-pure-spacetime/ Thanks to our Patrons Alex Sadashiv Nayak, Raviteja K, Lian Zhalka, Thomas Davis, Alex Yumashev, Charles Koehl, Matthew Maldonado, Aleksandre Khatiskatsi, Grig Coker, Matthew Ardebi…Decided to restructure names into comma-separated formatDecided to restructure names into comma-separated formatAlex Sadashiv Nayak, Raviteja K, Lian Zhalka, Thomas Davis, Alex Yumashev, Charles Koehl, Matthew Maldonado, Aleksandre Khatiskatsi, Grig Coker, Matthew Ardebili, Japeth Mitchell, Gaaary, Ian Patton, Casey Steelspine, Veninator, Hannu Latvakoski, Santiago Aguirre, Jonathan Caples, Ryan Wetmore, Dan Lepping, Chris Frank Betz, William Massar, Jason Durden, Jenny Patton, LiveAlive42, Kesh Iyer, scott ezell, Jesse Jensen, Javier E. Gonzalez, rk89, Courteney Kawalek, James Martin, seth forever, Why So Serious, Jared Kennington, andrew bash, David Pillado, Jennie Hrobak, Mariana, Robbie Rogers, Michael Huckey, Anthony Torres, Luke Gilliam, David Hasenauer, Libby Higgins, Willie Lo, and Mike B. for supporting us this week. Subscribe to SiriusXM Podcasts+ to listen to new episodes of StarTalk Radio ad-free and a whole week early.Start a free trial now on Apple Podcasts or by visiting siriusxm.com/podcastsplus. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

SpaceTime Series 29 Episode 80 Did ancient asteroid impacts prevent Earth's continents from forming A new study suggests the barrage of asteroid impacts that slammed into the ancient Earth during the Hadean Eon between 4.6 and four billion years ago may have prevented the formation of the planet's first continents. Could the ice giants Uranus and Neptune really be magma worlds A new study suggests that the solar systems two ice giants Uranus and Neptune might actually be magma worlds. World's biggest atom smasher powers down The world's most powerful atom smasher has been shut down for a four year major refit. The Science Report Sedentary behaviour linked to a 9% higher risk of death by cancer. Confirmation that mRNA vaccines are safe and highly effective. Artificial night time lighting has made planet Earth 16 percent brighter between 2014 and 2022. Study shows sending an electric current through black coffee can measure its strength and roast. Skeptics guide to AI and misinformation. Our Guests This Week: Professor Tim Johnson from Curtin University   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

Sponsor Link:This episode of SpaceTime is brought to you with the support of Incogni. If you worry about where your online data is going, you need Incogni. Worry no more. Check out our special SpaceTime offer (with 30 day money back guarantee) by visiting https://www.incogni.com/stuartgarySpaceTime Series 29 Episode 79 How Neutron Stars make heavy elements Physicists have achieved a significant breakthrough in understanding how Neutron Stars forge heavy elements. Aleutian subduction zone older than thought A new study has found that the subduction zone between the Pacific and North American tectonic plates are older than previously thought. The wobbling peanut asteroid Astronomers studying the inner main belt asteroid Donaldjohanson have found that its rotation wobbles. July Skywatch Planet Earth at its greatest distance from the Sun, the constellations Regulus and Leo, and one of the biggest known stars in the universe Antares are among the highlights of July's night skies on Skywatch.   Our Guests This Week: Uk Space Agency Programme Manager Rosemary Young Principle Investigator MIXS Instrument Emma Bunce Leicester University Planetary Geoscientist David Rothery The open University   And our regular guests: Alex Zaharov-Reutt from techadvice.life And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

More To The Story: Growing up in Los Angeles in the 1980s and '90s, a daughter and granddaughter of social justice activists, Chanda Prescod-Weinstein fell in love with math and the physical sciences and developed a profound curiosity about the cosmos (though the smoggy night sky of her childhood blocked her view of the stars). She soon developed a detailed plan for her life that led to a career writing and teaching about physics and gender studies at the University of New Hampshire. Today, Prescod-Weinstein's work stands out for the ways she weaves her identity as queer, Black, and Jewish into her work. In her latest book, The Edge of Space-Time: Particles, Poetry, and the Cosmic Dream Boogie, Prescod-Weinstein brings a Black feminist lens to cosmology, quantum physics, poetry, and popular culture to help unlock the mysteries of the physical universe. On this week's More To The Story, Prescod-Weinstein talks about the need for diversity and inclusivity in the sciences and puts science fiction's various hypotheses for space travel to the test with host Al Letson.Read: The Edge of Space-Time: Particles, Poetry, and the Cosmic Dream Boogie (Pantheon)Read: Has America Lived Up to Its Founding Promise? (Reveal)Watch: How We Could Solve the Dark Matter Mystery (TED Talks)Read: The Disordered Cosmos: A Journey Into Dark Matter, Spacetime, and Dreams Deferred (Bold Type Books)Learn more: Chanda Prescod-Weinstein's personal website Donate today at Revealnews.org/more Subscribe to our weekly newsletter at Revealnews.org/weekly Follow us on Instagram and Bluesky Learn about your ad choices: dovetail.prx.org/ad-choices

