Have you actually stayed at many AirBnB's? Doesn't sound like it. IMO, it's like a real vacation then. An actual house in the area you want to be in. You can pick amenities that matter to you - maybe a pool, or option to bring your dog, or nearby access to hiking trails, or lake access, etc.. TBH, the shittiest AirBnB I've stayed at was still as good or better than the Hotel options in the area. The one area where Hotels have a huge advantage is consistency - if you want to blindly pick a place to crash for the night, you can assume a $BRAND_NAME_HERE will be on par with your past experiences in other places.

Or maybe it's that they're capitalizing on said ill ease with AI, and leveraging it to instill a distrust of outsider AI's (Chinese, open source, etc..), enough so that they may manage to get the import thereof blocked, or at least pull a "nobody ever got fired for buying IBM" sort of management mentality.

Meanwhile, they didn't outline any features that existing worms and such are already capable of.

I may have misunderstood your original comment:

... misunderstood as the LLM directly recognizing some situation that is bad, as opposed to the LLM producing a result that can easily be tested for validity (like applying a test suite to a codebase and ensuring it still passes). If the latter, I mostly agree - with the same caveat you mentioned - your original "only" was a bit too restrictive, but not by much.

Agreed. And in those cases, a complex "if" statement can do the same job, but repeatably and accurately and cheaper.

IMNSHO, have this thing help identify those cases and guide in tool/monitoring solutions. Once a condition is determined, codify it - no need to have some LLM keep guessing at it every time. Hopefully, that's closer to what it's actually doing.

I may be arguing semantics here, but what they were doing in 2014 wasn't "AI"; That was "ML" - machine learning. But I'm also someone who tends to use "LLM" instead of "AI" when referring to most of the things labeled "AI" today, and I consider "SI" (per Executive Order 14434 - https://www.whitehouse.gov/pre... - aka. "Super Intelligence") to be even more backasswards (the "Artificial" part was the only accurate part of "AI").

Copyright applies to specific implementations. We're talking about product/game ideas. If one seems to be gaining traction, they can dupe it while dumping lots more resources into it to make it a polished and professional product with proper distribution channels and marketing/etc.., likely patenting as many bits of it as they can along the way and shutting down the original idea (or buying them out).

If it were simple to come up with a great new game idea, we'd be seeing them from the big studios instead of rehashing the same old IP. This seems like a great way to crowdsource game concepts and ideas, while simultaneously field testing them.

Won't surprise me if the licensing agreement means the get a royalty free license to your result. IE: they probably hope people DO create useful and profitable content so they can co-opt/steal it.

What's the next big idea? AI not giving it up? Get actual people to input the creativity into the system to solve that problem.

This sounds an awful like like, "Wait until the leopard eats your face before you draw any conclusions." There are well established benchmarks that provide insight into what we can expect, and that's a lot more useful than trusting the marketing and suggesting people wait for it to be released before deciding to buy the competition.

If you can't put two and two together and see what the benchmarks imply, that's on you. If you can, then do so and show where this went wrong (something like what AniMoJo implied, like maybe that benchmark was for a wildly different form factor or product... but was it?)

AFAICT, those "some models" that go up to 32gb don't exist yet. I looked at every one of them on google's site and none had a 32gb option.

They're advertising the local NPU. I know RAM is a shitshow right now, but if you're going to use it for local stuff, 32gb should at least be an option on all of these.

RR is a dick, and Coyote is a failure. Team Acme FTW.

While discussing US legal code, semantics matter.

Seems you've got nothing. Here, you're quoting a biased rephrasing rather than the actual quote. The actual quote from Jimmy is:

The "major political parties" is not the same as "competing political candidates", period. The part you quoted, "Candidates", and tried equating to "major political parties", is in the title of the US code, not the text thereof, and it still isn't the same.

The "competing political candidates" could include a Republican and someone from the Green party but no Democrats, in which case the "major political parties" would not all need equal opportunities (eg. Dems could be excluded if there was no Dem candidate).

Your second comment doesn't parse. I think that's related to your misguided idea that Kimmel chose not to broadcast Talarico's interview as a means of sidestepping regulation, rather than Kimmel planning on broadcasting it and the FCC strong arming them to prevent it from airing on broadcast. Do you have ANY evidence it was the former? Do you have ANY evidence that Kimmel or ABC did not offer a spot to Paxton as well (for broadcast)?

Actually... that could be a great way of making good of a bad situation. Bear with me...

Use it as a honeypot for shitty cops.
Keep the Flock cameras up.
Allow bullshit to be entered into request forms.
Periodically take action on those that enter bullshit into the request forms - mass firings and barring from future service.
If we're being really nice, maybe one round of warnings first, then the firings on the next review round.
Repeat the process with stricter controls on each pass (require verified ID, require a warrant, verify the user ID passes background checks, verify the warrant is valid and applies, add a judge to the loop, etc etc..).
If/when the system is no longer used frequently enough to justify its operating costs, so be it.

On a single user system, I mostly agree. But on a multi-user system, especially one where someone else also needs root for some action, sudo is a damned fine solution. You can lock it down to specific commands with specific arguments for specific users on specific machines, AND you get the logging as well (you are doing secure logging to a centralized system/cluster,right?).

A simple real world example: you've got a cluster of web servers and a support team that needs to be able to restart them.
Of course, there are other ways to implement that access (besides handing out the root password or using sudo), but go ahead and compare the options. Don't forget that you need admins familiar with those solutions, and there will be other cases needing similar solutions. A standard tool (sudo) with a config that can be copied around to all the servers (thanks to the host based stuff) is quite useful. If you only dish out specific access to specific people, and those are people you would otherwise trust with the root password, where's the downside?

That said, it's not a big stretch to say that the whole idea of sudo is fundamentally abused and misused so regularly that many would be better off without it enabling their poor security practices (Ex. "%sudo ALL=(ALL:ALL) NOPASSWD: ALL").

If you have something to say, just say it.