Hacker News new | ask | show | jobs
by OutOfHere 625 days ago
> The chat on the website is not their main product

Guess what: enterprises are made of people. People like to try things out. If people are not happy with something for their personal use, they most definitely are never going to recommend it to their employer. This is why OpenAI wins. It is in fact one of the factors that sets apart a hyper-successful product from a wannabe.

1 comments

Literally every single person I know who’s building anything in AI (dozens of early stage founders) prefers Claude over GPT at this point.

The AI emperor will not be the one who has the most consumers logging into product.com to use the chatbot.

Compound this with OpenAI’s continuous shedding of, as far as I can tell, every credible researcher… I find your position quite hard to believe, even accounting for the hysterical tone.

> Literally every single person I know who’s building anything in AI (dozens of early stage founders) prefers Claude over GPT

I don't believe this at all. I am not here to argue that Claude is a worse model, only that Anthropic is a worse company.

> The AI emperor will not be the one who has the most consumers logging into product.com

Your point only goes to show how much Anthropic hates its end users.

> OpenAI’s continuous shedding of, as far as I can tell, every credible researcher

OpenAI has zero trouble hiring great talent. As I see it, they lost a lot of dead weight that had no interest in bringing AI to the masses, but had an agenda of their own instead.

“Your point only goes…”

Huh? How so? Sorry not even clear what your complaint is… is it the Firefox (3% market share) login bug? The Claude chat experience has been superior for a while now, and Projects and Artifacts make it 100x so.

Good at hiring and bad at retaining is much worse than the reverse, especially for long-lived R&D projects.

> what your complaint is

It helps to read. It's noted in the original comment. It has nothing whatsoever to do with Firefox, as it manfiests only on the Anthropic website.

Ah that's right, the "sickening" pattern of emailing a login code. Not a fan of it myself, but boy is this dramatic.
It's that, but it's more. They promise API credits to try out the API, but then refuse to give any credits. Overall, it's as if they don't care even a little bit for individual customers.
Shredding you say?