SpaceTime Series 29 Episode 77 BepiColombo mission finally reaches Mercury space After eight years of powered flight towards the planet Mercury, mission managers have finally turned off the BepiColombo spacecraft's propulsion engines in preparation for planetary orbit insertion. Is our Sun changing A new study claims the Sun has been mysteriously changing over the last forty years. Strange new type of x-ray flare detected in deep space Astronomers have detected a mysterious never before seen new type of X-ray flare in the skies of the Southern Hemisphere. The Science Report A special report on the latest bombshell developments in the case of the COVID 19 pandemic. Our Guests This Week: Uk Space Agency Programme Manager Rosemary Young Principle Investigator MIXS Instrument Emma Bunce Leicester University Planetary Geoscientist David Rothery The open University   And our regular guests: Alex Zaharov-Reutt from techadvice.life And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

SpaceTime Series 29 Episode 76 A possible supernova remnant discovered in the galactic centre Astronomers may have discovered a supernova remnant near the supermassive black hole at the centre of our galaxy. A new quantum view of Big Bang A new study could change what science knows about the Big Bang and the earliest moments of cosmic history. Work begins on new Western Australian ground station for lunar missions Construction has begun on Kongsberg's new 20-metre parabolic dish antenna ground station at Mullewa in outback Western Australia. The Science Report Brain computer interface patient continues to communicate after two years. Powerful heatwave in Antarctica continues to push temperatures up. Study warns people eating ultra processed foods have higher risk of heart disease and death. Japan sends a transformer robot to the Moon. Skeptics guide to skeptical psychology.     Our Guests This Week: Dr Hadrien Devillepoix from Curtin University NASA Swift scientists Brad Cenko and Regina Caputo Katalyst CEO Ghonhee Lee Katalyst LINK lead Kieran Wilson   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

