Very elementary. British “A Level” school course.
I do remember a physics graduate telling me his first year was getting everyone on the same page with Newtonian physics and then the second year left a chunk of the class behind as it blew past intuitive billiard ball school-level experiments. Was a while ago so things may have changed.
Cornell researchers found that at the current rate of AI growth, the burgeoning industry could represent 24 to 44 million metric tons of carbon dioxide emissions by 2030, the equivalent of adding five to ten million cars to US roadways.
4 Likes
Nice post from Henry Farrell on AI arms races, citing Alan Moore sci-fi:
“Beneath Alan Moore’s jokes, there lurks a very useful general point. Abelard Snazz believes that the solution is always more robots , even when robot solutionism is obviously the root cause of the problem. So too for the Economist , arguing that the solution to the problems caused by AI is more AI , with a healthy dollop of rights-pruning on the side. As a general matter, trying to solve a systemic problem by doubling down risks creating out-of-control feedback loops.”
2 Likes
The downside of thinking -ism can be added to any word!
/BoringOldMan
I find the original premise outlandish. Will the state really buckle if people use AI to challenge their parking tickets?
AI is not magic. Even it allows people to make “demands as well-crafted as a first-class lawyer’s”, if you have a documented parking violation with proof, what can it really do?
1 Like
Yes, it will.
The state bureaucracy is built to deal with a fixed amount of requests. It is built intentionally with the expectation that not all possibly requests are ever written - and laws are written with exactly that expectation in mind. Laws can be, and are being, sloppily written because of human nature.
That it is too much effort to request something small has always been a filter needed by the bureaucracy - and the justice system, too.
Without that filter we will see either
- a bureaucracy drowned in work, and totally unable to prioritize in any meaningful way,
- a growing bureaucracy.
- a bureaucracy augmented with AI, which will lead to an onslaught of requests written in a way to use the AIs weaknesses.
- a more or less massive reduction of state services, prioritized not by the needs of the people, but the needs of the system.
or, maybe, the typical German solution to try all after each other, without any cleanup ever.
2 Likes
No. Not outlandish. A post from September.
I think part of the problem is that the LLMs can generate plausible sounding complaints far better than they can distinguish between reasonable and unreasonable complaints. And it’s not really parking tickets that is the issue, it’s repeated planning disputes, labour tribunals and other areas that are far less cut and dried.
It’s true that the issue of, say, planning and nimbyism is a problem in and of itself, but already you get repeated consecutive planning objections to the extent that infrastructure builders will go to great expense to try to anticipate any such objections, no matter how wild and unfounded, pushing up costs and stalling projects before they begin.
1 Like
The fact that one of the primary features of an LLM is to be persuasive should tell you a lot about the intentions behind them.
4 Likes
LLM Machinations
this reminds me of a trip from a camping site near the Loire, France to my second camping site in Oberai. Google was giving me a route of ~845km (shortest option, still does that, crossing major cities (I hate driving through cities during a holiday.) Apple maps did slightly better by giving a route of 822km. My own physical map, hand picked turned out to be 691km, albeit an hour longer, with beautiful landscapes, no busy cities to cross and a perfect picknick spot.
The culprit…a sand road of about 500m, which honestly I did not mind driving at all. That sand road saved me a 22km detour.
I use AI all the time at work, but carefully checked for nonsense. It breaks at the boundaries of science.
4 Likes
I am not sure about the “lack of conviction” narrative. To me, it looks like a rational business cutting its losses. Yes, Google has always been known as being keen to experiment, but at the same time it has a history of abruptly abandonning projects that are not working out.
1 Like
It also has a history of abandoning products that are working. 
4 Likes
IME it has a history of removing product features and calling it an “upgrade”, but I digress… 
2 Likes
Yes. I think this is from a certain “AGI-pilled”, as the bros say these days, perspective. That’s its point, I guess, that the fact that Google has shareholders (and not powerless ones like Meta or SpaceX) means it can’t set fire to other people’s money like OpenAi or Anthropic to stay at the “frontier”.
I heard an interview with a veteran VC guy and he was saying they were only really interested in crazy ideas because if the idea wasn’t crazy, there would be other people trying to do it and then you might not get a de facto monopoly (he didn’t say monopoly because he knows that’s a bad anti-free market idea but that’s what he meant).
Most, obviously, fail - and then you get an Airbnb (which he was referring to specifically) or Uber.
I’m not 100% sure I buy it, but it sort of made sense of the insane commentary in and around the SpaceX IPO and prospectus.
It wasn’t an IPO in the traditional sense, it didn’t have any track record on the market before it entered the index, it’s more or less pointless to try to value it using normal discounted free cash flow because the valuation makes no sense, it’s just a crazy story and you buy the story or you don’t.
Tim O’Reilly echoing your point:
“In short, SemiAnalysis is probably right that Google gave something up on August 5, but possibly wrong about why. The evidence does not necessarily show that Google has abandoned technological ambition. Sergey Brin still seems to have plenty. But at least for the moment it may show a shift in where Google is allocating scarce AI compute and where it expects to capture economic value. It may be a mistake to assume that the AI race is about who builds the best intelligence. It may turn out to be about who builds the electrical grid. Maybe Google didn’t decide to stop trying to win so much as to decide that a different race was more worth winning.”
2 Likes
Browsed in passing, and not worth backtracking to add the link but…
Guy asks Claude Opus 5 to do a backup. It realises that it put it in the wrong place and removes -r his home directory. responds, Sorry, typo.
It might just be me being characteristically slow, or may be the sort of haze that AI issues seem to cause, but, about twenty minutes later, my brain says…
Wait! This guy has a computer and he doesn’t know how to make a backup?
Earlier today, I read that someone I know changed the car he intended to buy, because AI told him he was making a wrong choice. Apparently, it looked at his needs and was right, but that isn’t the point. What next? What breakfast to eat?
5 Likes
Maybe what area of a photo to edit?
2 Likes
Maybe what area of a photo to take?
1 Like
I have a friend who goes food shopping with ChatGPT, presumably the speaking version, so it can scold her for putting unhealthy treats in her basket.
3 Likes