Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

I chatted with yash about MatX, chip design and where silicon designed for LLMs is headed (8:17) Tightly coupling SRAM and HBM on one chip (14:03) More MoE FLOPS, smaller KV cache load (16:08) Numerics: from 32-bit to 4-bit (19:02) Targeting both training and inference (22:14) Chip timelines (27:15)... show more

Reiner Pope

18,724 subscribers

71,855 görüntüleme • 4 ay önce •via X (Twitter)

Bilim & Teknoloji

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

New interview: Reiner Pope, co-founder/CEO of MatX A counterintuitive throughput insight: “Low latency means small batch sizes. That is just Little’s law. Memory occupancy in HBM is proportional to batch size. So you can actually fit longer contexts than you could if the latency were larger. Low latency is not just a usability win, it improves throughput.” We get into: • The hybrid SRAM + HBM bet, and why pipeline parallelism finally works • Why sparse MoE drives MatX to “the most interconnect of any announced product” • Why frontier labs are willing to bet on an AI ASIC startup • Memory-bandwidth-efficient attention, numerics, and what MatX publishes (and what it does not) • Why 95% of model-side news is noise for chip design • The biggest challenges ahead 00:00 “We left Google one week before ChatGPT” 00:24 Intro: who is MatX 01:17 Origin story: leaving Google for LLM chips 02:21 GPT-3 and the “too expensive” problem 04:25 Why buy hardware that is not a GPU 05:52 Overcoming the CUDA moat 08:46 Early investors 09:35 The name MatX 09:59 The chip: matrix multiply + hybrid SRAM/HBM 12:11 Why pipeline parallelism finally works 14:22 Reading papers and Google going dark 15:20 Research agenda: attention and numerics 17:06 Five specs and meeting customers where they are 19:24 Why frontier labs are the natural first customer 20:32 Workloads: training, prefill, decode 22:18 Little’s law and the throughput case for low latency 24:29 Interconnect and MoE topology 26:35 Inside the team: 100 people, full stack 28:32 Agentic AI: 95% noise for hardware 30:35 KV cache sizing in an agentic world 32:11 How MatX uses AI for chip design (Verilog + BlueSpec) 34:23 Go to market: proving credibility under NDA 35:12 Porting effort for frontier labs 36:34 Biggest skepticism: manufacturing at gigawatt scale 37:32 Hiring plug Vikram Sekar

New interview: Reiner Pope, co-founder/CEO of MatX A counterintuitive throughput insight: “Low latency means small batch sizes. That is just Little’s law. Memory occupancy in HBM is proportional to batch size. So you can actually fit longer contexts than you could if the latency were larger. Low latency is not just a usability win, it improves throughput.” We get into: • The hybrid SRAM + HBM bet, and why pipeline parallelism finally works • Why sparse MoE drives MatX to “the most interconnect of any announced product” • Why frontier labs are willing to bet on an AI ASIC startup • Memory-bandwidth-efficient attention, numerics, and what MatX publishes (and what it does not) • Why 95% of model-side news is noise for chip design • The biggest challenges ahead 00:00 “We left Google one week before ChatGPT” 00:24 Intro: who is MatX 01:17 Origin story: leaving Google for LLM chips 02:21 GPT-3 and the “too expensive” problem 04:25 Why buy hardware that is not a GPU 05:52 Overcoming the CUDA moat 08:46 Early investors 09:35 The name MatX 09:59 The chip: matrix multiply + hybrid SRAM/HBM 12:11 Why pipeline parallelism finally works 14:22 Reading papers and Google going dark 15:20 Research agenda: attention and numerics 17:06 Five specs and meeting customers where they are 19:24 Why frontier labs are the natural first customer 20:32 Workloads: training, prefill, decode 22:18 Little’s law and the throughput case for low latency 24:29 Interconnect and MoE topology 26:35 Inside the team: 100 people, full stack 28:32 Agentic AI: 95% noise for hardware 30:35 KV cache sizing in an agentic world 32:11 How MatX uses AI for chip design (Verilog + BlueSpec) 34:23 Go to market: proving credibility under NDA 35:12 Porting effort for frontier labs 36:34 Biggest skepticism: manufacturing at gigawatt scale 37:32 Hiring plug Vikram Sekar

Semi Doped

19,439 görüntüleme • 4 ay önce

Added 5 voice lines for each hero for situations when they are carrying the Spirit Jar and start stalling, standing still for too long, or delaying the delivery. (Timecodes for all heroes are below). #deadlock Abrams 00:00 - 00:32 Apollo 00:32 - 01:05 Bebop 01:05 - 01:42 Billy 01:42 - 02:07 Calico 02:07 - 02:39 Celeste 02:39 - 03:11 Paradox 03:11 - 03:50 Doorman 03:50 - 04:16 Drifter 04:16 - 04:49 Dynamo 04:49 - 05:22 Graves 05:22 - 05:55 Grey Talon 05:55 - 06:32 Haze 06:32 - 07:03 Holliday 07:03 - 07:40 Infernus 07:40 - 08:08 Ivy 08:08 - 08:34 Kelvin 08:34 - 09:09 Lady Geist 09:09 - 09:39 Lash 09:39 - 10:07 Sinclair 10:07 - 10:35 McGinnis 10:35 - 11:12 Mina 11:12 - 11:40 Mirage 11:40 - 12:08 Mo&Krill 12:08 - 12:32 Paige 12:32 - 12:58 Pocket 12:58 - 13:25 Rem 13:25 - 13:56 Seven 13:56 - 14:27 Shiv 14:27 - 15:01 Silver 15:01 - 15:39 Venator 15:39 - 16:10 Victor 16:10 - 16:42 Vindicta 16:42 - 17:21 Viscous 17:21 - 17:46 Vyper 17:46 - 18:18 Warden 18:18 - 18:47 Wraith 18:47 - 19:13 Yamato 19:13 - 19:47

Added 5 voice lines for each hero for situations when they are carrying the Spirit Jar and start stalling, standing still for too long, or delaying the delivery. (Timecodes for all heroes are below). #deadlock Abrams 00:00 - 00:32 Apollo 00:32 - 01:05 Bebop 01:05 - 01:42 Billy 01:42 - 02:07 Calico 02:07 - 02:39 Celeste 02:39 - 03:11 Paradox 03:11 - 03:50 Doorman 03:50 - 04:16 Drifter 04:16 - 04:49 Dynamo 04:49 - 05:22 Graves 05:22 - 05:55 Grey Talon 05:55 - 06:32 Haze 06:32 - 07:03 Holliday 07:03 - 07:40 Infernus 07:40 - 08:08 Ivy 08:08 - 08:34 Kelvin 08:34 - 09:09 Lady Geist 09:09 - 09:39 Lash 09:39 - 10:07 Sinclair 10:07 - 10:35 McGinnis 10:35 - 11:12 Mina 11:12 - 11:40 Mirage 11:40 - 12:08 Mo&Krill 12:08 - 12:32 Paige 12:32 - 12:58 Pocket 12:58 - 13:25 Rem 13:25 - 13:56 Seven 13:56 - 14:27 Shiv 14:27 - 15:01 Silver 15:01 - 15:39 Venator 15:39 - 16:10 Victor 16:10 - 16:42 Vindicta 16:42 - 17:21 Viscous 17:21 - 17:46 Vyper 17:46 - 18:18 Warden 18:18 - 18:47 Wraith 18:47 - 19:13 Yamato 19:13 - 19:47