In this episode, Ray Cochrane unpacks how two colliding black holes revealed a whirlpool in spacetime, a direct detection of frame dragging hidden in the cleanest gravitational-wave signal ever recorded. Additional stories cover the James Webb Space Telescope, counting 16.5 million stars in the Cigar Galaxy, SpaceX rolling out Starship V3, deadly back-to-back earthquakes in Venezuela, GitHub fighting a California law that could break open source, and Meta engineering a battery narrow enough to live in a pair of glasses. – Want to start a podcast? It’s easy to get started! Sign-up at Blubrry – Thinking of buying a Starlink? Use my link to support the show. Subscribe to the Newsletter. Email Ray if you want to get in touch! Like and Follow Geek News Central’s Facebook Page. Support my Show Sponsor: Best Godaddy Promo Codes Get 1Password Full Summary Cochrane opens with a personal update before the night’s lead story. He recently graduated with a Bachelor of Science in Computer Science from Portland State University, celebrated with family in town, and launched a new site at rayc.world. That site links to a final-project study he built on collaborative filtering using podcasting data, hosted at cohort.rayc.world and drawn from OP3 analytics. He also plans to return to the show’s classic twice-weekly cadence on Mondays and Thursdays. From there, he goes deep on a new black hole discovery, then pivots through space, earth science, climate, biotech, open source, cloud infrastructure, and consumer hardware. Colliding Black Holes Reveal a Whirlpool in Spacetime Two black holes spiraled together, merged, and sent a gravitational wave rippling across the universe. Researcher Neil Lu and colleagues at the Australian National University found the fingerprint of frame dragging buried in GW250114, the cleanest signal LIGO has ever recorded. Frame dragging means a spinning black hole drags spacetime around with it, like a spoon turning in honey, except the honey is reality itself. Remarkably, the wave changed the distance between your nose and your ear as it passed, by far less than the width of a single atom. Sponsor: GoDaddy Economy hosting is $6.99/month, WordPress hosting is $12.99/month, and domains are $11.99. Website builder trial available. Use codes at geeknewscentral.com/godaddy to support the show. Webb Counts the Stars in the Cigar Galaxy NASA released a striking new James Webb Space Telescope view of Messier 82, the edge-on galaxy nicknamed the Cigar Galaxy. Because Webb sees in infrared, it peers straight through the dust that normally hides the galaxy’s interior. Combined with archival Hubble data, the image resolves roughly 16.5 million individual stars. M82 is a starburst galaxy, meaning it forms stars at a furious rate, a frenzy likely triggered when it merged with a neighbor. SpaceX Rolls Out Starship V3 SpaceX officially introduced Starship V3, the third generation of the largest rocket ever built. The vehicle now flies on the Raptor 3 engine, pushing liftoff thrust to around 20 million pounds and making it the most powerful rocket ever flown. More importantly, V3 is designed to carry over 100 metric tons to low Earth orbit while staying fully reusable, roughly triple the previous version. SpaceX also added in-orbit refueling hardware, the capability that finally makes operational Moon and Mars missions realistic. The Asteroid Barrage That Kept Earth From Forming Continents A team led by Curtin University and the Queensland University of Technology argues that relentless asteroid impacts shaped the very young Earth. During the Hadean, more than four billion years ago, the planet was struck far more often than it is today. Each impact dumped heat deep into the interior, repeatedly melting and reworking the crust. Consequently, stable continents formed much later than calmer models assumed, painting a picture of a hotter, weaker, more chaotic early Earth. Back-to-Back Earthquakes Devastate Northern Venezuela Northern Venezuela was struck by two major earthquakes on June 24, a magnitude 7.2 foreshock followed by a magnitude 7.5 mainshock. Both hit only about six miles underground, so the shallow shaking delivered its full force at the surface. Tragically, at least 164 people died, and the region sits along the tangled boundary where the Caribbean and South American plates grind past each other. These were the largest quakes to hit the area since a magnitude 7.7 event near Caracas in 1900. The ‘Guerrilla Solar’ Era Has Arrived A quiet energy shift, nicknamed “guerrilla solar,” is spreading across Europe. These small plug-in panels deliver power to a home’s wiring via a standard wall outlet, with no electrician or permit required. Germany now counts roughly a million of these systems. However, the U.S. payoff remains modest, with savings estimates of around $15 per month against a $500 to $1,500 setup cost. Why a Broken-Up Forest Stores Less Carbon Researchers quantified what foresters long suspected: an intact forest stores far more carbon than the same acreage split into fragments. A hectare inside a large, continuous forest proved about 38 percent more productive than an isolated one. The culprit is edge effects, the extra wind, heat, and direct sun that stress trees at a forest’s boundary. Because a large forest maintains a large protected core while fragments are nearly all edge, planting trees together matters for carbon storage. Edited Human Embryos Reveal a Surprise Researchers used base editing, a precise cousin of CRISPR that rewrites a single DNA letter without cutting the strand, in human embryos. They discovered that a protein called NANOG plays a role in early human development that it does not play in mice. In humans, switching it off still let cells form that seed the placenta and yolk sac. The finding argues that understanding human development requires studying human embryos directly, which reignites a thorny ethical debate. GitHub Fights a California Law That Could Break Open Source GitHub joined Black Forest Labs, Hugging Face, and Mozilla to push for fixes to California’s AI Transparency Act. As written, the bill could force revocation of an open-source license when a downstream user fails to meet certain obligations, which clashes with the permanent, irrevocable promise of open source. Cochrane pointed to curl and its longtime maintainer, Daniel Stenberg, warning that the rule could destabilize the supply chain on which the whole tech world runs. Instead, the coalition points to the EU’s AI Act transparency code as a saner model. Rust Opens Its Maintainers Fund The Rust Foundation launched a Maintainers Fund to pay the people who keep the language’s ecosystem healthy. Backed by RFC 3931, it establishes a funding team and a new Maintainer-in-Residence program for the often thankless work on the compiler, standard library, Cargo, and Clippy. Individuals can donate through GitHub Sponsors, while companies can sponsor there or contact the foundation directly. Cochrane urged any business that depends on open source to invest in the projects it actually uses. AWS Gives Lambda Its Own Isolated Sandboxes AWS introduced MicroVMs inside Lambda, its serverless platform. Each session runs in a dedicated micro virtual machine with no shared kernel and up to eight hours of total runtime. The feature exists for the AI era, in which applications increasingly run code written by an AI agent rather than by the developer. Use cases include AI coding assistants, data analytics platforms, vulnerability scanners, and game servers running user-supplied scripts. Meta Engineers a Battery Narrow Enough for Glasses Meta built custom steel-can battery cells as narrow as seven millimeters to fit the temple arms of smart glasses like the Ray-Ban Meta and Oakley Meta Vanguards. These cells power cameras, speakers, and AI features in a space most engineers would call impossible. To prevent brownouts, Meta swapped wound electrodes for precisely die-cut stacked layers that lower electrical resistance. Now the company is spreading the technology across multiple vendors and eyeing other wearables. Polestar Gets Locked Out of the US Market Starting in 2027, Polestar will not be able to sell its new models in the United States. A federal Connected Vehicle Rule bars cars containing certain Chinese or Russian software or hardware on national security grounds. The painful irony is that Polestar moved production of the Polestar 3 to South Carolina specifically to dodge tariffs on Chinese-built EVs. Because the rule targets the technology’s origin rather than its assembly location, the company is shut out anyway. Retroid’s Pocket Nova Packs Serious Power for $229 Retroid returned with a new retro handheld, the Pocket Nova, starting at $229 with a step-up model around $269. It features a 4.5-inch AMOLED screen in a 4:3 aspect ratio, a shape well suited to classic games. On paper, it should handle GameCube- and PlayStation 2-era titles, though that remains an early expectation rather than a benchmarked promise. Retroid has earned a strong reputation for high-quality, genuinely portable consoles. Cochrane signs off with the usual ecosystem mentions: GNC Insider at geeknewscentral.com/insider, the show newsletter, email at geeknews@gmail.com, and modern podcast app recommendations at podcastapps.com. The post Colliding Black Holes Reveal a Whirlpool in Spacetime #1866 appeared first on Geek News Central.

