First Slice 002 is out! For Episode 002, I chatted with Dan Lambert (@peppybots), founder of Peppy. Dan is building “Lovable for robotics.” The idea: tell a robot what you want it to do in plain English, test it in simulation, then eventually deploy that behavior to a physical robot. Dan thinks robotics is hitting an inflection point right now. Robot brains are getting dramatically more powerful. Useful hardware is getting cheaper. He has robots on his desk that cost under $5,000. And AI coding tools are starting to make robotics development accessible to people who aren't robotics engineers. Dan's first experience setting up a robot with existing tooling took him more than 10 hours. Peppy wants to turn that into roughly 10 minutes. But vibe coding a robot is very different from vibe coding a website. If AI hallucinates while building software, the application might break. If AI hallucinates while controlling a physical machine, the robot could break something or hurt someone. We get into that problem, how Peppy is building a safety layer between AI and physical hardware, the shortage of real-world robotics data, NVIDIA's role in the industry, and why robots eventually need to get “punched in the face” by the real world before they become truly useful. Dan also explains why Peppy could eventually look less like a robotics company and more like Android: a common software, developer and application layer that persists even as the robot hardware underneath it changes. This is exactly what I want @FirstSlicePod to be. What's next, before it's obvious. TIMESTAMPS 00:00 Why robots, why now? 01:27 Turning 10 hours of robotics setup into 10 minutes 04:18 What's actually preventing general-purpose robots? 06:09 Peppy's oversubscribed ~$1M pre-seed round 09:41 Why useful robots are still a few years away 11:07 Why robots need to get “punched in the face” 11:57 From a plain-English prompt to physical motor movement 15:32 What would robotics look like without NVIDIA? 17:59 How experimenting with robots turned into Peppy 20:41 “Lovable for robotics” and the vibe coding moment 23:18 Peppy's December 2026 public launch 26:38 Should robots be allowed to use knives? 29:49 Do robots actually need to be humanoid? 31:15 The moment Dan became obsessed with robotics 32:49 Talking to a robot and seeing a glimpse of the future

FIRST SLICE 002 DAN LAMBERT (@peppybots), PEPPY FULL EPISODE TOMORROW (September 3, 2026)

FIRST SLICE 001 MATT LENHARD (@MattLenhard), VECTORAL (@vectoral_labs) FULL EPISODE TOMORROW (August 26, 2026)

I had no idea this market existed until I started researching @vectoral_labs for the first episode of @FirstSlicePod with @MattLenhard People are buying access to OpenAI, Anthropic, DeepSeek and other models for as much as 98% below list price. And there is an actual supply chain behind it.

AI has created a new kind of software attack. Matt Lenhard (@MattLenhard) calls it “denial of wallet.” An attacker doesn't necessarily have to take your application offline. If they can repeatedly hit an expensive AI endpoint, they can make your company pay for the attack. Lenhard says he has seen larger organizations get hit for seven figures before they notice. From First Slice 001 with Matt Lenhard, founder of Vectoral (@vectoral_labs) Watch the full conversation: nitter.net/jaredstaylor/status/20…

First Slice 001 is now out! I started this show to find people and ideas worth knowing early, then go deep before everyone else is talking about them. For the first episode, I sat down with Matt Lenhard (@MattLenhard), founder of Vectoral (@vectoral_labs). Matt discovered an underground market selling access to AI models at massive discounts. He saw companies getting hit with more than $1M in token fraud per day. The worst case he saw was $10M in 48 hours. A few days later, he quit his job. We get into how this market actually works, the brokers selling AI inference, denial-of-wallet attacks, how Vectoral is fighting it, and why Matt thinks hijacked AI agents could be one of the next major attack vectors. This is exactly what I want First Slice (@FirstSlicePod) to be. What's next, before it's obvious. TIMESTAMPS 00:00 The underground market for stolen AI tokens 01:33 $10M in 48 hours and why Matthew quit 02:38 How the stolen AI token market works 08:07 Who buys discounted AI inference? 10:01 How Matthew chooses what to build 13:11 Going inside the token broker market 17:22 Model distillation and frontier AI 19:33 Denial of wallet attacks 23:53 How Vectoral detects AI abuse 26:06 Where AI fraud goes next 29:40 AI regulation, jobs and moving fast 32:25 What's next for Vectoral

Episode 1 drops tomorrow morning!