Deadlock UA

176,606 görüntüleme • 1 ay önce

Reiner Pope (MatX) just raised a $500m round led by Leopold Aschenbrenner and Jane Street to build faster AI chips. I enjoyed having him on Cheeky Pint so I could ask all my questions about how chip design actually works, where the speed-up comes from, and how the industry will evolve. 00:00:15 Google’s AI revival 00:07:54 MatX 00:17:11 AI supply chain 00:21:48 Designing chips 00:37:11 TSMC 00:44:17 Token pricing 00:44:55 RL-ing chip design 00:49:26 Design to production 00:56:05 MatX culture 01:02:57 Rust 01:05:21 Cuckoo hashing 01:09:35 Unexplored model architectures

Reiner Pope (MatX) just raised a $500m round led by Leopold Aschenbrenner and Jane Street to build faster AI chips. I enjoyed having him on Cheeky Pint so I could ask all my questions about how chip design actually works, where the speed-up comes from, and how the industry will evolve. 00:00:15 Google’s AI revival 00:07:54 MatX 00:17:11 AI supply chain 00:21:48 Designing chips 00:37:11 TSMC 00:44:17 Token pricing 00:44:55 RL-ing chip design 00:49:26 Design to production 00:56:05 MatX culture 01:02:57 Rust 01:05:21 Cuckoo hashing 01:09:35 Unexplored model architectures

John Collison

53,613 görüntüleme • 5 ay önce

#카리나 #KARINA 춤 챌린지 모음 (2024.12 ver.) 00:00 Weekend 00:19 Second 00:43 Sparkling 01:04 Sneakers 01:16 Forever 1 01:37 Candy 01:59 나의 바람 (Wind And Wish) 02:11 UNFORGIVEN 02:27 퀸카 (Queencard) 02:34 FREEDOM 02:46 샤랄라 (SHALALA) 03:33 HARD 03:57 Vuja De 04:15 Picture 04:26 NEW DANCE 04:50 Baggy Jeans 05:14 Everyday 05:29 Guilty 06:08 Baddie 06:30 Perfect Night 07:02 Why 07:26 VROOM VROOM 07:47 Chill Kill 08:03 MANIAC 08:23 Sweet Venom 08:39 Grand Prix 09:02 첫 눈 09:13 Rodeo 09:24 Be There For Me 09:36 Underneath the Tree 09:52 Single Ladies 10:16 Two of Hearts 10:26 UNTOUCHABLE 10:46 Good Morning 11:01 Thot Shit 11:13 Grillz 11:27 Mr.Bean 11:43 갈매기 11:55 To. X 12:11 Tap 12:36 마녀배달부 키키 12:54 yeyeye 13:07 Rolex 13:23 GGB 13:32 See Tình 13:48 KARINA's bounce 14:04 Earth, Wind & Fire 14:23 Feel the POP 14:40 Mr. vampire 14:55 yeyeye

#카리나 #KARINA 춤 챌린지 모음 (2024.12 ver.) 00:00 Weekend 00:19 Second 00:43 Sparkling 01:04 Sneakers 01:16 Forever 1 01:37 Candy 01:59 나의 바람 (Wind And Wish) 02:11 UNFORGIVEN 02:27 퀸카 (Queencard) 02:34 FREEDOM 02:46 샤랄라 (SHALALA) 03:33 HARD 03:57 Vuja De 04:15 Picture 04:26 NEW DANCE 04:50 Baggy Jeans 05:14 Everyday 05:29 Guilty 06:08 Baddie 06:30 Perfect Night 07:02 Why 07:26 VROOM VROOM 07:47 Chill Kill 08:03 MANIAC 08:23 Sweet Venom 08:39 Grand Prix 09:02 첫 눈 09:13 Rodeo 09:24 Be There For Me 09:36 Underneath the Tree 09:52 Single Ladies 10:16 Two of Hearts 10:26 UNTOUCHABLE 10:46 Good Morning 11:01 Thot Shit 11:13 Grillz 11:27 Mr.Bean 11:43 갈매기 11:55 To. X 12:11 Tap 12:36 마녀배달부 키키 12:54 yeyeye 13:07 Rolex 13:23 GGB 13:32 See Tình 13:48 KARINA's bounce 14:04 Earth, Wind & Fire 14:23 Feel the POP 14:40 Mr. vampire 14:55 yeyeye

감자

219,376 görüntüleme • 1 yıl önce

Our conversation with MrBeast in full on X Let's see how it performs cc: Elon Musk Timestamps: 00:00 - Intro 00:44 - Fame 09:56 - MrBeasts Most Viral Video Ever 11:22 - Focus 12:11 - Fitness 14:18 - Stress 15:46 - Belief in YouTube 17:10 - Messages from MrBeast's Past 19:22 - Improving on YouTube 22:01 - Pete Davidson and Tom Brady 25:02 - Social Experiments 27:50 - Storytelling 32:19 - MrBeast vs Netflix 38:15 - Problem with YouTube, Beastification 43:18 - Work Life Balance and the New MrBeast 45:32 - Competition 48:22 - Secret to Short Form Content 55:51 - MrBeast of TikTok 57:25 - Feastables 1:07:56 - Money and Creativity 1:09:25 - Artificial Intelligence 1:12:15 - Voice Dubbing 1:14:25 - Next YouTube Icon 1:15:42 - Ryan Trahan 1:17:27 - Stress and Innovation 1:24:08 - Leaving the Creator World 1:28:53 - MrBeast the Mega Billionaire 1:30:22 - Marriage 1:31:40 - Growing Up 1:37:03 - What MrBeast Watches on TV 1:41:30 - Performance Anxiety 1:45:54 - YouTube's Next Chapter 1:50:58 - MrBeast Burger and documenting life

Our conversation with MrBeast in full on X Let's see how it performs cc: Elon Musk Timestamps: 00:00 - Intro 00:44 - Fame 09:56 - MrBeasts Most Viral Video Ever 11:22 - Focus 12:11 - Fitness 14:18 - Stress 15:46 - Belief in YouTube 17:10 - Messages from MrBeast's Past 19:22 - Improving on YouTube 22:01 - Pete Davidson and Tom Brady 25:02 - Social Experiments 27:50 - Storytelling 32:19 - MrBeast vs Netflix 38:15 - Problem with YouTube, Beastification 43:18 - Work Life Balance and the New MrBeast 45:32 - Competition 48:22 - Secret to Short Form Content 55:51 - MrBeast of TikTok 57:25 - Feastables 1:07:56 - Money and Creativity 1:09:25 - Artificial Intelligence 1:12:15 - Voice Dubbing 1:14:25 - Next YouTube Icon 1:15:42 - Ryan Trahan 1:17:27 - Stress and Innovation 1:24:08 - Leaving the Creator World 1:28:53 - MrBeast the Mega Billionaire 1:30:22 - Marriage 1:31:40 - Growing Up 1:37:03 - What MrBeast Watches on TV 1:41:30 - Performance Anxiety 1:45:54 - YouTube's Next Chapter 1:50:58 - MrBeast Burger and documenting life

