
Not just in this field. Crap is drowning out real work in the sciences as well as arts. Both in terms of where funding goes as well as the output. It’s a cultural disaster.

Not just in this field. Crap is drowning out real work in the sciences as well as arts. Both in terms of where funding goes as well as the output. It’s a cultural disaster.


Projection? Risk is played on a network with vertices representing the territories.

Me to, transformers, me too.
And by that we mean both that the people are dumb/gullible and that the LLMs are dumb/gullible.


Isn’t “intelligence” so ill defined we can’t prove it either way. All we have is models doing better on benchmarks and everyone shrieking “look emergent intelligence”.
I disagree a bit on “toys”. Machine summarization and translation is really quite powerful, but yeah that’s a ways short of the claims that are being made.
In a small company noone would try to label you “l5” or “l6” and probably an actual human would make your comp decision. You take the byzantine incentive structure away and people just try to do a good job.
Across all sports the US has over 3 times china’s all time medal totals in both summer and winter games. Not even having double is a pretty significant underperformance.
Maybe we should conclude gun culture makes a nation better at track and field because people get more practice running away 😂
Yeah fuck suicidal people and anyone killed by cops they had it coming. Jesus christ.
As for “justifiable homicide”, you mean like… when the other guy had a gun… You want to discount those deaths why?
If we take your number at face value. This is kind of a trolley problem thing right. Should 99 people get to have a nice day out at the range once a month, which they could easily replace with any other leisure activity, if it means one person has to die violently. If you honestly like those odds then fair play mad max.
You missed the point. Saying that calling it “fun” is problematic doesn’t imply that you can’t have fun, of course you can (I have, a tonne). I assume the above comment called it problematic because they weighed the “fun” of gun owners against 40k dead Americans a year and decided maybe we shouldn’t be focusing on the entertainment.
You can absolutely have more gun control and not really inhibit firearm sports or hunting BTW. The USA ought to have a monopoly on Olympic shooting medals if that weren’t the case 😅
Not wanting to to call the private ownership of deadly weapons fun is hardly “intense”. Of anything it undersells the point.
Speaking as someone who enjoyed regular target rifle practice and competitions for the best part of a decade since you insist that matters.


Agriculture is inherently “developing new products” you fundamentally cannot stop innovating because the growth of a plant is itself a selection step.
Take your point that it might discourage certain types of agri business. But aks yourself: given protection will they work to maximise the benefit to farmers or will they innovate to maximise their own profits (E.g. By creating quasi-monopolies or increasing farms’ dependence on them)?


Yeah I think if you peek behind the curtain a bit on the models it become harder to ignore the rampant hyperbole. Like really high dimensional transformers are a cool and powerful thing that didn’t exist 5 or so years ago, but it’s not magic and does have limitations. I guess the next 5 years will be a good test of whether we’re focused on the right things and allocating effort for the betterment of everyone when it comes to building and using models. I tend to think not but I hope to be proven wrong (and have no power to stop the hype train anyway).
I have views on vibe coding in particular since you mention it, but hey I do use an LLM a tonne while writing stuff myself so I guess there’s really only a fairly narrow margin separating me from the boosters.


Great writeup. Not sure if “conspiracy theory” is the right word, but i for sure see the magical thinking and the power dynamics described.


Yeah that’s the cleverer and faster bit. Generative AI has been a thing for at least 15 years (was in the AI field at the time and saw it presented at conferences). I would argue that attention structure in modelling was the new technology I guess. But most of the breakthrough is really just the money to scale.
Anyway pattern recognition is amazing and powerful and we can get excited about that without getting caught up in a superintelligence cult.


A normal full time job including holiday is only ~140 hours a month. Assuming you believe their user stats too that would be pretty grim. You probably shouldn’t though.


11 billion engagement hours a month! Does anyone seriously believe that lol. And still not profitable! What do they need, ever human beings every waking hour to spin a profit!?
Isn’t the biggest risk that you can’t shift your weight enough in emergency braking? The stopping distance must be horrible.
Yeah but how many users are actually that brave/stupid? 100kph feels FAST on a motorbike, I can’t imagine how it feels on one of these 😬


It’s getting a little shovey to be fair. The top of most Google results is an AI overview that you can’t opt out of. The top line of WhatsApp is “ask meta AI”. Meta is particularly bad for it, zuck explicitly wants to replace your real life friends with monetizable bots. Those of us working in big corporates are getting rapidly bored of the “assistants” and “copilots” that have nuked our tech budget and populate every app without assisting or copiloting in meaningful ways (and of correcting enthusiastically wrong output of colleagues blithely using LLMs for inappropriate tasks).
Some things it’s possible to just avoid but big tech is pretty pervasive. Tricky to really participate in society without at this point.
35 billion of 213 total. Wouldn’t call that most. I would guess it’s more important in it’s share of profit (but maybe not even right now with the capex explosion) or of profit/revenue growth.