The Author Events Series presents Chanda Prescod-Weinstein | The Edge of Space-Time In Conversation with Airea D. Matthews In her highly acclaimed debut, distinguished cosmologist and particle physicist Dr. Chanda Prescod-Weinstein shared with her audience an abiding sense of wonder at the cosmos, while imagining a world without the entrenched injustice that plagues her field. Now, in The Edge of Space-Time, she embraces that cosmic wonder, taking readers on a mind-altering journey to the boundaries of the universe, inviting us to spend time at the edge of what we know about space-time and about ourselves. Guided by her conviction that for humanity to go forward we must know our cosmic past and drawing on poetry and popular culture-from Langston Hughes, Queen Latifah, and Lewis Carroll, to Big K.R.I.T., Sun Ra, and Star Trek-Prescod-Weinstein renders accessible some of the most abstract concepts of theoretical physics to tell fascinating stories about the history and fundamental nature of our universe. Here we meet the quantum cat that is both dead and alive, learn the difference between dark matter and dark energy, explore the inner workings of black holes, and investigate the possibility of a unified theory of quantum gravity, following our guide out to the far reaches of the cosmic event horizon and down to the tiniest (and queerest) neutrino. Along the way, she calls on us to resist colonial approaches to space exploration and instead imagine a better path forward in our pursuit of humanity's undeniable connection with the stars. Through Prescod-Weinstein's clear-eyed and unique perspective, and informed by her deep knowledge of post-colonial history and Black feminist thought, The Edge of Space-Time argues that physics is an essential way for everyone to look at the universe and presents a compelling case that ''the edge'' is a powerful vantage point from which to see the big picture. Chanda Prescod-Weinstein is an associate professor of physics and astronomy and core faculty in women's and gender studies at the University of New Hampshire. She conducts award-winning theoretical physics research on dark matter, the early universe, and neutron stars, while also researching Black feminist science studies. Her first book, The Disordered Cosmos: A Journey into Dark Matter, Spacetime, and Dreams Deferred, won the 2021 Los Angeles Times Book Prize in Science and Technology, the 2022 Phi Beta Kappa Award in Science, and a 2022 PEN Oakland/Josephine Miles Award. A columnist for New Scientist and Physics World, she is originally from East L.A., California, and now divides her time between the New Hampshire Seacoast and Cambridge, Massachusetts. Because you love Author Events, please make a donation when you register for this event to ensure that this series continues to inspire Philadelphians. Books will be available for purchase at the library on event night! All tickets are non-refundable. (recorded 4/8/2026)

Dasa and Andy continue a conversation about what actually leads to peak performance. If it's not confidence and feeling good about your swing, then what actually matters? They discuss a theory of linking motion to outcome, and get help from Interstellar and Albert Einstein along the way.

SpaceTime Series 29 Episode 74 Why Venus spins backwards A new study suggests that the strange retrograde spin of the planet Venus is the result of a massive impact event. Could Dark Matter explain what's happening at the centre of our galaxy A new study has failed to rule out Dark Matter as the source of the so called Galactic Center Excess at the heart of the Milky Way galaxy. Trying to solve a meteor cold case Last month astronomers detected a small near Earth meteoroid on a collision course with our planet. The Science Report The deadly H5N1 strain of bird flu detected on the Australian mainland for the first time. Australia's Bureau of Meteorology has officially declared an El Niño. The risk of suicide among males can persist for years following a relationship break up. Research continues on nuclear diamond batteries that could last thousands of years. A new species of shark discovered in the tropical Pacific, north of Australia. Skeptics guide to five lessons on misinformation from the ancients. Our Guests This Week: Dr Hadrien Devillepoix from Curtin University NASA Swift scientists Brad Cenko and Regina Caputo Katalyst CEO Ghonhee Lee Katalyst LINK lead Kieran Wilson   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

Sponsor Link:This episode of SpaceTime is brought to you by Incogni, your first stop in reclaiming your online privacy.To check out our special offer for SpaceTime listeners, visit www.incogni.com/stuartgarySpaceTime Series 29 Episode 73 The earliest known flickering quasar Astronomers have discovered the earliest known flickering quasar dating back to a time when the universe was just 850 million years old. ExoMars to target vast clay beds in search for life on Mars The European Space Agency has selected a vast clay bed called Oxia Planum as the best place on the red planet to search for signs of life. Understanding neutron star mergers Scientists have used deep learning neural networks to better understand the violent events associated with the merger of neutron stars. The Science Report New GLP-3 drugs significantly improve blood sugar levels and lead to substantial weight loss. Ocean waves generated in the Southern Ocean tracked all the way to the shores of Alaska. Are dogs left or right handed? Skeptics guide to fish oil supplements.   Our Guests This Week: Kovi Rose from the University of Sydney   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

Sponsor Link:This episode of SpaceTime is brought to you with the support of Incogni. They can't spam or scam you,if they can't find you. Get details on our special deal and get your online pivacy back. Visit www.imcogni.com/stuartgarySpaceTime Series 29 Episode 72 Our ever-changing Milky Way Galaxy New observations are showing astronomers how our galaxy the Milky Way is being slowly changed through its gravitational interactions with our nearby neighbouring satellite dwarf galaxy the Large Magellanic Cloud. How Jupiter may have helped life start on Earth A new study suggests the solar system's largest planet Jupiter may have provided some of the key ingredients for life to Earth. Astronauts on the space station prepare for emergency evacuation Astronauts aboard the International Space Station ordered to prepare of emergency evacuation of the orbiting outpost as cosmonauts began working to try and repair a growing leak in the Russian Zvezda service module. The Science Report Global warming reaches 1.37°C above pre industrial levels in 2025. A new AI study claims laser-powered engines could one day support ‘intelligent' 6G networks. Kids with smartphone aren't more likely to end up depressed or overweight, but will be more sleepy. Alex on Tech computer tablet sales continue to rise.  Become a supporter of this podcast: https://www.spreaker.com/podcast/spacetime-with-stuart-gary--2458531/support.

