Definitely this. Our cognitive capacity is limited, we'd better not deplete it due to complexity before we even reach the core of the problem at hand.

Nice article showing the steps to port Rust code to run on deeply embedded systems. It highlights the difficulties and caveats of such a task.

It's actually unsurprising, all those tech and crypto bros have assets in jeopardy if some regulation is applied to their industry. No wonder they'd support the one with the most libertarian agenda after the current administration which did look into antitrust cases and increased regulation (even though marginally).

Want to put an end to the social media platforms weight on our lives? For once there's an individual solution which might work. This is a chance because as he rightfully points out individual solutions are generally too complicated to bring systemic change. Here this is actually doable.

It's tempting to use uv. It's probably fine on the developer workstation at this point. It looks a bit early to use it in production though, it's a bit young for that and carries questions regarding supply chain security still.

Nice graphic tricks when the hardware was harder to work with. It's amazing how much we could fit back then out of sheer motivation.

Our craft is based on shifting sands. This brings interesting philosophical questions, like why do it at all? I think the answer proposed in this short article is spot on. It can help bring new ideas on how to be in the world. This is more important than the code itself.

Definitely a funny hack. Not usable for compute workloads though.

I tend to side on the "boring tech" side, but indeed this is a good reminder that what we want is finding the right balance.

This is too often overlooked, but table lookups can help with performance if done well.

It's a very important project, it's really concerning that this attack went through. The service is still partly disrupted but they're showing signs of recovery. Let's wish them luck and good health. This archival service is essential for knowledge and history preservation on the web.

A good reminder that this is not the Google Chrome alternative you're looking for. It's the same privacy invading mindset with some bigotry on top.

People have to realize that tycoons like the ones from big tech companies can both be rich and mediocre. They were smart enough to seize opportunities at the right time but they are not exceptional. In fact, they're even boring and spineless.

The best quote in this paper I think is: "There is nothing special about Elon Musk, Sam Altman, or Mark Zuckerberg. Accepting that requires you to also accept that the world itself is not one that rewards the remarkable, or the brilliant, or the truly incredible, but those who are able to take advantage of opportunities, which in turn leads to the horrible truth that those who often have the most opportunities are some of the most boring and privileged people alive."

The real problem is that lots of journalists can't come to term with the fairy tale and so fall prey to all their publicity stunts as if it had any hidden meaning. This is dangerous because of all the political power they try to seize for their own gains.

Meanwhile, "the most powerful companies enjoy a level of impunity, with their founders asked only the most superficial, softball of questions — and deflecting anything tougher by throwing out dead cats when the situation demands."

Now you can go and read this long piece.

Now the impact seems clear and this is mostly bad news. This reduces the production of public knowledge so everyone looses. Ironically it also means less public knowledge available to train new models. At some point their only venue to fine tune their models will be user profiling which will be private... I've a hard time seeing how we won't end up stuck with another surveillance apparatus providing access to models running on outdated knowledge. This will lock so many behaviors and decisions in place.

This is definitely a good idea, I wish we had the same in France. This is too bad that they plan to raise the price, it's going to limit the impact of the measure.

Finally a path forward for logic programming? An opportunity to evolve beyond Prolog and its variants? Good food for thought.

Of course I recommend reading the actual research paper. This article is a good summary of the consequences though. LLMs definitely can't be trusted with formal reasoning including basic maths. This is a flaw in the way they are built, the bath forward is likely merging symbolic and sub-symbolic approaches.

Nice paper which debunks the choice of the language as an important factor for energy efficiency. The previous papers had a too simple model, this one puts forth a more complete causal model. There are many factors at play regarding energy efficiency, the programming language itself is not really one of them.

Neat little introduction on color manipulation using matrices. Mentions the things to pay attention to.

Data layout is essential for performance reasons. It is too often overlooked. If you want real speed you need to help the memory subsystem.