> Really I have no idea what this case accomplished besides being a tax on a successful company.
You forgot "evil" in your description of the company.
It is supposed to realign their incentives away from massively exploiting society for purely financial gain. There is ample evidence that Meta has gone out of its way to encourage the discussed issues, so clearly there was a problem with the incentive structure. Given all the lobbying they've done to prevent any passage of laws regulating their conduct, it's fallen to the courts to correct this (if, as other comments say, somewhat unsatisfactorily).
Unit testing can get you some of the way, but its not a full-on all-case guarantee. Sometimes the code is encapsulating some particularly complex system/behaviour. Sometimes the reason is interop/compatilibity issues or some kind of politics.
P.S. you managed to introduce a typo in your quote (were –> where)
I am very much weirder out by it, yeah. Seems some societies are just excessively desperate for some kind, any kind, of fuel for economic growth, to the point this is where attention is now. The term "post capitalism" being thrown around feels less ridiculous than it did in years gone past.
Truth is even weirder. plenty of growth is possible but modern liberal democracy requires that all progress must be contingent on the production of enormous amounts of text that nobody would ever read.
3 years ago, the Wall Street journal covered a company trying to use AI to generate documents required for the approval of new nuclear reactor reactor designs, which sounds dangerous, until you get to the point where they'd cite 2 million pages as necessary for a typical application. [1] I don't need to explain why no individual or institution could read that, much less examine it in detail. I remember a rather funny question I found in a comment to that story - "How many pages of those 2 million could contain pornographic images before anyone notices?"
It's obvious why companies are so desperate for training data - a text generator of sufficient quality is more conductive to the growth of the nuclear industry than any scientific breakthrough in nuclear physics. (And of course, if you want to prevent the development of a nuclear reactor by your competitors, being able to produce millions of pages of high-quality objections will do the trick.)
And it's not just nuclear power. When it comes to stuff like building rail lines, apartments or power plants (both conventional and renewable), you'd find that the main bottleneck is the necessity to produce documents. And of course, many documents can be subject to judicial review - a process that consumes even more text.
Well that's why a plan can take years to be approved.
There also is the liability angle, where no one necessarily reads the document until the document is relevant to some situation in the future. Then you better have that document on hand.
According to the article, the “application” (which I imagine was also broken down into smaller pieces as they went back and forth to regulators during the review process) was 12,000 pages in total.
The 2 million number is for “supporting” materials which isn’t really defined. It definitely makes for a better soundbite to claim millions.
> its 12,000-page application had around two million pages of support materials.
Today we are living this dystopian life. Imagine you train an llm on all the millions of complaints that say comcast or some other semi-disfunctional company gets. Or police reports of accidents. Or suicide reports, or any of the endless human misery that is out there in the world.
Even though I'm fortunate in so many ways to be in the west - even with the current destructive us government situation, the world is still full of endless misery that I don't ever experience. Like someone living in a war torn country, Gaza, lots of other places.
Any kind of fuel for giving active investors that FOMO tingle which then forces the steamroll of index funds to blindly follow.
I guess the appropriation "any sufficiently advanced stock market is indistinguishable from entertainment" doesn't quite stop at equating the trade floor with a casino. At some point, entertainment also becomes the modus operandi of corporations.
> So... did they list the "0GB" racks with removed ram cheaper you ask?
Could this be an indication that there's a supply shortage/price increase of the remaining components in the rack? Or do you interpret it as people just trying their luck out of sheer greed?
My understanding is that it's fundamentally a fab capacity limit. At a given node size/technology (think 7nm, etc) they can only output so many wafers. If AI data centers are generating enormous demand for memory, then that's what those wafers are going to be used to for, unless they can be outcompeted, hence higher prices.
I'm not actually sure if enterprise designated memory is equivalent or not to consumer designated. Could be they are the same just packaged differently and/or with better binning. Either way, it's the same supply/demand dynamic at work.
Yeah a RCD on the memory stick, it's PCB is designed around it. The actual DRAM chips themselves are not different though, which is why I mention it would be super easy to produce more consumer unregistered and less server registered at assembly time :)
As long as you are consciously aware you're not really talking to another human but just entertaining yourself I don't see the harm tbqh.
Anything can become addictive, also i'd say given the costs of operating llm inference the labs themselves when offering the products try to limit your usage at different levels, ChatGPT for example will recommend you take a break and chill if you go hours with it (even with a 200 plan)
I don't see how using an LLM is that different from surfing Wikipedia or Google, in all cases the computer is just giving me outputs i demand it from, in some or other shape, and can be abused etc
This is a strong assumption; we know that people tend to be overly confident about their mental states. That conscious awareness will last all of two minutes for 99.9% of the population, with everyone claiming they are the 0.1%.
> the labs themselves when offering the products try to limit your usage
Yeah clearly as long as they have compute constraints, ChatGPT is for sure not doing that out of concern for the user's health, believing so would be incredibly naive.
> I don't see how using an LLM is that different from surfing Wikipedia or Google, in all cases the computer is just giving me outputs i demand it from
This is by far the biggest stretch of all of your arguments. Following this logic, you go to the library and find a book that gives you said outputs - are books now == LLMs? You ask someone for the time and that interaction == LLMs? There is evidently a major qualitative difference here, starting (but not ending) with "you are absolutely right".
As usual a balance is needed. There is also a concern that society (friends, colleagues) that always shuts down your ideas just because they are different thus reducing creativity and progress. Sometimes you need a way to develop your ideas without the outright rejection and sometimes you need the filter of society to see if the ideas really hold up. Usually the time where ideas are most fragile to being destroyed is when they are just developed. Perfect time to brain storm and increase the strength of those ideas with AI before taking it to the filter of society.
Agreed about balance, but you lost me after that - I really struggle to see a serious concern about society over-eagerly shutting down ideas, not when we live in a world in which all manner of ideas - including those that are disruptive and destructive - are by default celebrated for their "innovation". Any existing filters cannot be that stifling; we are hardly living in the Dark Ages.
That's clearly not the point - it's not a function of everyone knowing, but rather a function of the attempt to cover it up. Waving defeatism as some kind of universal truth is neither particularly enlightening nor useful.
It's not clear to me that this is a result of corruption - the article notes that a big driver is overtime pay, and that SF has struggled with hiring in recent years, which could be a legitimate cause.
Also note about the mayor: "Lurie is required to receive that salary, but donates nearly all of it back to the government, accepting a net pay of $1 for the calendar year".
There were a number of reports and observations some years back of NYPD officers gaming overtime pay to pay themselves more, in some cases even claiming it while doing other jobs (like bouncer duty, private protection, etc).
Yeah so this would certainly be unacceptable, especially for NYC tax-payers struggling to get by I'm sure. My point is that the article leaves open room that the situation in SF isn't necessarily only due to unscrupulous behaviour. With that said, if there are other sources that indicate that this is happening in this case, I wouldn't be overly surprised either.
You forgot "evil" in your description of the company.
It is supposed to realign their incentives away from massively exploiting society for purely financial gain. There is ample evidence that Meta has gone out of its way to encourage the discussed issues, so clearly there was a problem with the incentive structure. Given all the lobbying they've done to prevent any passage of laws regulating their conduct, it's fallen to the courts to correct this (if, as other comments say, somewhat unsatisfactorily).
reply