Every time you open your eyes, you're not seeing reality. According to the mathematics of evolution, you never have. Donald Hoffman is a cognitive scientist and professor who has spent over 40 years building mathematical models of perception. His book The Case Against Reality makes one of the most unsettling arguments in modern science: your senses didn't evolve to reveal the truth. They evolved to hide it. Think of your body, your thoughts, your memories, and your entire experience of this world as a VR headset. Not a metaphor. A working model backed by mathematics. The headset gives you exactly what you need to play the game of life. It hides what is actually running underneath. Consciousness isn't a product of the brain. It's the other way around. Neurons don't exist when no one is looking. Nothing in your behavior is caused by neural activity. What's actually driving your life is something science is only now beginning to map. The way out isn't more information. It's a practice. The silence between your thoughts? That's you. When you learn to watch your emotions instead of becoming them, the grip of every story you've been telling yourself starts to release. The Case Against Reality: Why Evolution Hid the Truth from Our Eyes Amazon Ebook Audiobook Visual Intelligence: How We Create What We See Amazon Donald Hoffman on X TRACE Institute Trace Institute Instagram In this episode you will: Understand why your brain has zero causal power over your thoughts, feelings, and behavior Reframe failure, suffering, and loss using Hoffman's headset model to reclaim who you actually are Discover why the mathematics of evolutionary game theory proves no organism has ever seen reality as it truly is Learn how Hoffman's Trace Logic model of consciousness could transform our understanding of physics, the brain, and human potential Practice the watcher technique to step outside your emotions and release the grip of false identity For more information go to https://lewishowes.com/1941 For more Greatness text PODCAST to +1 (614) 350-3960 Follow The Daily Motivation for essential highlights from The School of Greatness More SOG episodes we think you'll love: Lewis Howes Solo [Everything You Want In Life Comes When You Let Go] Emily McDonald Dr. Daniel Amen TOPICS Donald Hoffman, The Case Against Reality, evolutionary game theory, Trace Logic, Planck scale, VR headset model, perceptual interface, watcher practice, consciousness, spacetime Get More From Lewis! Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

In this episode of Talk Nerdy, Cara is joined by theoretical physicist, and associate professor of physics and astronomy, and core faculty in women's and gender studies at the University of New Hampshire, Dr. Chanda Prescod-Weinstein. They discuss her newest book, The Edge of Space-Time: Particles, Poetry, and the Cosmic Dream Boogie. Follow Chanda: @chanda

SpaceTime Series 29 Episode 71 Universe expansion still accelerating after all A new study has confirmed that the universe is continuing to expand at an ever-accelerating rate under the force of dark energy and heading for a cold, dark and empty fate. Mysterious cosmic signals finally explained Astronomers have discovered that dead stars called white dwarfs located in binary systems are a primary source of mysterious signals from deep space called long-period radio transients. What made last week's New England meteor incident so rare? Last week we reported on a meteor that rocked the afternoon spring skies over New England. It now turns out that was a very rare event. The Science Report Sugar-sweetened drinks increase the risk of two types of liver cancer. New fish species swimming in the warm tropical waters of the Great Barrier Reef. A new study claims that living with cats does not worsen asthma or allergies in children. Skeptics guide to on line influencers.

Sponsor Link:This episode of SpaceTime is brought to you by NordVPN, where your online security starts. To check out our special offer for SpaceTime listeners, visit www.nordvpn.com/stuartgarySpaceTime Series 29 Episode 70 *The Small Magellanic Cloud is being ripped apart A new study reveals that the Small Magellanic Cloud, a satellite galaxy of the Milky Way, is slowly being torn apart by gravitational forces from the Large Magellanic Cloud. Researchers have utilised over a decade of observations to uncover the galaxy's dynamic state, challenging previous models of coherent rotation. *Blueprint for a lunar base NASA's plans for a lunar base at the Moon's South Pole are sparking innovative proposals for construction using local lunar materials. The Texas A&M Space Institute is leading research into using lunar regolith, a challenging construction material, to develop habitats for future lunar missions. *Meteor rocks New England A recent meteor explosion over New England has been confirmed as a sonic boom from a meteor entering the Earth's atmosphere, sending shockwaves across Massachusetts and Rhode Island. The meteor, travelling at 121,000 kilometres per hour, likely fragmented before falling into the North Atlantic Ocean. *The Science Robert Increased wildfire risks are predicted across parts of Australia, while a study reveals that Iceman Otzi's microbiome remains active even after 5,300 years. Additionally, video technology may allow for heart rate monitoring through facial recognition.Become a supporter of this podcast: https://www.spreaker.com/podcast/spacetime-with-stuart-gary--2458531/support.

