Go read the full paper of course, but in case you won't this is a quick summary which kind of does the job. There are obvious limitations to the study: most notably time as it's only a few minutes use each time. And yet that raises clear questions on longer term use... There is clear recovery after a few minutes but does it still exist after weeks or months of prolonged use? How fast would it be? Clearly we don't know yet.

Interesting essay even though I don't like the whole part about singularity. I found this section was a distraction to the more interesting arguments. Indeed, I liked a lot the concept of decoherence and its ties to complexity.

The essay also makes a good job of showing how it's something we have to deal with and likely the worst kind of technical debt. I also shows how it impact productivity obviously. It's all valuable and things I encounter on a regular basis.

And of course, since everything has to talk about LLM these days, it build an interesting case on how and why using those tools especially with harnesses to pump out more code is actually increasing decoherence. Exactly what we don't want.

This has a satirical undertone but it has a very good point. Those companies are just sucking the life out of very human activities.

I'm not agreeing with some of the points in here. That said the core position of "it should be about collaboration not automation" I'm very much aligned with. It's after all at the center of how we designed AI systems in my team twenty years ago. We really wanted systems helping the user in their task and so co-constructing the result with the support of the machine rather than some kind of "oracle"

Between this and the scrapers behavior... the disregard for infrastructure and best practices is obvious. It's just another externality those so called "labs" don't want to deal with. And well, it makes for a marketing stunt each time isn't it?

Interesting opinion piece on how autocracies benefit greatly from the current AI arm race. But the short term benefit will also create new weaknesses.

This is a rant yes but well deserved. Those companies selling giant LLMs are just DDoSing everyone...

Looks like the mess on hardware prices will continue for a while...

It makes sense to me, it's better to not fall prey to the marketing induced FOMO. Instead projects have a chance to do the right thing.

There's clearly a major issue in terms of environmental impacts with the AI industrial complex. Yes it's also about electricity consumption but not only. We have at last a comprehensive study on the matter. Without major changes it's just unsustainable and something we can't collectively afford.

Kind of an unusual read. But I think the message is powerful. We should definitely favor community over tools controlled by the new robber barons.

The slippery slop seems clear... If you start generating a bit you'll be pushed to loose full control in the end. Do you think you can learn this way? Improve? I doubt it.

Aren't we loosing sense of community and our humanity indeed?

Since I've got my PhD in AI almost twenty years ago now, this piece strikes hard. This fellow kept doing research while I didn't (the reasons to be told another time but it's linked to this article). Clearly AI researchers who started around that time or before have a very different vision to what "AI" means as a field (although I've a few difference I'm fairly aligned with this article)... Clearly we've been robbed of the term by the industrial complex.

The hype is very high right now indeed. This can lead only to misguided decisions. But really it's mostly distraction from the massive ethical concerns this industry carries. It would require for policymakers to step up but I'm afraid it unfortunately won't happen or they'll come up with misguided attempts if they don't first consult with field experts.

Looks like an interesting model to build decision assistance systems and classifiers.

Goes maybe a tad too far in the opposite direction at times leading to clunky phrases. That said this effort of naming things properly is badly needed. I hope the media will seize it and report about those tools properly.

Interesting initiative. I wish it'll be popular and many people will follow it. Maybe it'll cure some of them of their AI psychosis.

So apparently the low adoption is due to the lack of trust in the systems and regulations could help? Who knew...

Now this look like a very interesting and useful model you can run locally. Definitely helps for classification tasks or assisting in decision making.