(1/n) just binged this YC podcast and it completely flipped how am thinking about enterprise AI deals
https://t.co/CRRhyGhlMB
everyone's been doom posting about that MIT study saying 95% of AI projects fail but when you actually read it? it's basically a playbook for startups
In a recent batch talk, YC General Partner @t_blom broke down how to build a self-improving, AI-native company.
He walks through how to create recursive, self-improving AI loops, and why founders who get this right will run companies that improve while they sleep.
00:00 — Companies Are Roman Legions
00:54 — Copilots Are the Wrong Mental Model
01:55 — Extract the Domain Knowledge
02:24 — The Recursive Self-Improving Loop
04:12 — The Holy Shit Moment at YC
05:50 — Self-Optimizing Product and Support Loops
06:29 — Burn Tokens, Not Headcount
07:23 — Middle Management Is Over
08:05 — Make Everything Legible to AI
09:40 — Regenerating the YC User Manual
11:19 — Software Is Ephemeral, Context Is Valuable
12:18 — Where Humans Still Matter
Claude pro tip: ask claude to explain something in 'grug mode' or use the https://t.co/OjjFsj2A5S plugin.
It's stupid, but it works. And if it works, it's not stupid.
a company spends a percentage of their revenue on R&D to stay ahead
you need to do the same with your career to stay ahead
10% of your week should be spent on “personal R&D”
learning a new tool
talking to a mentor
building something
documenting the journey
reading docs from another part of the company
etc
it is the only way to stay ahead
and with the pace that things are moving
that is the bare minimum required imo
@snowmaker its high-stakes game
- radiation
- debris tax
- repair mission
- launch weight
- cost
but potential payoff is massive free cooling with 247 sun
Being a founder is hard. Being a solo founder is much harder. Kennan did YC in S20 with Skio. Applied with one idea, pivoted during the batch, then pivoted again. Never gave up. The last pivot worked. Today Skio sold for $105M in cash.
There are very few straight lines to success. Congrats, Kennan!
I've worked with a few companies on this and spent a ton of time thinking about it
The hard part is if you want your employees using Claude, you don't have enough centralized control over their system prompt or skills, so no matter how well-organized your Notion or whatever is, the agent likely won't use it effectively
But everyone wants to use Claude and is very happy to be locked in by them, assuming they'll build whatever you need in a few months if it's not already there (probably correctly?)
Hard to find the right lane for this as a startup
@cailynyongyong exactly ive seen this a lot doing consulting and audits
usually starts with clients.. saying their last few pilots didnt really work
and i give them this analogy all the time building one stop soln on their company its like putting f1 car on a track with potholes