Colin and Samir ✌🏼✌🏾

124,600 görüntüleme • 2 yıl önce

#운학 2024 챌린지 모음 00:00 드라마 00:17 허니 00:30 하입보이 00:47 Standing next to you 01:03 Dash 01:20 굿모닝 01:36 흠뻑 02:05 Mr. Vampire 02:33 Super Lady 02:55 Like I Do 03:17 때깔 03:32 Easy 03:47 잭팟 04:02 사랑스러워 04:15 구구단송 / 꽃집 챌린지 04:29 하이틴 04:45 Fadeaway 05:04 Obsessed 05:13 Nectar 05:31 마그네틱 05:51 👟⚡️ 06:05 스무디 06:30 Bad 06:55 Deja Vu 07:27 청춘에게 07:40 포켓몬 07:53 SPOT! 08:10 사이렌 08:27 드라마 08:42 길티 09:04 겟어기타 09:22 I'm So Sexy 09:49 진격의 방탄 10:07 giri giri 10:23 뉴스쿨 올드스쿨 10:41 XO 10:53 Sweetest Tune 11:18 티라미수 케익 11:33 댄스홀 11:54 터치 12:22 Lay it down 12:37 브링방방 12:51 Dollars 13:05 Pineapple Slice 13:27 ATE THAT 13:48 Dancing on the road 14:13 GGUM 14:31 Goodies 14:50 겟라우드 15:14 노메이크업 15:28 펑키 글리터 크리스마스

#운학 2024 챌린지 모음 00:00 드라마 00:17 허니 00:30 하입보이 00:47 Standing next to you 01:03 Dash 01:20 굿모닝 01:36 흠뻑 02:05 Mr. Vampire 02:33 Super Lady 02:55 Like I Do 03:17 때깔 03:32 Easy 03:47 잭팟 04:02 사랑스러워 04:15 구구단송 / 꽃집 챌린지 04:29 하이틴 04:45 Fadeaway 05:04 Obsessed 05:13 Nectar 05:31 마그네틱 05:51 👟⚡️ 06:05 스무디 06:30 Bad 06:55 Deja Vu 07:27 청춘에게 07:40 포켓몬 07:53 SPOT! 08:10 사이렌 08:27 드라마 08:42 길티 09:04 겟어기타 09:22 I'm So Sexy 09:49 진격의 방탄 10:07 giri giri 10:23 뉴스쿨 올드스쿨 10:41 XO 10:53 Sweetest Tune 11:18 티라미수 케익 11:33 댄스홀 11:54 터치 12:22 Lay it down 12:37 브링방방 12:51 Dollars 13:05 Pineapple Slice 13:27 ATE THAT 13:48 Dancing on the road 14:13 GGUM 14:31 Goodies 14:50 겟라우드 15:14 노메이크업 15:28 펑키 글리터 크리스마스

꼼이

62,198 görüntüleme • 8 ay önce

Updated: Vocals & Rap - #성화 #Seonghwa 00:00 Still Here 00:15 Wonderland 00:36 Halazia 00:47 Answer 01:21 Cyberpunk 01:27 Wave 01:37 Utopia 01:50 My Way 02:00 With U 02:16 Be With You 02:54 Star 1117 03:21 Turbulence (야간비행) 03:39 Days 04:26 Promise 04:38 Fever 04:54 Sunrise 05:19 Mist 05:43 Take Me Home 06:11 Better 06:34 All About You 06:43 Aurora 06:55 Blue Summer 07:23 Diamond 07:46 Time Of Love 07:58 Stay 08:07 Desire 08:21 WDIG 08:28 New World 08:35 Dreamers 08:53 Not Too Late 09:03 The Letter 09:19 One Day At A Time 09:28 Dazzling Light 09:40 Light 09:50 The Ring 10:04 Paradigm 10:18 This World 10:38 Bouncy 10:51 Not Okay 11:15 The King 11:23 Inception 11:52 Fireworks 12:03 Limitless 12:19 Precious 12:28 Rocky 12:41 Let’s Get Together 13:07 White Love 13:17 The Leaders 13:27 Crazy Form 13:43 Arriba 13:57 Silver Light 14:06 Dreamy Day 14:26 Matz 15:15 Outlaw 15:23 Twilight 15:31 Dune 15:38 Hala Hala 15:43 Deja Vu 15:50 Horizon 15:54 Say My Name 15:57 Sector 1 16:02 Good Lil Boy 16:09 Don’t Stop 16:12 Still Here (Acoustic Ver) 16:19 Siren 16:31 Guerrilla 16:45 We Know 17:00 Empty Box 17:28 Wake Up 17:40 Dancing Like Butterfly Wings 17:49 Work 18:09 Thanxx 18:26 If Without U 18:45 Shaboom 19:01 Outlaw 19:18 Treasure 19:33 Eternal Sunshine 20:03 Django 20:31 Turbulence Japanese Ver. 20:55 Blind 21:16 Pirate King 21:24 Forevermore 22:00 Birthday 22:11 Deep Dive 22:26 Ice On My Teeth 22:50 Man On Fire 23:10 Selfish Waltz 23:25 Enough

Updated: Vocals & Rap - #성화 #Seonghwa 00:00 Still Here 00:15 Wonderland 00:36 Halazia 00:47 Answer 01:21 Cyberpunk 01:27 Wave 01:37 Utopia 01:50 My Way 02:00 With U 02:16 Be With You 02:54 Star 1117 03:21 Turbulence (야간비행) 03:39 Days 04:26 Promise 04:38 Fever 04:54 Sunrise 05:19 Mist 05:43 Take Me Home 06:11 Better 06:34 All About You 06:43 Aurora 06:55 Blue Summer 07:23 Diamond 07:46 Time Of Love 07:58 Stay 08:07 Desire 08:21 WDIG 08:28 New World 08:35 Dreamers 08:53 Not Too Late 09:03 The Letter 09:19 One Day At A Time 09:28 Dazzling Light 09:40 Light 09:50 The Ring 10:04 Paradigm 10:18 This World 10:38 Bouncy 10:51 Not Okay 11:15 The King 11:23 Inception 11:52 Fireworks 12:03 Limitless 12:19 Precious 12:28 Rocky 12:41 Let’s Get Together 13:07 White Love 13:17 The Leaders 13:27 Crazy Form 13:43 Arriba 13:57 Silver Light 14:06 Dreamy Day 14:26 Matz 15:15 Outlaw 15:23 Twilight 15:31 Dune 15:38 Hala Hala 15:43 Deja Vu 15:50 Horizon 15:54 Say My Name 15:57 Sector 1 16:02 Good Lil Boy 16:09 Don’t Stop 16:12 Still Here (Acoustic Ver) 16:19 Siren 16:31 Guerrilla 16:45 We Know 17:00 Empty Box 17:28 Wake Up 17:40 Dancing Like Butterfly Wings 17:49 Work 18:09 Thanxx 18:26 If Without U 18:45 Shaboom 19:01 Outlaw 19:18 Treasure 19:33 Eternal Sunshine 20:03 Django 20:31 Turbulence Japanese Ver. 20:55 Blind 21:16 Pirate King 21:24 Forevermore 22:00 Birthday 22:11 Deep Dive 22:26 Ice On My Teeth 22:50 Man On Fire 23:10 Selfish Waltz 23:25 Enough

