But even the series handled this problem fairly well in the first couple of seasons by establishing a "senior hierarchy" of (powerful) races and aligning the Tau'ri (humans) with them as sort of a client race. I think they could have gone much further with that, but unfortunately didn't. Re-imagining the series with that constraint would still be a very interesting premise, although I don't think that would work in a continuation given how the series itself evolved.

It isn't clear what "perfectly random" means in any way that brings utility to the workflows, processes, and measurements that are implemented by products and systems in the "real world".

What's been good enough for a long time, and likely for a long time to come, is sufficiently random to meet some criteria in a risk/consequence trade. The quest for "better randomness" or maybe "perfect randomness" is just a means of saying I can completely retire one risk that is part of a number of risks in a risk assessment...the problem is most of the risk isn't from insufficient randomness, it's from other parts of the process.

Most of what you said is true through the 40xx series cards, but with the 50xx series, apparently all the AI they were putting in driver development came back to bite them in the ass (note: mostly a humorous assumption on my part, maybe true, maybe not). I don't have any interest in overpaying for hardware that has poor driver support...it's why I avoided AMD for quite some time. My last three video card buys were AMD (7900xt) or Intel (2xARCB580), and they've been more than good enough in SWAP-C, esp. given the ridiculous pricing on nVidia cards.

Combine that, and nVidia's clear dis-interest in the gaming market that "made" them, and it's unlikely I'll support anything other than the competition until after this AI bubble bursts and (hopefully) nVidia get's their crap together again. If they get out of the consumer market all together, it'd probably be good for the market as a whole at this point.

I think the key here is consent . If I started working at a company and such monitoring was not the norm/expectation/condition under which the offer was made, then my very reasonable expectation is that such monitoring is not part of the work product I am providing the company. Unilateral modification of a contract can be legal ("not a lawyer") but certainly isn't ethical without additional disclosure(s). It is therefore very reasonable that if one party is modifying a contract unilaterally, that some degree of compensation be required to offset the additional constraints/work products/expectations being imposed.

That said, if such monitoring was the disclosed expectation under which an offer was made and accepted, I think I'd pretty much agree with your point.

Oh come on, you mean all courses are now going to have a 5:1 student to professor ratio?

Where's the evidence that AI is replacing entry level jobs? Entry level jobs exist to get employees experience so they can be productive. Might it not be that the market conditions are such that the companies don't need more people? It is the norm to keep your experienced, productive employees during those times. It certainly can create long term problems for companies if they continue that behavior for too long, as it leads to an experience gap in the workforce.

Corporations are laying off people, but it isn't clear AI is even indirectly responsible for that, although it's certainly being used as a scapegoat, esp. by the AI pushers that want you to have more AI--I mean, it's what they're selling, it MUST be something you need, right?.

Where are my mod points when I need them. There's DEFINITELY some smoke from something, but it might just be a crap ton of GPUs on fire.

While you make some valid points prior, where the hell does this come from? I don't know about all countries, but the US at least has a progressive tax system. The total tax burden on someone making 50k a year is probably around 6k a year. Someone making 250k (5 times as much) is paying around 60k a year (10x the taxes as the 50k per year, i.e. twice the tax rate). NOTE: estimates from DDG.

Regarding "productivity" I haven't seen any real hard data showing productivity increases that relate to realized increased revenues for business that are not tied to cutting costs, i.e. "fire someone replace with (less capable) AI". If that data exists, please point me to it, I'd be interested in seeing it. That sounds like a short term savings while degrading the quality of the product being offered.

It seems reasonable that AI is replacing some jobs that really are completely obsolete, but it isn't clear what those jobs are, and if it's really "AI" or just other machine automation being lumped in the AI bucket (e.g., transcription, which automation and/or AI does "OK"). It also seems reasonable that what's really going on is companies embracing "AI" as an excuse to do some downsizing they likely need to do in any case, and what's probably going on is a work force reduction due to market changes...but I also note I'm an engineer that is only interested in what utility AI actually has and not a job market analyst.

It's not at all clear to me that the vast majority of "human intelligence" is really all that complex, and could at a fundamental level be (in many cases) almost as simple, or even simpler, than some of the existing LLMs.

