LS Latent Space

Podcast

Latent Space

Everything from Latent Space: articles, AI News briefings and podcast episodes.

Website ↗ Last published 31 October 2026 565 posts collected by Sputnik

Said on Latent Space

“It is very different to be a services firm that has access to AI tools versus a firm that was built to be AI native. The firm with AI tools requires their users to prompt the agents to produce work or to do work. But here, at ARCA, agents prompt people more than people prompt agents. And it is a proactive kind of service that is running for them 24-7.” — Aman Singh, Latent Space
Aman Singh · 31 Oct 2026 Listen →
“Building software is debatably a solved problem. Maintaining, like keeping things running, is a whole different beast.” — Paul Copplestone, Latent Space
Paul Copplestone · 28 Oct 2026 Listen →
“We don't have a single Rust developer in the room. … The two things that mattered most were the plan to do the thing and then the verification of the thing, because the actual implementation was Claude writing code. And so what we ended up doing was we literally built an entire fork of the entire API in Rust and then started running shadow traffic through it and verifying the outputs.” — Angela Jiang, Latent Space
Angela Jiang · 27 Oct 2026 Listen →
“Because of agent training data, it's very hard to deprecate things now. We see that agents often love using old versions of Modal syntax. And it's kind of hard to nudge them away from that. So we've actually had to be a little bit more conservative ... with how we deprecate and remove features.” — Erik Bernhardsson, Latent Space
Erik Bernhardsson · 22 Oct 2026 Listen →
“My view is code reviews in large part is people in many cases trying to just escape the accountability of merging things into production. And if you have that culture, then to me, that's a sign that you need to invest way more in CD and CI.” — Erik Bernhardsson, Latent Space
Erik Bernhardsson · 22 Oct 2026 Listen →
“Right now, if you took a B300 and you tried to run inference on it, you can make $40 an hour or so selling any model, basically. And so that's if you use it at 100%. And maybe if you only have like 50% utilization, so like $25 an hour. Meanwhile, you can buy B300s at like $5 to $6 an hour. One of these numbers has to change. It's a huge arbitrage.” — Evan Conrad, Latent Space
Evan Conrad · 20 Oct 2026 Listen →
“We really think that it's learning a lot of sociology. It's not learning ground truth. It's just learning the choices and the fingerprints of the people who are trying to develop these molecules. … But when you train models on the exhaust, you might be in trouble.” — Ian Quigley, Latent Space
Ian Quigley · 14 Oct 2026 Listen →
“Today, there's a lot of robotics companies that's talking about deployments, talking about pilots. But what's really hard is not putting your robot in one location or doing a pilot. What's really hard is that the customers believe or validate that your robot creates enough value that they want to deploy your robots across many of their sites.” — Jason Ma, Latent Space
Jason Ma · 13 Oct 2026 Listen →
“For coding agents, language model, it's okay if it has latency, right? … But for robots, the inference latency has to be very short. The robot needs to act. You can't just hold a knife and not do anything for three seconds. That's going to freak people out.” — Jason Ma, Latent Space
Jason Ma · 13 Oct 2026 Listen →
“I say proteins are the building blocks all the time, but actually, I don't believe in it. Proteins are extremely complex. They are disordered. Their shape might change depending on the context. … What we are trying to do is basically someone got a structure, deposited it in the PDB, and we are trying to replicate and get the same structure. That's what we did. And it just so happens that it's useful. But that does not mean that we have understood all of protein dynamics.” — Pushmeet Kohli, Latent Space
Pushmeet Kohli · 10 Oct 2026 Listen →
“Machine learning is really good at what's been trained on. But scientific discovery is almost by definition what you haven't been trained on.” — Ekin Dogus Cubuk, Latent Space
Ekin Dogus Cubuk · 8 Oct 2026 Listen →
“I don't think voice is necessarily low [effort]. It's more like how much information is in the prompt. You can um and uh and add some sentences and be like, oh, actually I changed my mind, in the middle of the prompt, and it will be able to follow that perfectly. … For a lot of people, it's just way easier to talk than to type. And if that gets more information out of you, that's better.” — Thariq Shihipar, Latent Space
Thariq Shihipar · 29 Sep 2026 Listen →

Hosts

Latent Space is hosted by Alessio Fanelli , swyx (Shawn) , Brandon Anderson , RJ Honicky , Shawn swyx Wang , Shawn Wang , Erik Torenberg , Latent Space , AI Charlie and Brian McCullough .

Recent Episodes

  1. 31 Oct 2026

    Agents Prompt People: Using Agents for Long-Horizon Financial Work

    With Aman Singh

  2. 31 Oct 2026

    [AINews] Google admits AI Mode hurts Search revenue; -15% $500B drop worst since 2004 IPO. AI Bear Market coming?

  3. 30 Oct 2026

    [AINews] Building AIE CODE: The top Coding Agent Conference

  4. 29 Oct 2026

    Can Frontier Models See Past a Decoy? Elorian Tests Visual Reasoning

    With Andrew Dai

  5. 29 Oct 2026

    [AINews] Gemini 4 Carbon tops leaderboards, launches for cyber defenders

  6. 28 Oct 2026

    Why Agents Love Supabase

    With Paul Copplestone and Ant Wilson

  7. 28 Oct 2026

    [AINews] Anthropic ends Coding Interviews: "Hunger beats tenure"

  8. 27 Oct 2026

    How Anthropic ships Claude with Claude: Separating the Brain from the Hands

    With Katelyn Lesse and Angela Jiang

  9. 27 Oct 2026

    How Anthropic ships the #1 LLM in the World safely: Separating the Brain from the Hands

    With Katelyn Lesse and Angela Jiang

  10. 27 Oct 2026

    Coding From the Himalayas With an Offline AI Agent

  11. 27 Oct 2026

    [AINews] Anthropic files for $2T IPO in 2 weeks: 1400% growth rate, 48% of revenue through cloud partners, 85% margin, $1B op profit in Q2, $134B cash in bank and $10B lifetime burn

  12. 24 Oct 2026

    [AINews] Kimi 3.1 Pro 104BA-2.8T: Fable-5.1-Class Open Weights narrows Closed-AI lead to 1 month... and Pro Highspeed API with 2-3× speed

Guests

Simon Willison (8 episodes), Alex Volkov (6 episodes), Michelle Pokrass (4 episodes), Raza Habib (3 episodes), Roland Gavrilescu (3 episodes), Thariq Shihipar (3 episodes), Aman Sanger (3 episodes), Shreya Rajpal (3 episodes), Pauline Brunet (3 episodes), Eugene Cheah (3 episodes), Graham Neubig (3 episodes), Jeremy Howard (3 episodes), and 311 more.