Hwanakin

62,764 görüntüleme • 1 yıl önce

Etched is deploying two new technologies in chip design: low-voltage inference and cluster-scale memory. CEO Gavin Uberti says they'll make their chips much more power-efficient and way, way faster than today's leading GPUs. He breaks it down: "We looked at a lot of early research directions, and we realized the key things that models need are way more compute and way faster memory." "If you think about inference, there are two key parts: prefill and decode. For prefill, it's a compute-bound problem. You need to have more FLOPS, more operations per second on each of your chips." "On our GPU, the bottleneck's actually thermals. You can't really run a GPU at more than around 50% of what it could theoretically do, or it'll melt." "So we're using a new technology today called low-voltage inference to try to solve this problem. You bring the voltage of the chip down dramatically, which allows us to have way, way better efficiency in terms of how much power is drawn per unit of math, and thus fit way way more flops onto the chip..." "For decode, it's all about bandwidth. Not just bandwidth on a chip, but bandwidth across your cluster. That's why we have this technology we call cluster-scale memory. It reduces the amount of time it takes to communicate from one chip to another dramatically." "As a result we can go use all of our HBM, HBM bandwidth, SRAM, SRAM bandwidth, and our scale-up domain as a single coherent pool. And that means if you're a user, you can go get much faster tokens per second, while still keeping your costs low."

Etched is deploying two new technologies in chip design: low-voltage inference and cluster-scale memory. CEO Gavin Uberti says they'll make their chips much more power-efficient and way, way faster than today's leading GPUs. He breaks it down: "We looked at a lot of early research directions, and we realized the key things that models need are way more compute and way faster memory." "If you think about inference, there are two key parts: prefill and decode. For prefill, it's a compute-bound problem. You need to have more FLOPS, more operations per second on each of your chips." "On our GPU, the bottleneck's actually thermals. You can't really run a GPU at more than around 50% of what it could theoretically do, or it'll melt." "So we're using a new technology today called low-voltage inference to try to solve this problem. You bring the voltage of the chip down dramatically, which allows us to have way, way better efficiency in terms of how much power is drawn per unit of math, and thus fit way way more flops onto the chip..." "For decode, it's all about bandwidth. Not just bandwidth on a chip, but bandwidth across your cluster. That's why we have this technology we call cluster-scale memory. It reduces the amount of time it takes to communicate from one chip to another dramatically." "As a result we can go use all of our HBM, HBM bandwidth, SRAM, SRAM bandwidth, and our scale-up domain as a single coherent pool. And that means if you're a user, you can go get much faster tokens per second, while still keeping your costs low."

TBPN

20,404 görüntüleme • 1 ay önce

Anthropic co-founder Jared Kaplan on scaling and the road to human-level AI. At AI Startup School in San Francisco. 00:17 - From Physics to AI 01:41 - Initial Skepticism and Shift to AI 02:12 - AI Training Phases 02:32 - Pre-Training 03:16 - Reinforcement Learning 04:19 - Scaling Laws in Training 08:19 - Unlocking AI Capabilities 11:27 - Organizational Knowledge and Memory 12:19 - Oversight and Nuanced Tasks 13:38 - Preparing for the Future 15:48 - Claude 4 and Beyond 21:18 - Human-AI Collaboration 29:50 - Scaling Laws and Compute Efficiency 35:26 - Audience Q&A

Anthropic co-founder Jared Kaplan on scaling and the road to human-level AI. At AI Startup School in San Francisco. 00:17 - From Physics to AI 01:41 - Initial Skepticism and Shift to AI 02:12 - AI Training Phases 02:32 - Pre-Training 03:16 - Reinforcement Learning 04:19 - Scaling Laws in Training 08:19 - Unlocking AI Capabilities 11:27 - Organizational Knowledge and Memory 12:19 - Oversight and Nuanced Tasks 13:38 - Preparing for the Future 15:48 - Claude 4 and Beyond 21:18 - Human-AI Collaboration 29:50 - Scaling Laws and Compute Efficiency 35:26 - Audience Q&A

Y Combinator

116,444 görüntüleme • 1 yıl önce

I sat down with Silas Alberti, Head of Research Cognition, to chat about the intersection of RL and inference -- from training frontier coding models to running inference at scale. Watch till the end for bonus doggo! :o 0:00 — Intros 1:04 — What's hardest to get right in an RL run 3:39 — How Cognition started using Modal 4:03 — What to weigh when buying inference 4:58 — The latency/throughput Pareto frontier 7:14 — DFlash reveal (!!) 8:44 — Why faster runtime matters 9:56 — Memory-bound vs. compute-bound 10:57 — Tree-based speculation 14:20 — Hot take: RL and inference optimization converging 15:28 — Online training of the speculator 16:22 — "Auto inference" 17:27 — Agents competing on optimization problems 18:27 — What inference providers still get wrong

I sat down with Silas Alberti, Head of Research Cognition, to chat about the intersection of RL and inference -- from training frontier coding models to running inference at scale. Watch till the end for bonus doggo! :o 0:00 — Intros 1:04 — What's hardest to get right in an RL run 3:39 — How Cognition started using Modal 4:03 — What to weigh when buying inference 4:58 — The latency/throughput Pareto frontier 7:14 — DFlash reveal (!!) 8:44 — Why faster runtime matters 9:56 — Memory-bound vs. compute-bound 10:57 — Tree-based speculation 14:20 — Hot take: RL and inference optimization converging 15:28 — Online training of the speculator 16:22 — "Auto inference" 17:27 — Agents competing on optimization problems 18:27 — What inference providers still get wrong

