mh-

22 ms·

421.
by mh- 1y ago
And the law in question: https://en.wikipedia.org/wiki/The_Walker_Montgomery_Protecti...
422.
by mh- 1y ago
Can you expand on that? What would ideal look like to you?
423.
by mh- 1y ago
This is just factually incorrect. Yes, business efficiencies have come from all sorts of tactics, but this being HN: the engine on the 737 MAX uses >30% less fuel at cruise than the JT8D from the 1970s-era 737s. Not to mention airframe i
424.
by mh- 1y ago
In no particular order: Air travel (thanks to efficiencies) costs a fraction of what it used to, look up what % of people have traveled by air and compare that to previous decades. Cars are safer than ever, per mile driven, for their occupa
425.
by mh- 1y ago
My assumption is they want to eliminate/prevent schemes where a ton of apps are signed as a service by a small number of centrally controlled keys. Someone elsewhere in the thread said this is how F-Droid works, but I can't confir
426.
by mh- 1y ago
Yeah, I think anything that has Google Play would fit that qualifier. So that's basically all major devices (in the West, at least). Oof.
427.
by mh- 1y ago
Thanks. It's worthwhile context that I perhaps should have disclosed, but I don't think it affects my opinions in this thread. My opinion was simply in reaction to an, IMO, nonsensical claim: > AI is now as good as it's g
428.
by mh- 1y ago
More than you'd think, from what I've seen.
429.
by mh- 1y ago
Sure, slowing is natural for the reason you say. But the statement we're commenting on is: > AI is now as good as it's going to get And that's just silly, from my point of view.
430.
by mh- 1y ago
You can order a hamburger with "no salt" now via the app (and probably the kiosk?). But I'm not sure if there are other seasonings you're referring to, though.
431.
by mh- 1y ago
That's a strange source of confusion. A plain cheeseburger without cheese has a different name: a plain hamburger. I can't imagine saying "cheeseburger", no matter what qualifying words around it, and being surprised i
432.
by mh- 1y ago
Taking out the fact we're talking about "AI" for the moment.. doesn't it seem unusual to speculate that despite recent progress, it's just going to be.. flat from here out? (That's directed at El Reg, not you.)
433.
by mh- 1y ago
I agree. I'm confused that this idea is even controversial. I would absolutely use it in the way you're describing. I've wanted something like this since around when Alexa/Echo launched.
434.
by mh- 1y ago
I believe half of the comments here are just dumping on AI-related ideas because they see it as their duty to counter the hyperbolic claims about capabilities being tossed around. I enjoy reading both sides of the argument when the argument
435.
by mh- 1y ago
Those "refine your results" buttons is clever UX. I like the Choose your own adventure feel to it. Nicely done.
436.
by mh- 1y ago
Supposing I accept that's a likely outcome, it's exactly the same thing that would have happened if a typical human shopper searched for Vitamin D and picked the top result, right? The cynicism on these topics is getting exhaustin
437.
by mh- 1y ago
Yeah, sorry I was talking about the "why does saying no elephants cause elephants". Phrasing it as do rather than don't is probably still more effective on both humans and LLMs. :)
438.
by mh- 1y ago
Agreed. I wasn't necessarily thinking for cost optimization, simply for capacity purposes. Whether because they're using a bunch themselves (like you're saying, via training for example), or otherwise.
439.
by mh- 1y ago
I have no inside info, however, I would be shocked to find out that OpenAI does not have several knobs for load shedding across their consumer products. Had I been responsible for implementing that, the very first thing I'd reach for
440.
by mh- 1y ago
I just replied to OP with an explanation and some links you might enjoy.
441.
by mh- 1y ago
I see where you're coming from, but if you take a look at the system prompts for these models (some are public, some have partially or fully leaked), you'll see that is no longer a concern. At least not for the kind of models bein
442.
by mh- 1y ago
I would totally expect a different result even on the same model. Especially if you're doing this via a chat interface (vs API) where you can't control the temperature parameters. But yes, it'll be more effective on a differe
443.
by mh- 1y ago
That disables GPT's ability to use that Tool altogether. Despite the confusing location, it doesn't have anything to do with whether it gets your Custom Instructions or not.
444.
by mh- 1y ago
It's quite incredible how fast the generative media stuff is moving. The self-hostable models are improving rapidly . How capable and accessible WAN 2.2 (text+image to video; fully local if you have the VRAM) is feels unimaginable fro
445.
by mh- 1y ago
Isn't that just an argument for always doing open book tests, then? Seems like there's no downside, and as already mentioned, it's closer to how one works in the real world.
446.
by mh- 1y ago
..oh. That's clever. Thanks for posting this.
447.
by mh- 1y ago
> Currently I scroll and stitch the images together if a full page screen is required. Actually, I wish this was exposed as an alternative full-page screenshot method in CDP. The dev tools approach very frequently does not work with SP
448.
by mh- 1y ago
100% agree with that, as I alluded to in my last sentence. And that honestly seems like it might be a good product strategy in the short term.
449.
by mh- 1y ago
I don't know how to say this less flippantly, and I honestly tried: you could have simply posted a comment phrased as a question, and 20 people would have jumped in to answer. (To your point, >15 of them would have had different ans
450.
by mh- 1y ago
I never played much of RA2, but played hundreds of hours of RA1 skirmish. Must have been that. Thanks for the insight! Skirmish was a blast- I'd turtle until I had the enormous battleships (cruisers?) that could fire onto land. Loads o

More ›