It's 6B down the drain. Saying grok 1.5 is competitive is a joke, if it was any good it would be ranked well in chatbot arena (https://chat.lmsys.org/). Elon is a master in hyping underperforming things and this is no exception.
No, there is no ranking for Grok. It’s not participating.
It would be hard to judge rate of improvement at this point, since the company has only been around for 1.25 years, and grok 1.5 is yet to be released for general access.
You really think investors like sequoia and a16z are dumb enough to fall for Elon hyping things up? They know who he is and They’ve seen him operate at levels basically no other entrepreneur can snd are betting on that
> You really think investors like sequoia and a16z are dumb enough to fall for Elon hyping things up?
a16z invested $350 in Adam Neumann's real estate venture - after WeWork. VCs will absolutely knowingly invest on hype if they think it's going to last long enough for them to cash out with great returns.
Well, the most common ML problems can be expressed as optimization over smooth functions (or reformulated that way manually). We might have to convince the ML world that branches do matter :) On the other hand, there are gradient-free approaches that solve problems with jumps in other ways, like many reinforcement learning algorithms, or metaheuristics such as genetic algorithms in simulation-based optimization. The jury's still out on "killer apps" where gradient descent can outperform these approaches reliably, but we're hoping to add to that body of knowledge...
> Why do you think similar approaches never landed on jax?
Isn't this just adding noise to some branching conditions? What would take for a framework like Jax to "support" it, it seems like all you have to do is change
> if (x>0)
to
> if (x+n > 0)
where n is a sampled Gaussian.
Not sure this warrants any kind of changes in a framework if it's truly that trivial.
Semantically it seems truly that trivial, but in practice handling expectations in AD requires some additional machinery not found in implementations that were not written for nondeterminism.
How is it serious if money is the motor of freedom of speech? The suing culture in the US ensures freedom of speech up until you bother someone with money.
Yes, you can learn how to sue. You can learn how to be a doctor too. You can also learn rocket science. The third one is the easiest to me, personally.
Sounds like the standard disrupt formula should apply. Cant we stuff the court into an app? I kinda dislike the idea of getting a different sentence for anything related to appearance or presentation.
Quite frankly, I see a lot of text in this post and no numbers.
For something to be production-ready I'd expect you to at least cover major things like "latency to serve x in Elixir instead of lang y is k% better" or "EMFU we got when training x in Elixir was comparable to lang y".
These are two random metrics that are of course biased to my experience but the article just feels empty without numbers.
Absolutely would not work in the US (as, unfortunately, most public services). There's already a huge lobby from TurboTax and other players. All countries that got to this level of automation already had publicly developed, free software for tax payers previously.
How did you sleep last night? Your cynicism is slowing.
Just this year, the IRS started a direct file pilot program that flies in the face of "it absolutely would not work". It's not everything, but it's a start.
Not everything has to be clear and to the point. The sentence's murkiness adds emotion to the text. I'm not a native speaker (just like the writer isn't) and maybe that's why we're not so bothered.
But this comment is exactly what I expected from HN.
The intro query "youtube downloader" already showed me relevant results (some website where you paste an URL and bam download). I think there's a big tech bias in the whole post (how relevant is a mastodon poll, for real).
Not saying the current landscape doesn't suck with ads everywhere and incentives to not give exactly relevant results at times, but I think google is pretty good still.
It's 6B down the drain. Saying grok 1.5 is competitive is a joke, if it was any good it would be ranked well in chatbot arena (https://chat.lmsys.org/). Elon is a master in hyping underperforming things and this is no exception.
reply