jeanlucas

well, do it as soon as it opens up;

if you really believe it you found a free money glitch

The data to be open, in my case.

The "independent" METR that is composed by... Checks notes... Previously employees from the top labs.

And any exponential growth is also a multiplication, it's not a big deal, just giving example on what they asked before.

Somehow people are interpreting this as some kind of math trick.

It's not a trick, our brains are not wire to deal with numbers growing quickly, but we do understand the math.

Of course it is logical, we are not discussing the math.

But it isn't intuitive, you're not used to numbers that big or that grow that fast.

Few things go from a million to a billion, especially when related to time.

One million seconds is 11 days, a billion seconds is 31 years.

You need to catchup with the last 15 years then. It's not new at all. But I agree it is getting worse, but it is getting worse every year for a while.

Recent?

Oh dear.

If you really wanna worry about all of these things you better not use any open source from now on. Or even private software as well, because you can't even audit and know what "vibe code" touched.

YES YES YES YES

That's perfect

wrong link...

this guy gets it

So if they paid for pre-training they would not be able to open their weights?

Mostly: Discord, Github issues (as forum), X (twitter at the time), and Discourse! Discourse was growing steadily, even startups like OpenAI and neobanks like nubank (Brazil) had forums for their users because keeping your community active and close was something considered valuable (devrel since then has experienced a rapidly decline).

The sale is relevant. The spike in activity right before the sale already shows the management hand and how they actively ran the org to optimize for this sale. The rebound right after is hard, and it was (right) before chatgpt

They accelerated the downfall with this and then chatgpt came over

Nice! The showcase of companies is really nice