gongy

78,284 görüntüleme • 21 gün önce

2025 #지훈 챌린지 & 릴스 모음💝 00:00 sticky 00:16 샹하이 로맨스 00:40 Boom Boom Base (remix ver.) 01:01 오늘만 I LOVE YOU 01:30 💥💥💥 01:44 Make A Wish 02:20 뱅(Bang) ! 02:39 Kiss Kiss Shy Shy 02:52 wall🪼❤️‍🔥 03:07 🍉🍓🍊👌 03:24 Sweet Dreams 03:52 Lucifer 04:21 WE LIKE 2 PARTY 04:51 かわいいだけじゃだめですか？ 05:08 Ayo 05:36 BEBE 05:56 NIN 06:10 Come on!⚡️ 06:24 GO 지훈🎉 GO 지훈🥳 06:40 🪼💸 07:07 BANDIT 07:32 KNOW ABOUT ME 07:53 뭘봐 08:15 Cool & hot 08:38 SOURPATCH 08:55 Chains 09:24 이뻐이뻐 09:55 행복 10:22 move like that🎵 10:40 Shake It To The Max✌️ 10:52 boy for the 42🩵 11:08 JIHOON Calling📞 11:13 I Feel Good 11:34 HANDS UP 12:12 ROCKSTAR 12:36 Thunder 12:59 Bad Desire 13:17 시끄러!! 13:26 Touch It🪼 13:43 good vibes☀️ 13:52 Killin' It Girl 14:16 ECHO! 14:41 🚘🕶️ 14:56 Buzzin💭 15:07 FAMOUS 15:18 빌려온 고양이 15:35 Domino 15:51 The Blue Sun 16:07 POP 16:32 シュガリードライブ 16:52 훈제란즈✌️✌️ 17:02 되고파 너의 17:17 🪼vs🪼 17:34 before sleep 🛌 17:44 Walk Like👟 17:54 🌼 18:02 Barbie Girl 18:11 HOWL 18:23 Tempo 18:43 🪼🔄 18:52 Dougie🤙 19:10 What You Want 19:37 😎😎 19:51 Bang🪼 19:58 ✌️✌️ 20:10 EZ 20:29 M.O. 20:55 I‘m right here 💐 21:12 산책 21:24 body 21:42 SPAGHETTI 22:15 Blue Valentine 22:50 Hollywood Action 23:17 BURNING UP 23:40 하얀 그리움 24:20 Break your legs👟 24:29 Let‘s 🪩 24:47 Talk to You 25:16 My Body 25:37 MISMATCH 26:02 Do you wanna dance?🎵 26:12 🦾 26:41 All I Want for Christmas Is You 26:58 Hmm🕶️ 27:08 ˏ₍⸜̠̇⸝̠̇₎ˎ ̀⁽⸌̠̇⸍̠̇⁾ ́ˏ₍⸜̠̇⸝̠̇₎ˎ ̀⁽⸌̠̇⸍̠̇⁾ ́ 27:22 You like it? 27:28 락 (樂) 27:48 Love Shot

2025 #지훈 챌린지 & 릴스 모음💝 00:00 sticky 00:16 샹하이 로맨스 00:40 Boom Boom Base (remix ver.) 01:01 오늘만 I LOVE YOU 01:30 💥💥💥 01:44 Make A Wish 02:20 뱅(Bang) ! 02:39 Kiss Kiss Shy Shy 02:52 wall🪼❤️‍🔥 03:07 🍉🍓🍊👌 03:24 Sweet Dreams 03:52 Lucifer 04:21 WE LIKE 2 PARTY 04:51 かわいいだけじゃだめですか？ 05:08 Ayo 05:36 BEBE 05:56 NIN 06:10 Come on!⚡️ 06:24 GO 지훈🎉 GO 지훈🥳 06:40 🪼💸 07:07 BANDIT 07:32 KNOW ABOUT ME 07:53 뭘봐 08:15 Cool & hot 08:38 SOURPATCH 08:55 Chains 09:24 이뻐이뻐 09:55 행복 10:22 move like that🎵 10:40 Shake It To The Max✌️ 10:52 boy for the 42🩵 11:08 JIHOON Calling📞 11:13 I Feel Good 11:34 HANDS UP 12:12 ROCKSTAR 12:36 Thunder 12:59 Bad Desire 13:17 시끄러!! 13:26 Touch It🪼 13:43 good vibes☀️ 13:52 Killin' It Girl 14:16 ECHO! 14:41 🚘🕶️ 14:56 Buzzin💭 15:07 FAMOUS 15:18 빌려온 고양이 15:35 Domino 15:51 The Blue Sun 16:07 POP 16:32 シュガリードライブ 16:52 훈제란즈✌️✌️ 17:02 되고파 너의 17:17 🪼vs🪼 17:34 before sleep 🛌 17:44 Walk Like👟 17:54 🌼 18:02 Barbie Girl 18:11 HOWL 18:23 Tempo 18:43 🪼🔄 18:52 Dougie🤙 19:10 What You Want 19:37 😎😎 19:51 Bang🪼 19:58 ✌️✌️ 20:10 EZ 20:29 M.O. 20:55 I‘m right here 💐 21:12 산책 21:24 body 21:42 SPAGHETTI 22:15 Blue Valentine 22:50 Hollywood Action 23:17 BURNING UP 23:40 하얀 그리움 24:20 Break your legs👟 24:29 Let‘s 🪩 24:47 Talk to You 25:16 My Body 25:37 MISMATCH 26:02 Do you wanna dance?🎵 26:12 🦾 26:41 All I Want for Christmas Is You 26:58 Hmm🕶️ 27:08 ˏ₍⸜̠̇⸝̠̇₎ˎ ̀⁽⸌̠̇⸍̠̇⁾ ́ˏ₍⸜̠̇⸝̠̇₎ˎ ̀⁽⸌̠̇⸍̠̇⁾ ́ 27:22 You like it? 27:28 락 (樂) 27:48 Love Shot