Sponsor Link:This episode of SpaceTime is brought to you with the support of NordVPN...where your online security starts. To check out our special offer for SpaceTime listeners, visit www.nordvpn.com/stuartgarySpaceTime Series 29 Episode 69 *Evidence of planetary destruction in the early solar system Scientists have confirmed a cosmic collision in the early solar system which saw the complete destruction of a planet possibly as big as Mars. *Are we missing a planet A new study suggests that one of our planets might be missing, and it could explain why the solar system looks the way it does. *SETI investigates interstellar comet 3I Atlas The search for extraterrestrial intelligence SETI institute says it's found no evidence of any alien technology associated with the interstellar comet 3I Atlas. *The Science Report Global average temperatures likely to continue at or near record levels over the next five years. A new study has discovered two distinct subtypes of autism with different underlying biology. Does reading stuff on paper help you better understand than reading it on a digital device. Alex on Tech Computex 2026.Become a supporter of this podcast: https://www.spreaker.com/podcast/spacetime-with-stuart-gary--2458531/support.

Dive into the nature of spacetime and data center backlash in this Edge of Wonder Live. Visit https://rise.tv for more exclusive content! Visit https://metaphysicalcoffee.com for coffee that's out of this world! Physics & Time: The Soviet Union once classified a strange discovery in space. In 1985, a Russian cosmonaut watched a spinning wing nut flip 180° every few seconds in zero gravity—behavior that seemed to defy normal physics. This Dzhanibekov Effect (or Tennis Racket Theorem) remained secret for years because scientists at the time had no clear explanation. That's just the beginning. As humans explore deeper into space, we keep uncovering bizarre truths about space-time itself. By some measurements, time literally runs slower closer to Earth and speeds up the farther you move away from its gravity. The more we study spacetime, the more it appears to change our everyday understanding of reality. Data Centers: The legendary environmental activist Erin Brockovich is now leading the pushback against massive AI data centers. She's launching a nationwide map to track community complaints. Meanwhile, some Big Tech voices are blaming Chinese Communist Party propaganda for the growing American resistance. Simultaneously, a company backed by Nvidia is offering to pay homeowners to host mini data centers right next to their houses—in exchange for discounted electricity and internet. Would you let them put one in your yard? Join Edge of Wonder for all of this plus the weekly “Bendela Effect.” Rise.TV Exclusive: During the “Dig Deep” Live Q&A segment, ask your questions directly. In the fan-favorite Top 10 Weirder News of the Week, hear hilarious and bizarre stories to end your week on a high note. And as always, see you out… on the edge!

Sponsor Link:This episode of SpaceTime is broughtto you by NordVPN, where your online security starts. To check out our special discount with bonuses offer, simply visit www.nordvpn.com/stuartgarySpaceTime Series 29 Episode 68 *How black holes shape the cosmos A new study has revealed how powerful jets generated by black holes shape the universe. *NASA forced to end its MAVEN Mars Mission NASA has been forced to shut down its MAVEN mission orbiting Mars following a mysterious spacecraft failure in December. *Earth gets a rare blue micro moon Skywatchers have just experienced a rare blue micro-moon. *The Science Report An El Niño climate event to develop this month and last at least until the southern hemisphere spring. One in six cases of COVID-19 might have resulted in patients suffering long covid. Palaeontologists have identified fossils of a new species of raptor-like dinosaur in Patagonia. Skeptics guide to antivaxxers change of heart. Our Guests This Week: Dr Steve Prabu from Curtin University Beth Johnson from the search for extraterrestrial intelligence SETI institute Texas A&M Space Institute Director Dr Nancy Currie-Gregg Texas A&M Space Institute lead Professor of Mechanical Engineering Dr Rob Ambrose NASA Johnson Space Centre Director Vanessa Wyche   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian SkepticsBecome a supporter of this podcast: https://www.spreaker.com/podcast/spacetime-with-stuart-gary--2458531/support.

SpaceTime Series 29 Episode 67 *Are we in a cosmic void after all? It's an hypothesis which has been around for decades and refuses to go away: Are we in a cosmic void? *New study confirms a black hole that formed before its galaxy Astronomers using the Webb Space Telescope have identified a supermassive black hole in the early universe that formed before its host galaxy. *Another win for SpaceX over Boeing NASA has just awarded SpaceX six more crew transfer missions to the International Space Station because Boeing still can't certify its Starliner spacecraft as safe for human operation. *SkyWatch June The June Solstice, the constellation Sagittarius, and the Taurids meteor shower are among the highlights of the June night skies on Sky watch.   Our Guests This Week: NASA Administrator Jared Isaacman NASA Associate Administrator Lori Glaze NASA Moon Base executive Carlos García-Galán   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

SpaceTime Series 29 Episode 65 *NASA confirms its moon base plans and first contracts NASA confirms its plans to have humans living on the Moon by 2032. The agency has released the latest draft of its lunar south pole base project and signed its first contracts. *A surprising core reversal deep inside the Earth The European Space Agency has discovered a mysterious flow reversal of Earth liquid iron outer core. *Red dwarf stars detected 'eating' Earth-like planets Astronomers have discovered some of the strongest evidence yet that stars eat their offspring. *The Science Report The healthy tomato-soy juice cocktail that could lower inflammatory proteins. A new species of giant mosasaur fossil discovered in Texas. Sodium-ion batteries could become a genuine low-cost rival to lithium-ion technology. Skeptics guide to secret flying saucers hidden in plain sight. Our Guests This Week: NASA Administrator Jared Isaacman NASA Associate Administrator Lori Glaze NASA Moon Base executive Carlos García-Galán   And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics And Senior science writer and Sky and Telescope magazine contributor Jonathan Nally  

