I got all the way to round 53, but it turned out that one of my semiaquatic tetrapod ancestors from the Carboniferous Period didn't perform on land as well as they would have liked, so that was it for me.
"Traffic vioaltion fee" is a great translation. As far as I understand the logic behind them, they're meant for relatively minor violations where a fine would be kind of overkill and specifically have to be "directed" at the right person.
The downside is that unlike fines which scale by income here – the term is "päiväsakko" or "day fine", a fine unit that scales with net income – the fees are fixed sums, so unless a person with high income really does something heinous with their car, they're not as likely to get 200k€ (really) speeding tickets.
So now if you're rich you can speed all you want and pay a relatively small fee for it, as long as you're not doing 200km/h in a school zone or something like that
In Finland automatic camera fines (they're not exactly fines but I have no idea how to translate "liikennevirhemaksu" so work with me here) are the problem of whoever owns the car. If the owner wasn't the one driving the car, then it's up to them to inform the police who was actually driving
I'm sure some bright spark will soon show up to say that it was actually NATO who was violating our airspace for decades
, just like they're claiming that NATO is the one cutting cables here
Do you understand that this has been going on for much longer than the US's Venezuelan murder spree, and longer than Trump has been president (this time around)?
Also, as I said, we have a crew of a Russian-operated ship on the record admitting to cutting a cable by dragging their anchor, and all the previous cases have also been traced to other Russian-operated ships (well, I think one was Chinese though) using AIS and radar data, and this has been done by OSINT folks in addition to the local authorities here around the Baltic. Are all of these people being controlled by NATO and the US?
Pro-Russian people like you assume that other countries will always just let the US or "NATO" do whatever they want and have absolutely zero autonomy at all, and you're absolute experts at ignoring everything that doesn't fit your insanely simplistic narrative that's predicated on the idea that Russia is just a perpetual victim and a spooky spooky NATO CIA USA cabal is actually doing everything bad that the Russians get up to.
Yeah sure, we keep cutting our own telecoms cables multiple times per year, using Russian-operated ships as a front.
The Eagle S (I think it was?) case was brought to court here in Finland and they even admitted to dragging heir anchor but steadfastly maintained that it was due to their own incompetence (which the judge unfortunately believed.)
But the thing is that they aren't even making money; eg. OpenAI lost $11 billion in one quarter. Big LLMs are just so fantastically expensive to train and operate, and they ultimately really aren't as useful to eg businesses as they've been evangelised as so demand just hasn't picked up – plus the subscription plans are priced so low that most if not all "LLM operators" (OpenAI, Anthropic, etc) apparently actually lose money on even the most expensive ones. They'd lose all their customers if the plans actually cost as much as they should.
Apropos to that, I wonder if OpenAI et al are losing money on API plans too, or if it's just the subscriptions.
LLMs really can't be improved all that much beyond what we currently have, because they're fundamentally limited by their architecture, which is what ultimately leads to this sort of behaviour.
Unfortunately the AI bubble seems to be predicated on just improving LLMs and really really hoping that they'll magically turn into even weakly general AIs (or even AGIs like the worst Kool-aid drinkers claim they will), so everybody is throwing absolutely bonkers amounts of money at incremental improvements to existing architectures, instead of doing the hard thing and trying to come up with better architectures.
I doubt static networks like LLMs (or practically all other neural networks that are currently in use) will ever be candidates for general AI. All they can do is react to external input, they don't have any sort of an "inner life" outside of that, ie. the network isn't active except when you throw input at it. They literally can't even learn, and (re)training them takes ridiculous amounts of money and compute.
I'd wager that for producing an actual AGI, spiking neural networks or something similar to them would be what you'd want to lean in to, maybe with some kind of neuroplasticity-like mechanism. Spiking networks already exist and they can do some pretty cool stuff, but nowhere near what LLMs can do right now (even if they do do it kinda badly). Currently they're harder to train than more traditional static NNs because they're not differentiable so you can't do backpropagation, and they're still relatively new so there's a lot of open questions about eg. the uses and benefits of different neural models and such.
What about content that is illegal in the country that your "silo" is hosted in, like, say, CSAM (but you can really really substitute anything else illegal there, like eg. planning terrorist attacks)? If a "silo" is CSAM-friendly or its express purpose is posting it and its moderators don't want to remove illegal content, what then?
Also speaks to a lack of understanding on the author's part; people who truly understand some subject are generally much more adept at explaining it in simpler terms – ie without adding complexity beyond the subject's essential complexity
This'd be a valid analogy if all compiled / interpreted languages were like INTERCAL and eg. refused to compile / execute programs that were insufficiently polite, or if the runtime wouldn't print out strings that it "felt" were too silly.