FinnLobsien

And government benefits which kick in no questions asked in old age

I think it's similar to those note taking apps that were big a few years ago (Roam Research, Obsidian etc.) Being weird enough to not use the free incumbent and pay for a niche solution instead m…

They are for profit, they're just a "public benefit corporation". I don't know exactly what that means and I don't think it comes with any obligations (like being a non-profit…

Yeah that's exactly what I thought: Infra is usually more usage based

The question is how much that first-mover advantage still means. I guess with video games and movies (Doom and Ghibli), there's a massive difference between telling ChatGPT to make your selfie lo…

Yeah exactly. I feel like it's the same on every marketplace: Their incentive is to have good ratings on the platform. So even if it's not down to taste, but to actual objective factors the …

Sorry, I meant more like the best designers wouldn't sign up to review some random website for $20. So it doesn't actually select for great reviewers.

Hasn't this always been true? He makes the the point that curiosity, initiative etc. irrespective of your skills will thrive in the future. Which I think it always has? It's always been the …

oh this is interesting because I feel like it wouldn't work with infra as that runs in the background and you almost never have zero usage on a DB.

I've come to distrust many reviews. First, it's all Goodhart's Law. I heard the average rating on Airbnb is now 4.8. If everyone in the system (guest, host, platform) is incentivized to…

I've used platforms like these for copywriting in the past. I think they all have the same problems at some point: 1. The very best don't use them because it's not worth their time. 2. …

> There were a million Doom clones, none of which were as good as Doom. The same will be true of AI art copycats. True, but this was also during a time when it was incredibly hard to make a video g…

Yeah, that logic doesn't work. I guess the point is that you shouldn't categorically judge its production because of the worst examples, which I guess applies to any meat/fish productio…

Also, in 2005 there wasn't that much of an acquisition ecosystem. I feel like fewer founders back then actually thought about getting acquired and instead shot for building big companies. You had…

Yeah I think they do it every batch or at least every year. https://www.ycombinator.com/rfs

Yeah it's a good question to ask: Was the startup success rate in 2005 overall higher or was it actually due to YC selecting so well. Also, success is relatively defined. A $50m exit might be som…

There are many factors (YC needs to allocate more capital, not founder-led anymore, needs to fill more batches...) etc. But another one is that the first batch has had 20 years to grow and compound wh…