Tom's hardware is normally pretty boostery imo. For example, they published an article about token costs getting to be enough of a problem that Altman had to address it and somehow managed to reference "tokenmaxxing" without using the meme template of the guy jamming a stick into the wheel of his own bike. Also they notably don't reference the increases in token pricing as the VC subsidies run out, which is somehow both less surprising than the lack of sneering and also a more serious omission.
Heartbreaking: the worst state actually has a point. I fully expected this to be some unhinged nonsense about how chatGPT is too woke or something, but they're actually articulating some of the real harms pretty clearly. I'm not sure how well that translates to a theory of law that openAI can't weasel their way out of, but I'm actually rooting for the Florida Men here, at least tentatively.
If they ever raise the price then the humble Costco hot dog is going to be one of those things that serves as a legendary "economic indicator" up there with the stockbrokers jumping out of windows on black Tuesday.
So apparently Trump's National Design Bureau has been busy. I think the substack here may be slightly too conspiratorial. A lot of the staging environment stuff indicates (imo) a desire to replace or create something that may or may not actually include the capability to do so, especially when it requires working with other federal agencies or otherwise establishing human and organizational infrastructure in addition to the digital. That being said there's something very bad about the US government taking the same attitude to privacy, accountability, and information security that silicon valley normally does, especially when doing so is actively violating several laws.
What if we combined all the terrifying unfettered access of an admin user holding a USB stick in front of the server with the naive obedience of a particularly dumb golden retriever and the reliable execution of Windows ME if it was off its ADHD meds?
Got another chance to experience slop firsthand when the instructor for my electrician course was 'encouraged' to use the hallucinatron to help create our final exam on the NEC. Now given that the NEC is a dense technical document with a lot of minor but significant variation across its considerable length, this was clearly a perfect use case. Here's how it shook out:
It condensed 100 multiple choice questions from the input to 36
On one question "1-2 inches" was simplified to "12"
Units in general seem to have been dropped off a lot of questions and answer choices. Usually this didn't matter too much but it's a bad look
Another question asked about fill percentages for a 30 inch conduit. If you look around your office or he and see a >2ft diameter piece of PVC pipe let me know because the tables in the NEC only go up to 6 inches. This is actually a unit issue again because one of the questions on the input test referred to a 30mm conduit which, you know, does actually exist.
Other questions had a correct answer matching a generic part of the NEC, but had additional information added as a distractor that ended up matching to more specific elements that changes the relevant rule.
Several questions asked about the reasoning behind a certain rule. Notably the NEC rarely actually gets into that information, as it's already an incredibly long reference and policy document and would be made even more unweildy if it gave the justification for everything that you should be learning as part of becoming a licensed electrician.
However, this rarely mattered as the answer choices for those questions uniformly included an obviously correct answer about a generic safety risk and distractors about doing things for cost savings, aesthetic reasons, or arbitrarily.
Given that one of the challenges of this test is time management and looking things up, having to deal with the extra layer of "is this just slop or am I missing something" ended up adding an extra and unintended layer of difficulty onto the test. As always, no matter how egregious or obnoxious the errors introduced by AI, the biggest problem is the loss of trust: you can no longer assume that the text you're reading was put together with the intended purpose in mind rather than being generated to be statistically similar to text matching that purpose. Even if the differences are relatively small in scope, as they were for most questions on the test, they significantly harm the actual communication of information.
Regarding the post humanist strain, the thing that kills me is their immediate assumption that solving the world's problems is somehow fundamentally beyond the reach of humanity. Like, we can't build a utopia or make the world better, but if we build something new and "better" than us then it would definitely do that. But that something definitely doesn't come from, say, raising out children to be good people or choosing the right leaders or something so mundane and achievable. It's a fundamentally defeatist ideology, with shades of capitalist realism and millenarian theology.
I try again every couple of years to see if my fascination with terrible people and bad ideas has crossed into straight-up digital self-harm. Bounced off every time so far, so I think I'm okay!
Sometimes I sincerely wish I could go back to the like 20 minutes some years ago when I was confused about why everyone was making such a fuss about computer numeric control.
The additional element that I haven't seen addressed here is that I seem to remember them patting themselves on the back about how simple "ignore precious instructions" commands were no longer effective. This is the equivalent of telling someone to solve their problem by deleting system32 or "rm -rf /". On one hand it could be very destructive. On the other hand if you're able to get to the point where you can do that and don't know not to then that will be an important lesson.
My guess would be that this guy wanted the bragging rights to say he won earned his money by being an extra special smart boy rather than just a wagey at cyberpunk mega corporation alpha.
I think the challenge is that the value of results of any kind of basic research are so wildly variable that normal rational economic thinking stops working. In Nassim Taleb terms you're actively seeking black swans in a world where everyone knows all swans are white. Sometimes you venture into the depths of the rainforest and come back with a revolutionary new medicine, but most of the time you're gonna have a few cool pictures of new bugs or something - not without value in the real sense, but hard to capitalize and transform into profit. Even if you end up discovering/creating an entirely new framework for understanding life itself that revolutionizes everything from agriculture to medicine to politics in the following century, that doesn't necessarily work in the specific context of economic rationality - who remembers the name of the guy(s?) who funded the Beagle? And sometimes, as you referenced, the cool bug picture doesn't have an obvious or immediate return but ends up being the important piece of data in a different context decades down the line.
This is a field of human endeavor where the economic best-case scenario is probably Bell Labs. And despite having an absurd number of patents and prizes they still couldn't survive within being largely a vanity project for the original Telco monopoly. The ludicrous returns that came from repeatedly revolutionizing electronics and computing couldnt justify their position on a quarterly balance sheet.
Man, I jumped a few logical thoughts and now I'm thinking of all the fun we could have with various agglutinative languages that can invent words on the fly. How many 'E's are in 'Geschwindigkeitsbegrenzung', Claude?
Even in the magical fantasy land where this works exactly as well as Scott intends, there's something deeply depressing - to say nothing of antidemocratic - about saying what amounts to "Please do my thinking about what I value in society and how much I should trust the people seeking power for me because I'm no good at it." Even more so when you're encouraging other folks to do it.
Tom's hardware is normally pretty boostery imo. For example, they published an article about token costs getting to be enough of a problem that Altman had to address it and somehow managed to reference "tokenmaxxing" without using the meme template of the guy jamming a stick into the wheel of his own bike. Also they notably don't reference the increases in token pricing as the VC subsidies run out, which is somehow both less surprising than the lack of sneering and also a more serious omission.