🏝️

103,918 görüntüleme • 7 ay önce

내가 보려고 모은 #조이 역대 브릿지 파트 모음.zip (Ice Cream Cake~Sweet Dreams) 💚 * 레드벨벳 앨범만 포함했습니다 00:00 Somethin Kinda Crazy 00:05 Take It Slow 00:13 사탕 (Candy) 00:20 Dumb Dumb 00:22 Campfire 00:31 Lady’s Room 00:33 Day 1 00:43 Coold World 00:51 My Dear 00:55 Bad Dracula 00:58 Happily Ever After 01:03 빨간 맛 (Red Flavor) 01:09 여름빛 (Mojito) 01:13 I Just 01:17 Kingdom Come 01:23 두 번째 데이트 (My Second Date) 01:27 Attaboy 01:36 Perfect 10 01:42 달빛소리 01:53 Bad Boy 01:59 Time To Love 02:09 #Cookie Jar 02:13 Aitai-tai 02:20 Power Up 02:23 한 여름의 크리스마스 (With You) 02:30 Mr.E 02:35 Mosquito 02:41 RBB (Really Bad Boy) 02:47 So Good 02:51 멋있게 (Sassy Me) 02:54 Swimming Pool 02:57 Sayonara 03:04 짐살라빔 (Zimzalabim) 03:12 Sunny Side Up! 03:16 Milkshake 03:21 친구가 아냐 (Bing Bing) 03:29 안녕, 여름 (Parade) 03:32 LP 03:41 음파음파 (Umpah Umpah) 03:48 카풀 (Carpool) 03:57 Love Is The Way 04:05 Jumpin’ 04:11 눈 맞추고, 손 맞대고 (Eyes Locked, Hand Locked) 04:17 Psycho 04:24 In & Out 04:34 Remember Forever 04:44 Milky Way 04:52 Queendom 05:00 Pose 05:07 Knock On Wood 05:15 Better Be 05:25 Feel My Rhythm 05:29 Rainbow Halo 05:36 Beg For Me 05:43 In My Dreams 05:48 Marionette 05:50 WILDSIDE 05:54 Jackpot 06:00 Snap Snap 06:08 ‘Cause it’s you 06:14 Color of Love 06:16 Birthday 06:19 BYE BYE 06:22 롤러코스터 (On A Ride) 06:29 Zoom 06:40 Celebrate 06:48 Chill Kill 06:56 Knock Knock (Who’s There?) 06:59 Nightmare 07:03 One Kiss 07:10 Wings 07:21 풍경화 (Scenery) 07:32 Sunflower 07:38 Love Arcade 07:44 Bubble 07:51 Night Drive 07:57 Sweet Dreams

내가 보려고 모은 #조이 역대 브릿지 파트 모음.zip (Ice Cream Cake~Sweet Dreams) 💚 * 레드벨벳 앨범만 포함했습니다 00:00 Somethin Kinda Crazy 00:05 Take It Slow 00:13 사탕 (Candy) 00:20 Dumb Dumb 00:22 Campfire 00:31 Lady’s Room 00:33 Day 1 00:43 Coold World 00:51 My Dear 00:55 Bad Dracula 00:58 Happily Ever After 01:03 빨간 맛 (Red Flavor) 01:09 여름빛 (Mojito) 01:13 I Just 01:17 Kingdom Come 01:23 두 번째 데이트 (My Second Date) 01:27 Attaboy 01:36 Perfect 10 01:42 달빛소리 01:53 Bad Boy 01:59 Time To Love 02:09 #Cookie Jar 02:13 Aitai-tai 02:20 Power Up 02:23 한 여름의 크리스마스 (With You) 02:30 Mr.E 02:35 Mosquito 02:41 RBB (Really Bad Boy) 02:47 So Good 02:51 멋있게 (Sassy Me) 02:54 Swimming Pool 02:57 Sayonara 03:04 짐살라빔 (Zimzalabim) 03:12 Sunny Side Up! 03:16 Milkshake 03:21 친구가 아냐 (Bing Bing) 03:29 안녕, 여름 (Parade) 03:32 LP 03:41 음파음파 (Umpah Umpah) 03:48 카풀 (Carpool) 03:57 Love Is The Way 04:05 Jumpin’ 04:11 눈 맞추고, 손 맞대고 (Eyes Locked, Hand Locked) 04:17 Psycho 04:24 In & Out 04:34 Remember Forever 04:44 Milky Way 04:52 Queendom 05:00 Pose 05:07 Knock On Wood 05:15 Better Be 05:25 Feel My Rhythm 05:29 Rainbow Halo 05:36 Beg For Me 05:43 In My Dreams 05:48 Marionette 05:50 WILDSIDE 05:54 Jackpot 06:00 Snap Snap 06:08 ‘Cause it’s you 06:14 Color of Love 06:16 Birthday 06:19 BYE BYE 06:22 롤러코스터 (On A Ride) 06:29 Zoom 06:40 Celebrate 06:48 Chill Kill 06:56 Knock Knock (Who’s There?) 06:59 Nightmare 07:03 One Kiss 07:10 Wings 07:21 풍경화 (Scenery) 07:32 Sunflower 07:38 Love Arcade 07:44 Bubble 07:51 Night Drive 07:57 Sweet Dreams

맑음쨩 🍏

20,678 görüntüleme • 1 yıl önce

THE 2023 BESTIE AWARDS! join the besties as the recap the best, worst, and craziest from this past year! (live on X) timestamps: (0:00) welcome to the fourth annual bestie awards! (4:14) biggest political winner (10:26) biggest political loser (15:14) biggest political surprise (23:02) biggest business winner (26:50) biggest business loser (30:32) biggest business surprise (35:57) best science breakthrough (40:30) biggest flash in the pan (42:34) best ceo (44:53) best investor (47:07) best turnaround (50:04) worst company (54:21) best meme (56:53) best new tech (59:57) best trend (1:02:12) worst trend (1:07:42) favorite media (1:12:40) the rudy giuliani award for self-immolation (1:17:02) final thoughts

THE 2023 BESTIE AWARDS! join the besties as the recap the best, worst, and craziest from this past year! (live on X) timestamps: (0:00) welcome to the fourth annual bestie awards! (4:14) biggest political winner (10:26) biggest political loser (15:14) biggest political surprise (23:02) biggest business winner (26:50) biggest business loser (30:32) biggest business surprise (35:57) best science breakthrough (40:30) biggest flash in the pan (42:34) best ceo (44:53) best investor (47:07) best turnaround (50:04) worst company (54:21) best meme (56:53) best new tech (59:57) best trend (1:02:12) worst trend (1:07:42) favorite media (1:12:40) the rudy giuliani award for self-immolation (1:17:02) final thoughts

The All-In Podcast

728,223 görüntüleme • 2 yıl önce

E133: Sam Blackshear - How Libra Sparked the Move Language and Why Sui Is the Real Endgame! Sam Blackshear is the Co-founder and CTO of MystenLabs.sui , the company behind the Sui, and the Creator of the Move programming language that's revolutionizing smart contract development. Timestamps: 0:00 Introduction 1:54 Partnerships: Jupiter, KAST (old), , Sui, Mantle, Forza! BTC 2:44 The Power of Preparation 5:14 Discipline Behind the Podcast 6:27 Translating Thought Into Code 8:19 Who is Sam Blackshear? 9:27 Choosing What Truly Matters 10:25 Self-Custody with Trezor 11:18 Crypto vs. AI Thinking 12:28 The Power of Support 16:04 From Court Dreams to Reality 17:55 Challenging the Limits of Code 22:52 Chose Learning Over a Job 24:09 The Internship That Changed Everything 27:35 PhD Skills Meet Facebook 29:05 Entering Crypto Through Facebook 32:15 Why Libra Needed Move 33:42 Solving Scarcity in Code 36:37 Bitcoin & Ethereum Mistakes 38:45 Creating a New Language 41:41 Problem-Driven Innovation 44:47 Avoiding Analysis Paralysis 48:11 What is Unstructured Thinking? 50:50 Why Unstructured Thinking Works 53:14 Future of Crypto Protocols 54:21 Why Move is the Best Programming Language 55:00 What is the Sui Network? 56:25 What Makes Sui Different? 57:05 Why is Sui The Best Blockchain? 59:05 Managing Energy Long-Term 1:01:02 Satisfaction Without Closure 1:03:37 90% Love, 10% Grind 1:05:37 Mental State of Surfing 1:06:47 Non-Consensus Beliefs 1:07:34 What is Memory Safety? 1:09:07 What Was the Equifax Hack? 1:10:53 Rethinking Software Safety 1:12:00 Right Dose of Regulation 1:13:00 Biggest Prediction for the Next 24 Months? 1:14:02 Scaling Crypto Developers 1:16:07 Concluding Remarks

