The State of Startups in 2026

影片總結

YC says hard-tech startups rose to 20% of its batch as AI helps founders build faster and solo founders approach one-fifth.

Y Combinator’s latest batches reveal a shift from software toward companies building in the physical world: hard tech rose from 8% to 20%. Robotics reached 6–7% of the batch, manufacturing 10%, defense 5%, and compute and power infrastructure also grew sharply. AI coding tools are lowering the software burden for technical startups, while demand for data centers, chips, robotics and defense technology creates new opportunities. At the same time, software is changing rather than dying: YC says end-to-end AI products grew from 10% to over 25% of accepted companies, and median monthly revenue by batch’s end rose from $8,000 to $20,000. Some startups now go from zero to seven-figure revenue in three months. More than a dozen YC-backed data and RL-environment companies each earn over $10 million annually. Nearly one in five founders is now solo, and experienced founders are gaining ground as AI lets them turn judgment and product ideas into working businesses faster.

章節

1. 0:00Intro: Experienced Founders Return as AI Models Suddenly Improve
2. 1:06Startups Are Moving From Bits to Atoms: Hard Tech Reaches 20% of YC’s Batch
3. 3:29Why AI Is Making Hard Tech Easier: One in Six Summer-Batch Founders Has a PhD
4. 5:21Defense, Manufacturing, and Hard Tech’s Return: Exosat, Icarus, Nine Mothers, and Knox Metals
5. 9:48AI Compute Becomes a Physical Infrastructure Problem: Rising A100 Costs and Dipole Labs’ Optical Switch
6. 12:34Robotics Nears Its ChatGPT Moment: Astra’s 60–70% Benchmark Result and Renewed Hard-Tech Interest
7. 15:40Software Isn’t Dead; It’s Becoming the Harness: Slack AI and Astra’s Reported 90%+ Score
8. 17:59Why AI Startups Are Growing Faster: End-to-End Agents Lift Revenue and Some Reach Seven Figures in Three Months
9. 22:52The Hidden Boom in Data and RL Environments: Dozens of Millions in Revenue and Reported Billion-Dollar Lab Spend
10. 27:08Why Robotics Will Need Specialized Models: Boost Robotic, Physical Intelligence, and Task-Specific Training
11. 29:38The Rise of the Solo Founder: YC’s Share Climbs from 5% to 18–19%
12. 32:54Why Experienced Founders Are Back: Peter Steinberger and the Value of Judgment
13. 34:59What Founders Should Do Right Now: Prompt GPT-6 to Revisit Unsolved Bugs

完整分析還有什麼

13 章 · 62 個重點 · 13 段段落總結 · 心智圖

其餘 59 個重點、段落總結與心智圖在完整分析裡。 分析這支影片,看完整內容 →

這支影片 37 分鐘,免費方案可分析 30 分鐘以內的影片;Plus 方案起可分析 2 小時以內。