Kürzlich aktualisierte Kanäle

Wählen, um nur diesen Kanal zu sehenAlle Kanäle →

Ergebnisse

68 Artikel
Sortieren
32 Folien1:50:48EN
Y Combinator· 24. Okt. 2026Why Inference Matters 12× More Than TrainingFrançois Chaubard, Dan Fu, Jeff Morgan, Joe Fioti, Liangsheng Yin · Y CombinatorYC Paper Club walks the inference stack from an API request down to GPU kernels. Four talks from Together AI, Ollama, Luminal, and SGLang argue that serving quality, not just training FLOPs, is where most of the compute and the money now sit.InferenceGPU programming & kernelsKV caching & prefix reuse+2
2:00:51EN
Cognitive Revolution "How AI Changes Everything"· 10. Okt. 2026PodcastResample Every Token and You Can Watch the Model DecideModeration: Nathan LabenzGoodfire's Eric Bigelow on forking paths, in-context learning, and why chain-of-thought monitoring is getting harder to trust.AgentenPrompting & in-context learningReasoning & test-time compute+3
1:09:44EN
No Priors: AI, Machine Learning, Tech, & Startups· 9. Okt. 2026Podcast · 2 Stimmen10,000 GB300s for Four Weeks: Reflection’s Open-Model LedgerModeration: Sarah Guo, Elad GilMisha Laskin tells No Priors how Reflection trained Beam, why reinforcement learning outspent pre-training, and why the West needs its own open weights.Open Source & CommunityOpen-weight modelsReinforcement Learning+3
56:31EN
Unsupervised Learning: With Jacob Effron· 6. Okt. 2026PodcastPost-Training Wins Inference: Why Owning the Weights Beats Betting on One LabModeration: Jacob EffronApplied Compute CEO Yash Patil argues that control, private evals, and post-training on the biggest workloads matter more than fearing that frontier labs will turn malicious.Modelle & TrainingPost-trainingReinforcement Learning+3
2:06:11EN
Cognitive Revolution "How AI Changes Everything"· 14. Aug. 2026PodcastLindy Runs on DeepSeek. Its CEO Still Wants Chinese Models BannedModeration: Nathan LabenzFlo Crivello tells Nathan Labenz how Lindy Teammate lives in Slack, reaches two billion tokens of memory in two model calls, and why he would still outlaw the cheap Chinese models that currently power the product.AgentenAgent memoryContext Engineering+3
2:16:29EN
Cognitive Revolution "How AI Changes Everything"· 2. Aug. 2026PodcastFrom Two Papers a Month to Sixty: What Two Weeks in China Revealed About AI SafetyModeration: Nathan LabenzNathan Labenz argues that the familiar 'China will never slow down' line does not survive contact with the actual Chinese AI safety ecosystem.SicherheitSicherheit & GovernanceAI security+3
51:31EN
Sequoia Capital· 21. Juli 2026PodcastThey Refunded Nearly $2 Million. Then Factory’s CEO Explained the Dark FactoryMatan Grinberg on being two years early, why a model-agnostic harness beats lab co-design, and why 90 percent of coding tokens will soon run with the lights off.Coding-AgentenHarness engineeringModel routing & gateways+3
1:47:38EN
Cognitive Revolution "How AI Changes Everything"· 4. Juli 2026Podcast19 Neurons Can Drive a Car. Liquid AI Wants Foundation Models on Every ChipModeration: Nathan LabenzRamin Hasani traces liquid neural networks from a worm's 300-cell brain to a 600-megabyte model that will talk from inside a Mercedes, and explains why architecture search now beats architecture avengers.Modelle & TrainingLocal inferenceScaling Laws+3
1:06:36EN
Unsupervised Learning: With Jacob Effron· 12. Juni 2026PodcastBills Doubled Overnight. Next, the APIs Might Vanish.Moderation: Jacob EffronSix months after their last roundup, Ari Morcos and Rob Toews tell Jacob Efron why coding agents finally work, why near-frontier open weights may have peaked, and why a compute famine could pull the plug on public APIs.Unternehmen & StrategieAI StrategyOpen-source ecosystems+3
42:57EN
No Priors: AI, Machine Learning, Tech, & Startups· 1. Mai 2026Podcast · 2 StimmenInference Is the Last Market: Baseten’s 30x YearModeration: Sarah Guo, Elad GilCEO Tuhin Srivastava on a billion-dollar inference cloud, why the app layer survives, and a GPU crunch worse than the narrative.Infrastruktur & DeploymentInferencePost-training+3
54:26EN
Cognitive Revolution "How AI Changes Everything"· 22. Feb. 2026PodcastTen Billion Active Parameters, and the Unsexy Work of Making Them Agent-ReadyModeration: Nathan LabenzMiniMax researcher Olive Song explains how M2 was trained with expert developers as reward models, interleaved thinking, and a perturbation pipeline that treats every scaffold change as the real generalization test.Coding-AgentenReinforcement LearningAgent Engineering+3
4:25:13EN
Lex Fridman· 31. Jan. 2026PodcastThe Architecture Barely Changed. Data, Cost, and Open Weights Did.Moderation: Lex FridmanNathan Lambert and Sebastian Raschka walk Lex Fridman through a year in which training methods and datasets moved more than model design, and 2026 looks like a fight over inference cost, agents, and who owns truly open models.Modelle & TrainingScaling LawsOpen-weight models+3
