You also have people who just like bangers and love throwing some tracks together, like to create sets or playlists, think they might be a DJ one day, to move the crowd, no time/inclination in making music per se but care about the created experience and will learn the 'plumbing' if they have to....
I had an idea the other day that children could have some kind of jewellery that they can't remove like a curfew device. Probably impractical but at least it matches - children=limited rights and need protection. Once of age, no need, come of age, can remove.
Children already have less rights. Like voting rights missing, the right to buy alcohol (at least here in Germany), not allowed to drive a car, Forced to go to school, not allowed to watch quite a lot of movies - just to name a few, but also special protections. As they can't be held responsible below the age of 14 (so not court cases and prison time for kids <14). Just as one example.
So we already treat kids quite differently from adults. And tht is - imho - fine. Because kids are not just "small grown ups", they are just growing up.
I just do not think, that KYC/age verification is the right way to go - or even feasible in how it is being approached. I am for denying Meta, TikTok and the others the opportunity to exploit kids. But I would rather much have them regulated in a way that just forbidds them their predatory ways overall, then to give them even more data in the form of verified idetities and the knowledge that this account is within a specific age range ad also a "real human being", making that data even more valuable for advertisers.
There is quite a difference between say having to hold a driver's license when driving or having to show the cashier your ID when buying beer, and having something attached to your body, for years, for the crime of simply having been born.
I abhore the idea of having something attached to ones body. I wanted to show, that we already have systems in place that work and do not need such a horrible intrusive way of marking children.
Yes! I was just thinking that - what if they become conditioned.. still I think it's worth considering, as a thought experiment in light of what seems to be happening - you get the same chains when you come of age instead, although they are virtual....
If you're concerned with the fact that the government might have misaligned incentives to use a system beyond the originally given intent, I think the actual approaches that are in danger of being implemented should probably be something you focus on more than something that one person on an internet forum casually suggested
well, I agree with this, however... if you make it soo bad that it seems that adulthood will be a joy, maybe it will reduce the crash as the responsibilities start coming in? The restrictions that would ensure should be pretty transparent along the lines of well you can't do that because you're a kid, right?
I was thinking that some kind of permanent physical attachment with passive electronics could be given to children, like an ankle bracelet used for home curfew, monkey's headband, a dogs shock collar, or just a nice bracelet, call it MoB, which couldn't be removed until they are of age. Devices they are given could be associated with those devices and not usable without them, if they disappear from passive scanning then they have been tin-foiled, etc etc. I've not seen any discussion of this type of approach which gives children something to aim for - freedom, and tallies with human historical culture as well.
go on, spell it out - its just a thought experiment - how is it worse than the death march we're already on? and the inexorable march to control adults, this is the opposite of that, and it reflects the control we're supposed to exert over our children, in a pretty neutral way (it seems - waiting for your arguments to burst my balloon!). Society and parents control you while you need a parental guardian, then once of age you're free to do whatever we like.
Something similar happened to me with twilio - years ago I had a number 'parked' with them, my card expired, they never notified me, account went into arrears, and they cancelled my number. I was quite a special number, a UK '0200' number they'd released somehow, which shouldn't have been released. When they then acquired sendgrid, to add insult to injury I eventually found out that I was blacklisted and had to go through various validation procedures in order to do business with them, which I declined to do.
Right, but if these things are so rare that we all only know the one viral example, I feel like that lends credence to the models basically generally not having this problem.
Researchers built the Winnograd Schema Challenge more than a decade ago to assess common sense reasoning, and LLMs beat that challenge task around GPT 4.
They're not so rare. Hallucinations have been spotted everywhere, but the "driving a car to the car wash" is an amusing one that's been recently publicised. Developers aren't going to point out every time an LLM hallucinates an entire library.
I'd add to this, any moderately involved logical or numerical problem causes hallucinations for me on all frontier models.
If you ask them in isolation they may write a script to solve it "properly", but I guess this is because they added enough of these to the training set. But this workaround doesn't scale.
As soon as I give the LLM a proper problem and a small part of it requires numeric reasoning, it almost always hallucinates something and doesn't solve it with a script.
If the logic/math is part of a larger problem the miss rate is near 100%.
LLMs have massive amounts of knowledge, encoded in verbal intelligence, but their logic intelligence is well below even average human intelligence.
If you look at how they work (tokenization and embeddings) it's clear that transformers will not solve the issue. The escape hatches only work very unreliably.
If you ask this of any current day AI it will answer exactly how you would expect. Telling you to drive, and acknowledging the comedic nature of the question.
That's because AI labs keep stamping out the widely known failures. I assume without actually retraining the main model, but with some small classifier that detects the known meme questions and injects correct answer in the context.
But try asking your favorite LLM what happens if you're holding a pen with two hands (one at each end) and let go of one end.
Not unlikely that you're talking to a lot of AI-based AI boosters. It's easier to create astroturfed comments with chatbots than fixing the inherent problems.
Nice. My test was always a blond bald guy. It always adds hair. If you ask for bald you get a dark haired bald guy, if you add blond, you can't get bald because I guess saying the hair color implies hair (on the head), while you may just want blonde eyebrows and/or blond stubble.
I believe that that's the stuff you buy in the shop, the non-refillable containers. If you buy a proper refillable balloon gas cylinder it's the higher grade stuff. Source: bought the shop stuff, got disappointed, bought the cylinder, happy.
