jsnell

17 ms·

301.
by jsnell 1y ago
If a protocol is getting implemented wrong repeatedly and independently, what is the value in it being nominally "nice and simple"? You might not be understanding just how complex the H1.1 feature set really is, and how how much f
302.
by jsnell 1y ago
This is the fifth time I've seen this site be linked this week, and I feel I must be missing something. There's nothing there yet, folks. It's still just an announcement of an announcement! No longer does a vulnerability need
303.
by jsnell 1y ago
Dupe: https://news.ycombinator.com/item?id=44622637
304.
by jsnell 1y ago
See also the discussion on the technical description last week: https://news.ycombinator.com/item?id=44591605 (IMO much more interesting article than this announcement, and that probably should have gotten more attention th
305.
by jsnell 1y ago
All written by hand, and you can check the style against thousands of posts I've made on HN over almost two decades. (I explained the issue briefly in the first paragraph, but the article is 15k words. Hard to convincingly rebut that w
306.
by jsnell 1y ago
The analysis is just bogus. He is basically comparing two years of inflated AI capex estimates to a low-ball estimate of one year of trailing revenue. Let's unpack that a bit. Capex is spending on capital goods, with the spending being
307.
by jsnell 1y ago
From a quick skim, at least 90% of the article is about profitability. The remaining 10% is mostly bragging.
308.
by jsnell 1y ago
Ah, a new entry in the "throw every possible AI criticism at the wall, and see what sticks" genre. As always, going for quantity rather than quality undermines the entire endeavor, as the article gets dominated by the stale and mo
309.
by jsnell 1y ago
> The only way I found is to block the agent running certain commands (requiring my consent) but that can only take me so far, since there are infinite command line tools the agent can run. You're doing this the wrong way around. Yo
310.
by jsnell 1y ago
Not really? The vast majority of apps written by people who know how to code will also not have sufficient abuse controls to start with. Actually most will have no abuse protections, and certainly nothing that would be effective against j
311.
by jsnell 1y ago
The thing is, people claimed already a year or two ago that we'd reached peak data and progress would stall since there was no more high-quality human-written text available. Turns out they were wrong, and if anything progress accelera
312.
by jsnell 1y ago
As far as I can tell, OpenAI didn't participate, and isn't claiming they participated. Note the fairly precise phrasing of "gold medal-level performance": they claim to have shown performance sufficient for a gold, not t
313.
by jsnell 1y ago
I feel you're misunderstanding something. That's not "this exact experiment". Matharena is testing publicly available models against the IMO problem set. OpenAI was announcing the results of a new, unpublished model, on
314.
by jsnell 1y ago
> It is very likely OpenAI solved the IMO problems in a matter of minutes, maybe even seconds Really? My expectation would have been the opposite, that time was a constraint for the AIs. OpenAI's highest end public reasoning models
315.
by jsnell 1y ago
That seems like a non sequitur. He is making two arguments. One is that AI capex is starving other industries. And the other that AI capex is causing major GDP growth, attributed to both the direct investments themselves, as well as the mul
316.
by jsnell 1y ago
For a first approximation, yes? Their earnings have been growing steadily for years ($4-5B growth in each of the last 8 quarters), with no seasonal effects. Q1*4 is highly likely to be a better estimate of their eventual 2025 calendar reven
317.
by jsnell 1y ago
I don't see how you can make the argument that a large portion of funds used for AI capex were diverted from other investments (and starving other industries), while simultaneously applying the economic multiplier to the whole sum when
318.
by jsnell 1y ago
> Can OpenAI stop spending on research to get their costs down? They do not need to. Their costs are already really low given the size and nature of their user base. > “Can OpenAI do this simple thing” is the whole question! There was
319.
by jsnell 1y ago
Dupe: https://news.ycombinator.com/item?id=44522772 (750 points, 478 comments) https://news.ycombinator.com/item?id=44526912 (276 points, 272 comments)
320.
by jsnell 1y ago
But they're not asking for your consent for that. They're saying that if you use these features, they have a legal basis (legitimate interest) other than consent to use your data for product improvements. The only thing they'
321.
by jsnell 1y ago
I'm pretty sure what's happened is that you'd agreed to it and forgotten. These features were very explicitly turned off in the EU half a decade ago, and all European users had to make an explicit decision of whether to opt i
322.
by jsnell 1y ago
That's not my experience from the short samples in the article, I could get the correct meaning out of all of them at a glance. The only slowdown was in returning to the text after the initial read, to try to puzzle out the exact defin
323.
by jsnell 1y ago
You're in the EU, right? In that case these features have been off by default for years (I posted a link to the 2021 documentation showing that, but I'm 95% sure it was already the case pre-Covid). If you've been using these
324.
by jsnell 1y ago
Workspace has published policies about how the customer's data may be used: https://support.google.com/a/answer/15706919 > Your content is not used for any other customers. Your content is not human review
325.
by jsnell 1y ago
Your data is not being used to train an LLM even if you click "yes" to this.
326.
by jsnell 1y ago
> Either allow us to process and use your data however we want, or disable these existing features we've offered for many years That's not an accurate description. It's not "however they want", it's pretty c
327.
by jsnell 1y ago
I mean, the GP made a point about "per-user costs" that I believe was false, so that was the specific thing I was commenting on. Steering the discussion to a totally different topic of research costs doesn't help us reach clo
328.
by jsnell 1y ago
> Do you know why it's so expensive? I'd thought serving html would be cheaper, particularly at Facebook's scale. I don't know about Facebook specifically, but in general people underestimate the amount of stuff that
329.
by jsnell 1y ago
Indeed, OS X goes down, and obviously none of us actually believe that. But not only does Statcounter report that clearly faulty number, but they have yet to fix the problem. This happens all the time. When their numbers are clearly wrong,
330.
by jsnell 1y ago
The statscounter data is not reliable, and it is just embarrassing how often these posts make it to the HN frontpage. You even have a demonstration in this very article, with the surge of classic Mac OS to 7% for several months . The data

More ›