Hacker News Reader

Comment by nileshtrivedi | Hacker News Reader

> we're not talking of Google or Facebook investing a chunk of their billions cash

OpenAI had raised $1B from Microsoft in 2019 and used it to train a 175B param model. Now, they have raised $10B and are training GPT-4 with 1.5T params. GPUs are capital intensive and as long as there are returns to bigger models, that's exactly where things will go.

View on HN

I can't find any source on the 1.5T params number. I'd love to read more if you have any links to share. Thanks

afaik, gpt-4 is mostly rumours so far, same thing for the 1.5T number. gpt-4 is suerly coming.

Maybe it will be called GPT-XP by then, with Microsoft owning half of it.

Or GPT-365

Then they can bring back the talking paperclip, but this time actually useful.

It could actually work. It would be an incredibly gutsy move and I love it, and they'd probably earn a lot of respect. They’d get so much press for it. And if it held up, it’d probably be one of the things that MS is remembered for.

Or GPT One.

Why not ask GPT itself what it wants to be called?

Looking forward to see GPT-4 recommending Linux and Libre Office instead of Windows/Office as the logical choice out of 250 IQ ML Model...

GPT-10 will be evergreen and 'the last version of GPT'.

And then three years later GPT-11 will be required to run the latest games.

In my imagination, OpenAI does what Bungie did when MS bought them, and open-sources what used to be their crown jewels.

That said, GPT-AlephOne only makes sense if there's a preceding GPT-∞.

They have got to release GPT-3.11 For Workgroups first.

Will 1.5T parameters be possible to run in the public way GPT-3 is? I can’t wait to see what happens with this much learning!

Reply on news.ycombinator.com