1:54:30EN
Cognitive Revolution "How AI Changes Everything"· 9. Jan. 2026PodcastFrom One in Ten Cancer Cells to One in a MillionModeration: Nathan LabenzNathan Labenz's solo AMA updates his son's lymphoma treatment, calls Claude Opus 4.5 software AGI, and ranks the live players of early 2026.Unternehmen & StrategieAI StrategyCoding-Agenten+3
1:16:21EN
No Priors: AI, Machine Learning, Tech, & Startups· 8. Jan. 2026Podcast · 2 StimmenJensen Huang’s Year-End Case That Token Costs Can Fall a Billion-Fold Without Killing JobsModeration: Sarah Guo, Elad GilOn No Priors, NVIDIA’s CEO uses radiology, open source, and a $2 trillion R&D budget to argue that 2025’s extremes missed the stack.Unternehmen & StrategieReasoning & test-time computeAI cost & performance+3
1:18:40EN
Cognitive Revolution "How AI Changes Everything"· 1. Jan. 2026PodcastWhen AI Becomes the Oil Well: Luke Drago on the Intelligence CurseModeration: Nathan LabenzThe Workshop Labs co-founder tells the FLI Podcast why a world of non-human production could recreate the resource curse — and how private data, open weights, and moonshot careers might still break it.Unternehmen & StrategieAI StrategySicherheit & Governance+3
1:32:38EN
Cognitive Revolution "How AI Changes Everything"· 6. Dez. 2025PodcastPolish Is 1% of Frontier Training Data, So Poland Built Its Own ModelModeration: Nathan LabenzMarek Kozlowski of Poland's National Information Processing Institute explains how PLLuM uses language adaptation, organic instructions, and on-prem small models to keep a mid-size country in control of its AI.Unternehmen & StrategieOpen-weight modelsData quality & curation+3
1:09:48EN
Unsupervised Learning: With Jacob Effron· 5. Dez. 2025PodcastChina's AI Labs Went From 15 to Six. The Compute Gap Is 15-to-1.Moderation: Jacob EffronJordan Schneider tells Unsupervised Learning that Chinese models were a private-capital story, that Beijing still thinks in hardware, and that the West's real edge is chips it can serve — not a state master plan.Unternehmen & StrategieGeopoliticsAI Strategy+3
1:25:54EN
Cognitive Revolution "How AI Changes Everything"· 3. Dez. 2025PodcastHours after training ends, Z.ai puts the weights on the internetModeration: Nathan LabenzZixuan Li of Z.ai tells ChinaTalk how a Chinese lab ships GLM, treats open weights as a ticket into Western ecosystems, and still sees itself as an upstart.Unternehmen & StrategieOpen-source ecosystemsOpen-weight models+3
43:09EN
Sequoia Capital· 9. Sept. 2025PodcastA $100 Arm and a $300 Robot: Hugging Face Wants an App Store for Physical AIThomas Wolf tells Sequoia's Training Data why LeRobot, cheap hardware, and open datasets are Hugging Face's bet that software developers can become roboticists.Open Source & CommunityRoboticsOpen-source ecosystems+2
47:18EN
No Priors: AI, Machine Learning, Tech, & Startups· 14. Aug. 2025Podcast · 2 StimmenThe Analyst Tracking 200 GPU Clouds Breaks Nvidia Into a Three-Headed DragonModeration: Sarah Guo, Elad GilSemiAnalysis founder Dylan Patel on OpenAI’s open-weight model, a coming neocloud wipeout, and why a chip startup has to beat Nvidia on hardware, networking, and software at once.Infrastruktur & DeploymentInferenceAI cost & performance+3
36:48EN
No Priors: AI, Machine Learning, Tech, & Startups· 31. Juli 2025Podcast · 2 StimmenThe Weekend DeepSeek Dropped, the White House CalledModeration: Sarah Guo, Elad GilSriram Krishnan walks No Priors through America’s 28-page AI Action Plan, a token-share definition of winning, and why the lead over China is too small to take for granted.Unternehmen & StrategieAI StrategyInfrastruktur & Deployment+3
2:14:55EN
Cognitive Revolution "How AI Changes Everything"· 26. Juli 2025PodcastFrom Character AI to a 1% Warning Shot: Gabriel Weil's Case for Liability as AI GovernanceModeration: Nathan LabenzA Touro law professor tells Nathan Labenz how negligence, strict liability, and punitive damages could make frontier labs treat public risk as a hit to their own balance sheet.SicherheitSicherheit & GovernanceAI Regulation+3
16 Folien23:39ZH
跟李沐学AI· 23. Juli 2025李沐:给文本大模型加入一千万小时语音,开源 Higgs Audio李沐 · Boson AI李沐介绍团队如何在文本大语言模型的训练里加入大量语音数据,做成能听能说的 AudioLLM,并把 Higgs Audio 开源。他讲了架构取舍、token 压缩、数据从哪来,以及怎么在 3090 上跑起来。Sprach-KIMultimodal AIModelle & Training+2
54:53EN
Unsupervised Learning: With Jacob Effron· 17. Juni 2025PodcastA Free Voting Board Burned $2 Million. Then It Raised $100 Million.Moderation: Jacob EffronDatabricks co-founder Ion Stoica explains why LMArena exists, why China is winning open-source models, and why reliability—not another scaling law—is still the thing holding AI back.Evals & ObservabilityEvalsBenchmarks+3
Mehr laden24 von 68 angezeigt