E133: Sam Blackshear - How Libra Sparked the Move Language and Why Sui Is the Real Endgame! Sam Blackshear is the Co-founder and CTO of MystenLabs.sui , the company behind the Sui, and the Creator of the Move programming language that's revolutionizing smart contract development. Timestamps: 0:00 Introduction 1:54 Partnerships: Jupiter, KAST (old), , Sui, Mantle, Forza! BTC 2:44 The Power of Preparation 5:14 Discipline Behind the Podcast 6:27 Translating Thought Into Code 8:19 Who is Sam Blackshear? 9:27 Choosing What Truly Matters 10:25 Self-Custody with Trezor 11:18 Crypto vs. AI Thinking 12:28 The Power of Support 16:04 From Court Dreams to Reality 17:55 Challenging the Limits of Code 22:52 Chose Learning Over a Job 24:09 The Internship That Changed Everything 27:35 PhD Skills Meet Facebook 29:05 Entering Crypto Through Facebook 32:15 Why Libra Needed Move 33:42 Solving Scarcity in Code 36:37 Bitcoin & Ethereum Mistakes 38:45 Creating a New Language 41:41 Problem-Driven Innovation 44:47 Avoiding Analysis Paralysis 48:11 What is Unstructured Thinking? 50:50 Why Unstructured Thinking Works 53:14 Future of Crypto Protocols 54:21 Why Move is the Best Programming Language 55:00 What is the Sui Network? 56:25 What Makes Sui Different? 57:05 Why is Sui The Best Blockchain? 59:05 Managing Energy Long-Term 1:01:02 Satisfaction Without Closure 1:03:37 90% Love, 10% Grind 1:05:37 Mental State of Surfing 1:06:47 Non-Consensus Beliefs 1:07:34 What is Memory Safety? 1:09:07 What Was the Equifax Hack? 1:10:53 Rethinking Software Safety 1:12:00 Right Dose of Regulation 1:13:00 Biggest Prediction for the Next 24 Months? 1:14:02 Scaling Crypto Developers 1:16:07 Concluding Remarks

MR SHIFT 🦁

164,299 görüntüleme • 1 yıl önce

Steph Ango (kepano) is a designer, entrepreneur, and the CEO of Obsidian. In this episode, we discussed his journey from creating Winamp skins to leading Obsidian, and the fundamental patterns and inclinations that drive everything he explores and makes. We discussed Obsidian on multiple different dimensions, his deluxe chocolate chip cookie recipe, his exploration through furniture design, and his default inclination to push “too far” into anything he is exploring. Steph Ango demo'd the design process of several furniture projects he’s been working on at home as well as a custom piano learning app he created to teach himself music theory. This is a conversation about craftsmanship, pursuing a deep understanding, and what that reveals about who we are. The Other Stuff is hosted by internetVin — filmmaker, entrepreneur, and possibly the most curious man on Earth. Produced by New. The Other Stuff #28 — Kepano: The Interconnectedness of Everything — Timestamps 00:03:55 Origin of Kepano 00:10:34 Relationship with Computers 00:15:25 Online Communities & Video Games 00:30:47 Winamp Skins 00:40:04 Uniformity in Design 00:51:05 Obsidian 00:59:20 Bidirectional Linking, And The Relationship Between Everything 01:23:18 Obsidian Sync, Publish, and Bases 01:32:06 Why File Over App 01:43:15 The Deluxe Chocolate Chip Cookie Taste Test 01:49:48 Living Fully 02:09:50 Early Project Walkthrough 02:19:03 Furniture Design Walkthrough 02:49:00 Constraint and Patterns 02:58:47 Kepiano 03:13:00 The Future of AI and Startups 03:31:10 Closing

Steph Ango (kepano) is a designer, entrepreneur, and the CEO of Obsidian. In this episode, we discussed his journey from creating Winamp skins to leading Obsidian, and the fundamental patterns and inclinations that drive everything he explores and makes. We discussed Obsidian on multiple different dimensions, his deluxe chocolate chip cookie recipe, his exploration through furniture design, and his default inclination to push “too far” into anything he is exploring. Steph Ango demo'd the design process of several furniture projects he’s been working on at home as well as a custom piano learning app he created to teach himself music theory. This is a conversation about craftsmanship, pursuing a deep understanding, and what that reveals about who we are. The Other Stuff is hosted by internetVin — filmmaker, entrepreneur, and possibly the most curious man on Earth. Produced by New. The Other Stuff #28 — Kepano: The Interconnectedness of Everything — Timestamps 00:03:55 Origin of Kepano 00:10:34 Relationship with Computers 00:15:25 Online Communities & Video Games 00:30:47 Winamp Skins 00:40:04 Uniformity in Design 00:51:05 Obsidian 00:59:20 Bidirectional Linking, And The Relationship Between Everything 01:23:18 Obsidian Sync, Publish, and Bases 01:32:06 Why File Over App 01:43:15 The Deluxe Chocolate Chip Cookie Taste Test 01:49:48 Living Fully 02:09:50 Early Project Walkthrough 02:19:03 Furniture Design Walkthrough 02:49:00 Constraint and Patterns 02:58:47 Kepiano 03:13:00 The Future of AI and Startups 03:31:10 Closing

The Other Stuff Podcast

71,626 görüntüleme • 8 ay önce

"The world is probably going to be a better place without authority and trust." Gavin Wood on why he's spent over a decade building Web3. 0:00 - Childhood: Lego, self-taught programming 10:00 - Video games 14:02 - Board games 19:34 - Game design as blockchain architecture 25:11 - Philosophy on authority, living as peers 30:57 - Motivations for Web3 38:16 - JAM vs Polkadot development 42:48 - JAM technical: scalability 52:32 - JAM as neutral protocol 55:00 - AI dangers, Web3 as counterbalance 1:01:33 - Ethereum origin: Vitalik and whitepaper

"The world is probably going to be a better place without authority and trust." Gavin Wood on why he's spent over a decade building Web3. 0:00 - Childhood: Lego, self-taught programming 10:00 - Video games 14:02 - Board games 19:34 - Game design as blockchain architecture 25:11 - Philosophy on authority, living as peers 30:57 - Motivations for Web3 38:16 - JAM vs Polkadot development 42:48 - JAM technical: scalability 52:32 - JAM as neutral protocol 55:00 - AI dangers, Web3 as counterbalance 1:01:33 - Ethereum origin: Vitalik and whitepaper

