@neeuumm@koltregaskes@FT Thank you so much, your gift link worked for me. I can't believe they spent the article talking about models like Haiku 4.5. That model series hasn't had an update since last October, which is forever in AI terms...
@techikansh@ChrisGPT 'Tesla’s Roadster is supposed to debut with a set of SpaceX-developed cold gas thrusters that let it lift off the ground, and the company has been dangling a “flying” demo since 2018.' (source: https://t.co/pRS4Zq7Rrm )
@5StringsInTX 2. Look for the missing window in History
Open ⋮ → History and look under ‘Recently closed’ for an entry showing a number of tabs, perhaps ‘60 tabs’. Open that entry and choose ‘Restore window’
Hope these help! I had this problem once before
@5StringsInTX 1. Try the restore shortcut
In Chrome, press Ctrl + Shift + T on Windows/Linux, or Cmd + Shift + T on a Mac. This can bring back an entire closed window, not just one tab. Try it several times if it initially restores something more recent
OpenAI and Anthropic staff can't work well together, but their models *can*...and should.
If you subscribe to ChatGPT and Claude, one of the best ways I've found to maximize value is to use frontier models together.
That's why I built:
> 'Astrable'
A free, open-source Codex plugin that empowers you to use Astra and Fable in partnership.
Details below 👇
@petergostev@BBCNews The BBC article looks like it's aiming to make the explanations accessible to a general audience, but in the process it equates tokens with lines of code…'130 billion output tokens, or the individual lines of text and code an AI model produces in answers' https://t.co/k2k45gGv60
Astra is fully rolled out to Plus, Pro, Business, and Enterprise users in Codex and ChatGPT Work.
Go build!
And if you need inspiration, watch Astra in action, live: https://t.co/VmBLksCNZW
Tristan Buckmaster says OpenAI told him an internal model had produced a roughly 100 page proof of finite time blow up for forced Navier–Stokes.
If that proof is correct this would be world changing math news.
But please keep in mind Buckmaster says he has not personally seen or verified OpenAI’s proof.
Now the drama part is Buckmaster alleges that Sébastien Bubeck proposed publication arrangements that would remove Levent Alpöge from authorship partly because Alpöge works at Anthropic, and that after Buckmaster threatened to disclose what happened, he was asked, “Why would you ruin your career?” Buckmaster also says OpenAI’s effort began only after information about his research had reached the company.
I don’t think this is going to end pretty.