SpaceTime Series 29 Episode 64 *A new explanation for how stars explode  A new study suggests that neutrino which are some the least massive objects in the universe may trigger some of the biggest explosions in the cosmos – supernovae the explosive death of massive stars which are so bright they can outshine entire galaxies. *Neptune's mysterious moon Nereid A new study suggests the planet Neptune's distant moon Nereid may be the last of the ice giant's original satellites which somehow managed to survive a cosmic collision.. *A safe return to Earth for a hypersonic test vehicle Varda Space Industries' W-6 capsule has safely returned to Earth, parachuting down into the Australian outback. *The Science Report New study claims your eyes could indicate of how strong your bones are. Scientists confirm insects feel pain. Researchers show most Australian Wild Dogs have mostly dingo ancestry. Skeptics guide to bigfoot visits the Marines at Quantico.     Our Guests This Week: Dr Finn Stokes from Adelaide University Dr. Kirsty Duffy from Fermilab Dr. Jessica Turner from the University of Durham.     And our regular guests: Alex Zaharov-Reutt from techadvice.life Tim Mendham from Australian Skeptics  

Can we influence the strong nuclear force? Neil deGrasse Tyson and Paul Mecurio answer grab bag questions about sci-fi laser guns, the Roche Limit, how we interact with the fundamental forces, and more! NOTE: StarTalk+ Patrons can listen to this entire episode commercial-free here:  https://startalkmedia.com/show/cosmic-queries-scars-in-spacetime/ Thanks to our Patrons Gladys Strickland, Jonathan Marino, Petri Rajama, Benjamin Cross, Smooth, Cecelia Linley, John Burgin, Elizabeth Shope, Barrett Mayes, Paweł Szczypa, Ivan Ocampo, Angelo Rios, Luisangel Araujo, B-RO RTR, Sebastian Poehlmann, Kendra, Charles, LateGame, Stephanie, Denis, Joseph Hodge, Daniel Smith, Matt Sutton, Ziyod Yusupov, TheAceIsHere _, Robert Baughman, Patricia Weaver, Scott Jones, Luis Figueroa, TheJosh, Justin Garrity, J. Michael Mastro, Andreas Sorteberg Vik, Christian Di Patria, Steve Kingan, Martha, Nick, Jeff Ferren, Louise Keyte, Ann Hosler, Darren, Roni Gi, Salacious B Crumb, Tero Tommola, Dhaval, Andy Roberts, Brian Simmons, Toney, Remedy, Terry Melman, David Smith, Andrew M Gross, Conan, Raz, Joseph Watkins, Joe, Dom WB, Mike Bertuccio, Deepak Mani, Adam Dockerty, Mike, Habib Hassan, Exercise Enlightenment, Everett, Twisted Universe, Jason Prechtl, Luis Antonio Leon, SwillisBolt, Switchblade91, Linda Hall, Bo J, Megan Marler, Dalton, Jim, Chris Brown, Krisztian Unpronounceable, Donce, Jay, Jacob, Suzan Wallace, Ted, Steve James, TERP Radio, Sublimis, Alexander Casian, Onlymeami, Zack Blankenship, John Perez, Specter, DJ, Kristian Jeremiassen, Adam Flores, Dan Herman, Zef Correal, Maddie, Adam, Mark, Mary, Andrew494, and Matthew Grieve for supporting us this week. Subscribe to SiriusXM Podcasts+ to listen to new episodes of StarTalk Radio ad-free and a whole week early.Start a free trial now on Apple Podcasts or by visiting siriusxm.com/podcastsplus. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Most people think that space-time and the physical universe are the ultimate reality—something solid and unchangeable, governed by the laws of physics. But what if that's just the tip of the iceberg? What if our everyday experience is nothing more than a kind of VR headset, a useful interface that hides the deeper layers of reality from us? I find this idea not only fascinating, but increasingly convincing, especially as breakthroughs in both physics and computational theory keep bending the boundaries of what's possible. To challenge our assumptions and help us explore what's really behind the headset, I bring you today's guest—a cognitive scientist who argues that space-time is not fundamental, and that consciousness might be the true base reality. He believes that if we figure out the code underlying the simulation, we could unlock possibilities that make nuclear weapons look like firecrackers—and that the future of science is about to blow our collective minds. With that in mind, I bring you Donald Hoffman. Ketone IQ: Visit https://ketone.com/IMPACT for 30% OFF your subscription orderQuince: Free shipping and 365-day returns at https://quince.com/impactpodMonetary Metals: Future-proof your wealth at https://monetarymetals.com/impactTruemed: Check your eligibility and start saving at https://truemed.com/impactAT&T Business: Switch to AT&T Business at business.att.comIncogni: Take your personal data back with Incogni! Use code IMPACT at the link below and get 60% off an annual plan: https://incogni.com/impactShopify: Sign up for your one-dollar-per-month trial period at https://shopify.com/impactNetsuite: Right now, get our free business guide, Demystifying AI, at https://NetSuite.com/TheoryQuo: Try for free PLUS get 20% off your first 6 months at https://quo.com/impact Learn more about your ad choices. Visit megaphone.fm/adchoicesSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Most people think that space-time and the physical universe are the ultimate reality—something solid and unchangeable, governed by the laws of physics. But what if that's just the tip of the iceberg? What if our everyday experience is nothing more than a kind of VR headset, a useful interface that hides the deeper layers of reality from us? I find this idea not only fascinating, but increasingly convincing, especially as breakthroughs in both physics and computational theory keep bending the boundaries of what's possible. To challenge our assumptions and help us explore what's really behind the headset, I bring you today's guest—a cognitive scientist who argues that space-time is not fundamental, and that consciousness might be the true base reality. He believes that if we figure out the code underlying the simulation, we could unlock possibilities that make nuclear weapons look like firecrackers—and that the future of science is about to blow our collective minds. With that in mind, I bring you Donald Hoffman. Ketone IQ: Visit https://ketone.com/IMPACT for 30% OFF your subscription orderQuince: Free shipping and 365-day returns at https://quince.com/impactpodMonetary Metals: Future-proof your wealth at https://monetarymetals.com/impactTruemed: Check your eligibility and start saving at https://truemed.com/impactAT&T Business: Switch to AT&T Business at business.att.comIncogni: Take your personal data back with Incogni! Use code IMPACT at the link below and get 60% off an annual plan: https://incogni.com/impactShopify: Sign up for your one-dollar-per-month trial period at https://shopify.com/impactNetsuite: Right now, get our free business guide, Demystifying AI, at https://NetSuite.com/TheoryQuo: Try for free PLUS get 20% off your first 6 months at https://quo.com/impact Learn more about your ad choices. Visit megaphone.fm/adchoices