Polkadot

292,437 görüntüleme • 8 ay önce

From Airbnb to the Trump Administration's DOGE initiative—my conversation with Airbnb co-founder Joe Gebbia covers his remarkable journey and what's next. 0:00 - Introduction 1:14 - Joe’s background 3:13 - First website 4:57 - Airbnb foudning 11:07 - First guests 15:04 - South by Southwest 17:48 - Trough of sorrow 18:56 - Selling cereal to keep dream alive 20:43 - Holding a grudge 21:13 - First investor pitch 23:54 - Building trust 24:37 - Greatest unforeseen challenge 26:59 - Favorite Airbnb 30:48 - Backing President Trump 32:19 - Macrobiotic diet 34:00 - Make America Healthy Again 34:37 - Biden’s border crisis 36:47 - America’s Chief Design Officer 39:43 - Improving the retirement process 41:46 - National Design Studio 45:46 - The private sector versus government 46:50 - First government website redesign 48:39 - Serving in the Trump Administration 50:37 - Being a billionaire 55:50 - Rapid fire questions 57:51 - Would You Rather 59:52 - Departing Airbnb 1:01:29 - Worst idea 1:02:07 - Best investment 1:02:38 - Genuine intentions 1:04:27 - Dream dinner party 1:06:02 - Next endeavor

From Airbnb to the Trump Administration's DOGE initiative—my conversation with Airbnb co-founder Joe Gebbia covers his remarkable journey and what's next. 0:00 - Introduction 1:14 - Joe’s background 3:13 - First website 4:57 - Airbnb foudning 11:07 - First guests 15:04 - South by Southwest 17:48 - Trough of sorrow 18:56 - Selling cereal to keep dream alive 20:43 - Holding a grudge 21:13 - First investor pitch 23:54 - Building trust 24:37 - Greatest unforeseen challenge 26:59 - Favorite Airbnb 30:48 - Backing President Trump 32:19 - Macrobiotic diet 34:00 - Make America Healthy Again 34:37 - Biden’s border crisis 36:47 - America’s Chief Design Officer 39:43 - Improving the retirement process 41:46 - National Design Studio 45:46 - The private sector versus government 46:50 - First government website redesign 48:39 - Serving in the Trump Administration 50:37 - Being a billionaire 55:50 - Rapid fire questions 57:51 - Would You Rather 59:52 - Departing Airbnb 1:01:29 - Worst idea 1:02:07 - Best investment 1:02:38 - Genuine intentions 1:04:27 - Dream dinner party 1:06:02 - Next endeavor

The Katie Miller Podcast

1,266,089 görüntüleme • 11 ay önce

This is my sixth conversation with Gavin Baker. As always with Gavin, the conversation covers a lot of ground, but we spend the most time on watts and wafers. We discuss: - Why the wafer shortage may prevent an AI bubble - Data centers in space (reframed) - Elon's Terafab and the new chip companies challenging Nvidia - Usage-based pricing - The disaggregation of GPUs - DRAM, frontier tokens, and open source Enjoy! Timestamps: 0:00 Intro 7:55 Anthropic and OpenAI Valuations 12:58 Watts, Wafers, and Infrastructure 14:39 Orbital Compute and Data Centers in Space 22:49 Avoiding the AI Bubble 28:26 Terafab and the Future of US Manufacturing 32:16 Returns to the Frontier 37:23 Continual Learning 42:03 New Chip Companies 48:52 Extending GPU Lifespans and Private Credit 51:22 The Application Layer 57:32 The Token Path and Open-Source Dynamics 1:01:37 Cybersecurity 1:05:46 Diversity Breakdown 1:11:59 Assessing the Big Tech Players in AI 1:19:02 Geopolitics, Personal Safety, and the AI Horizon

This is my sixth conversation with Gavin Baker. As always with Gavin, the conversation covers a lot of ground, but we spend the most time on watts and wafers. We discuss: - Why the wafer shortage may prevent an AI bubble - Data centers in space (reframed) - Elon's Terafab and the new chip companies challenging Nvidia - Usage-based pricing - The disaggregation of GPUs - DRAM, frontier tokens, and open source Enjoy! Timestamps: 0:00 Intro 7:55 Anthropic and OpenAI Valuations 12:58 Watts, Wafers, and Infrastructure 14:39 Orbital Compute and Data Centers in Space 22:49 Avoiding the AI Bubble 28:26 Terafab and the Future of US Manufacturing 32:16 Returns to the Frontier 37:23 Continual Learning 42:03 New Chip Companies 48:52 Extending GPU Lifespans and Private Credit 51:22 The Application Layer 57:32 The Token Path and Open-Source Dynamics 1:01:37 Cybersecurity 1:05:46 Diversity Breakdown 1:11:59 Assessing the Big Tech Players in AI 1:19:02 Geopolitics, Personal Safety, and the AI Horizon

Patrick OShaughnessy

1,676,303 görüntüleme • 3 ay önce

My conversation with Kanishka Rao, who went from working on particle physics at CERN to driving robotics at Google DeepMind. We get into the Gemini Robotics 2 release, trending training recipes in the industry, physical AGI, world models, and more. 1:00 Intro 1:47 Does AI's pace feel crazy? 4:04 From particle physics PHD to robotics 6:03 Speech recognition's deep learning moment 8:43 Whole-body humanoid control 10:47 The trash-bag tying demo 13:07 The robotics data pyramid 14:13 Transferring skills across robots 16:30 How does Motion Transfer work 16:56 Why a separate reasoning model 19:02 Can distillation work in robotics? 20:02 The ASIMOV safety benchmark 22:47 VLAs vs world models 25:10 The missing tactile sensing piece 27:48 How far to physical AGI? 30:07 Fundamental breakthrough needed? 31:53 World models for evaluation 33:45 Will the model layer commoditize? 35:11 Humanoid as a form factor

My conversation with Kanishka Rao, who went from working on particle physics at CERN to driving robotics at Google DeepMind. We get into the Gemini Robotics 2 release, trending training recipes in the industry, physical AGI, world models, and more. 1:00 Intro 1:47 Does AI's pace feel crazy? 4:04 From particle physics PHD to robotics 6:03 Speech recognition's deep learning moment 8:43 Whole-body humanoid control 10:47 The trash-bag tying demo 13:07 The robotics data pyramid 14:13 Transferring skills across robots 16:30 How does Motion Transfer work 16:56 Why a separate reasoning model 19:02 Can distillation work in robotics? 20:02 The ASIMOV safety benchmark 22:47 VLAs vs world models 25:10 The missing tactile sensing piece 27:48 How far to physical AGI? 30:07 Fundamental breakthrough needed? 31:53 World models for evaluation 33:45 Will the model layer commoditize? 35:11 Humanoid as a form factor

The Humanoid Hub

11,426 görüntüleme • 20 gün önce