The major challenge is not keep it building and compatible. The major challenge is to keep up with security issues.

I wish op all the best, it's a tough journey

++

I don't know if they will never go public,but their business is clearly not profitable as it is today.

I'm a heavy user but the moment they stop subsidizing I will pick something else because I know how damn expensive it actually is (my usage is ~4k usd/month, but paying 200 in their max plan).

I guess lot of people will do the same.

The funny thing is the govt would prob prefer all the resources were concentrated (e.g. across MSFT and Alphabet) rather than diversified across many firms in a head to head fight vs China.

So therefore theres no argument to keep anthropic and OAI afloat if they cannot do so on their own.

Interesting how that whole conversation about "AGI tomorrow" completely disappeared and it's all about making money now.

I'm a fan of claude and I have been using their top tier plans because it's so damn cheap comparing to my real usage. They will have to stop subsidizing this at some point, and people (include me) will flee. What they are gonna do?

Interesting times we live in.

Sam Altman being like "AGI in <6 months/8 months/2023/2024/2025/2026>" is just sad to listen to now. They don't have it, and never will.

"confubabulating.... You're definetely right"

Shame it isn't Luis Pareto.

A historical reminder: the Nazis were highly effective at persecuting Jews in the Netherlands precisely because of the country's meticulous population registry.

You never know who will be in power tomorrow. The convenience of a centralized system today does not outweigh the catastrophic danger of that system in the wrong hands in the future. Specially nowadays.

Interesting how op describes his own experience and then assume that every other company around the globe experience exactly the same.

He generalizes CEO's behavior but provides no evidence. Cool.

What do you expect from a person other than the generalizations they see from the experiences they have had...? What 'evidence' could possibly be given other than extensive anecdotes?

The article presents hypothesis as fact with insufficient science. I have no problem discussing speculation, but if the author wants to promote their claims to any more than that I would want to see more journalism.

What journalism would you like to see in a blog post about the authors opinion?

Not everything runs on paid models. Claude and Codex are frontier models, but some people have much higher usage needs and finite budgets that force them to self-host. And if you're self-hosting, you're very likely running a Chinese model

Especially if you're pre-funding, which the sampled startups (those pitching A16Z) were.