You should read my essay about why your startup might not want to “move fast and break things.”

You should read my essay about why your startup might not want to “move fast and break things.” 

https://samhcarter.com/posts/do-things-that-dont-break/

VC analyst investing in data center hardware, spatial computing & cool stuff 

I write — http://pursuit.beehiiv.com

cs & english @harvard | 🇺🇸🇰🇷

Working on Farcaster and Warpcast. Longer thoughts at https://dwr.email

I agree that speed is imperative. But very little that OpenAI releases breaks. They’ve been even more conservative with testing/post-training since the day it crashed, which is when Claude found some power users IMO

Any startup that’s worth < $10 BN can’t take market share if they break. The standard is too high

FWIW this is mostly about hardware startups and how software investors undervalue their early signs of traction.

I disagree. OpenAI is valuable because they are ahead and move quickly. If they move slower, they will lose their advantage.

Researcher at MIT Media Lab | Creating futures | https://cyrus.website/info

Answers in the essay :) 

It describes hardware startups that qualify product with enterprise customers over 2-3 years before sales. This is not new, except to VCs.

The bar has gone up for software, too, which is also touched upon. Can’t scale to $50M ARR in 12 months if product breaks. Bar is higher than 2000s.

So we have to move fast and build  unbreakable things now? Sounds fun!

How do you think that approach will impact innovation?

Thoughts on paragraph.xyz/@antidote

Agents and Intents @NEAR | prev investments and BD 
@ZkSync, New Enterprise Associates and Balderton Capital

I don't agree with this really. This seems like a gross generalization for a phenomenon that mostly occurs in AI. Trying to "break" the AI is more about a test how human this AI can appear and what its limits are. 

I do not really see this phenomenon with other technologies.

Certainly applies to smart contract development