Tag
Discovered Materials, a YC P26 startup, launches AI agents for discovering new materials, showcasing a benchmark where frontier LLMs find stable new materials but struggle to propose plausible synthesis recipes.
A Y Combinator method for compressing a month of go-to-market research into 3 hours using Claude, with a specific prompt workflow for market research and validation.
Advaith Sridhar introduces Discovered Materials, a startup building AI scientists to discover new semiconductor materials, releasing hundreds of discoveries and a benchmark.
Y Combinator CEO Garry Tan shares insights from his Startup School 2026 talk on building personal AGI, emphasizing that context and skill files matter more than model choice, and that capturing reusable skills can multiply output dramatically.
Flexport CEO Ryan Petersen and Y Combinator CEO Garry Tan discuss how startup founders' initial visions often evolve and expand as they learn from customers, arguing that too much certainty about the vision is probably wrong.
Herdr, now a YC-backed company, launched a Vercel Sandbox plugin that lets developers run multiple coding agents in isolated environments from a local Herdr pane.
Y Combinator shares a Startup School talk in which Science Corp CEO Max Hodak discusses the company's retinal implant that has enabled a blind patient to read a 300-page novel, along with lessons on startup infrastructure, hiring, and judgment.
Maingen launches on YC with simulations of real industrial operations, allowing AI agents to control the physical world.
Herdr, an open-source runtime and TUI for terminal coding agents, is joining Y Combinator's F26 batch. The project, which has 25k stars and 340k downloads, will remain Apache-2.0 licensed.
YC's Garry Tan talks at Startup School 2026 about the rise of personal AGI, explaining why AI agents running on your own infrastructure will enable smaller startup teams and why founders should own their intelligence rather than rent it.
ProvenMetal, a YC S26 startup, launches a US-based PCB manufacturing service that delivers circuit boards in seven business days as standard, with no expedite fees, targeting drone and defense teams.
A space startup built its first satellite for $2M including launch, versus traditional quotes of $75-100M, illustrating the power of iterative execution over waiting for certainty. The story comes from Y Combinator's Lightcone Podcast featuring Starcloud's Philip Johnston.
A short post highlighting a Y Combinator podcast with Bryant Chou about his startup Ploy, which uses AI to revive and redesign old-school websites, described as a nostalgic and impressive AI application.
The cost of starting a software company has plummeted from $5M in 2000 to about $500 today, driven by open source, cloud computing, and AI. The piece argues that traditional startup advice like 'validate before building' may now be outdated, as building and shipping is cheaper and faster than conducting interviews.
Ultrasonium, a YC S26 startup, announces a new method for manufacturing complex metal parts that is 15x faster and 75% cheaper than current techniques.
Meta's chief AI officer Alexandr Wang claims that with the right agentic loop and evaluation system, a swarm of agents can outperform a team of 100 engineers, speaking at Y Combinator Startup School.
YC shares a podcast episode with Starcloud CEO Philip Johnston, whose company launched an Nvidia H100 GPU into orbit and trained the first LLM in space, raising $200M and planning 88,000 satellites.
A YC-backed startup pitches 'Air Defense as a Service' — autonomous interceptor boxes to protect data centers from drone attacks, referencing an incident where drones took AWS data centers offline.
Y Combinator open-sourced QM, the multiplayer AI agent harness they use to run their company. It provides per-team workspaces, per-project memory and scheduled jobs, plus a keychain that shares credentials without exposing raw API keys.
EdotEnv is a Y Combinator-backed platform that provides market-derived reinforcement learning environments to teach LLMs and agents applied machine learning and long-horizon planning under adversarial noise.