The vast majority of most human interactions are handled at a level that typically doesn't require any thought at all, driven by ingrained correct/incorrect patterns we've imprinted mentally in some form...true thoughtful innovation seems to be and have historically been, very rare.

The fundamental difference that is pretty obvious is the ability to "train" or "force rapid learning" upon the various LLMs that are all the rage today...but at a low level it appears to be a complex version of repetition of existing patterns, which seems similar to typical "human intelligence". It's why I question if accessing content (e.g. "NYT archives" or similar) for AI training can or should be different than accesses used for human consumers...the only real difference is the rate of consumption.

It all seems to originate from an X post (linked above), where the poster simply claims that "They wrongfully froze all of my tokens, stripped me of my right to vote on governance proposals, and have threatened to permanently destroy my tokens by “burning” them—all without any proper justification."

There's literally nothing else here to even hope to understand what is going on, other than one guy claiming he's getting screwed...and a bunch of anti-trump slash-dotters saying WHAT DID YOU EXPECT

You'll have to point at something that backs this statement up, since it seems fairly clear to me your statement could follow simply from the premise, i.e. people that arm themselves may be more likely to be attacked by armed individuals in general, and that's why they arm themselves. Additionally, it seems likely that gun owners do have an increased risk of some kind of gun-related incident, simply because owning a gun implies accepting the risks of having one (e.g., being subject to potential accidental discharge, amongst a host of potential other things).

A quick googling (duckduckgoing, technically) of "are armed people more likely to be killed by armed assailants?" found these in the top responses:

The Science of Guns Proves Arming Untrained Citizens Is a Bad Idea, an article that quotes statistics that purport to say it's a bad idea, where many of the items quoted don't necessarily follow based on the data.

and

Do Armed Civilians Stop Active Shooters More Effectively Than Uniformed Police?, an abstract that purports that "Civilians with permits stopped the attacks more frequently and faced a lower risk of being killed or injured than police"...but the link itself is only an abstract. In the article, it extends the FBI data with other data sets (likely ones supporting the claim).

NOTE: Did not dig deep into either article / publication.

In short: good data is hard to come by, data sets are not typically exceptionally large, and biased views seem to be the norm here ("left" or "right").

Take it a step further:

You hire someone to manage/investigate a specific thing, and 90 out of 100 things that cross their desk every month (yes, yes, numbers made up) are completely trivial, and can be handled by the one person you can afford to pay to handle this kind of thing. Now that person wants to spend all of their time handling the 10 things that aren't trivial, and doing 9 of those 10 things would take ALL month, leaving 91 things left undone. What's worse, that 1 remaining thing would require the whole month on it's own. Suddenly you have one person trying to do the work of at least three people, and you can afford one person...and now that one person wants to spend their ENTIRE MONTH on that 1 thing, leaving 99 things left undone.

To me it looks like a Pareto numbers game, and unless you're will to invest a large amount of time and money to get answers, it'll always be more efficient to maximize the number of answers per unit time per cost invested, which means you'll only ever get to address the most trivial things in the inbox.

Don't see those too often, but they do happen.

Wikipedia on Betteridge's Law

Wasn't this whole approach to "repair" that's being used here enabled by the DMCA? So I don't think we can say something like "crime doesn't pay" when this company has just taken a horrible law and applied it to profit making. Is what they're doing unethical as hell? Oh Yeah. Illegal? Probably not.

Don't get me wrong...no one should be doing business with a company that takes this kind of position ("screw our customers all day long 'till the cows come home") w/r/t such expensive investments that are typically required for a small (farm) business to operate.

Repeal the damn DMCA already...it remains a bad law and it should be gone. Happy to see this outcome, but really, it isn't enough.

It's a systems engineering trade. It's been a while since I've worked on a space system, but...

You might bother because transmitting all that data down and waiting for a response takes more time than you have if you want to respond on the current orbit.
You might bother because the power cost of transmitting the data is on par with using stored energy on the platform to perform the computation locally.
You might bother because you want to be more efficient in your use of your communication link regarding the function you're performing on the AI HW (i.e. reserving link bandwidth for other stuff).
You might bother because you need to generate some heat to keep your electronics warm, and you get free results that would incur other costs (see above) when doing so.

There's probably a number of other reasons you'd bother, but I do agree AI in space seems to be the tone of the headline, and doing that just for the sake of doing that is kinda dumb.