When the world gets to be too much, contemplating the endless wonder and beauty of the cosmos can be a huge relief. After all, we're insignificant in the grand scale of space and time. But cosmic thinking can also teach us so much about ourselves. This week, Adam sits with Chanda Prescod-Weinstein, professor of physics and faculty member in women's and gender studies at the University of New Hampshire, to talk about the truths we uncover about ourselves when we search for the truths of the universe. Find Chanda's new book, The Edge of Space-Time: Particles, Poetry, and the Cosmic Dream Boogie, at factuallypod.com/books--SUPPORT THE SHOW ON PATREON: https://www.patreon.com/adamconoverSEE ADAM ON TOUR: https://www.adamconover.net/tourdates/SUBSCRIBE to and RATE Factually! on:» Apple Podcasts: https://podcasts.apple.com/us/podcast/factually-with-adam-conover/id1463460577» Spotify: https://open.spotify.com/show/0fK8WJw4ffMc2NWydBlDyJAbout Headgum: Headgum is an LA & NY-based podcast network creating premium podcasts with the funniest, most engaging voices in comedy to achieve one goal: Making our audience and ourselves laugh. Listen to our shows at https://www.headgum.com.» SUBSCRIBE to Headgum: https://www.youtube.com/c/HeadGum?sub_confirmation=1» FOLLOW us on Twitter: http://twitter.com/headgum» FOLLOW us on Instagram: https://instagram.com/headgum/» FOLLOW us on TikTok: https://www.tiktok.com/@headgum» Advertise on Factually! via Gumball.fmSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Erin and Alyssa check in on the latest Bravo-level drama from Trump's wack job cabinet, two recent chilling tragedies in Virginia and Louisiana, Planned Parenthood's foray into cosmetic offerings, Reese Witherspoon's suspicious call for women to use more AI, and more. Then professor Chanda Prescod-Weinstein drops by to talk about her new book, The Edge of Space-Time, what people are getting wrong about the Artemis II mission, and what Star Trek and Octavia Butler can tell us about our current political moment.For a closed-captioned version of this episode, click here. For a transcript of this episode, please email transcripts@crooked.com and include the name of the podcast.The FBI Director Is MIA (The Atlantic 4/17)FBI director Kash Patel files $250M defamation lawsuit against The Atlantic (CNN 4/20)Labor Dept. Investigates Texts Among Secretary's Family and Staff (NYT 4/15)Feud between Mace and Mills flares as the Republicans trade barbs, expulsion threats (CNN 4/21)Ex-Virginia deputy governor kills wife and himself, police say (BBC 4/17)Haunted by ‘Dark Thoughts,' Louisiana Father Kills 8 Children (NYT 4/19)The Shreveport Mass Killing Isn't Just About ‘Mental Health' by Brittany Cooper (The Cut 4/20)A Planned Parenthood Clinic, in a Pinch, Turns to Botox (NYT 3/11)The Woman Who Knows Too Much: An Interview with Amanda Ungaro (Courier 4/18)Reese Witherspoon Declares “It's Time” For Women To Embrace AI: “Want To Learn With Me?” (Deadline 4/17)We Need Space w. Dr. Chanda Prescod-Weinstein