oersted

Very Crusader Kings :)

Frankly, it sounds like you are being a bit short-sighted with the whole endeavour.

The lowered production costs enabled by AI should to a degree insulate you from the usual financial risks, and should allow you to be more authentic. It might seem risky to make niche content, but I think it could help attract a real initial audience, never underestimate the demand for true authenticity.

It might sound strange talking about authenticity when it comes to AI, but I do agree with your perspective, that it can empower creative thinkers to produce content with real authorial intent directly, and without the compromises inherent to large production teams.

You seem to be operating like a corporate production company with large costs that needs to appeal to the lowest common denominator to get ROI, but you are doing it without the benefits of the higher production budgets.

You are wasting your advantage. Do the things they cannot afford to do because they are too risky or too niche. If you say that AI allows you to empower your writers, don't waste their talents on brainless content!

I must admit, the short sample for “How Asians Pay the Bill” was kinda fun :)

That’s a good visualization, although I am a bit mistrustful of Arena’s scores. It does get around the fact that models are getting trained for the benchmarks, but the methodology of letting random people compare outputs side-by-side is a very shallow judgement method in my opinion.

EDIT: Indeed looking at the overall rankings for text again, the list is rather strange, a lot more about writing style than intelligence.

I was confused too, but I believe that it refers to the Pareto Frontier: the best set of solutions to a multi-objective problem.

Look it up, it’s a bit difficult to explain concisely in words but it is intuitive visually.

If we are thinking of intelligence and price, a model will be in the Pareto Frontier if there’s no cheaper model of the same or higher intelligence. Or if there’s no more intelligent model for that price or lower.

EDIT: See this chart from Artificial Analysis: https://artificialanalysis.ai/#intelligence-comparison-tabs

So for example DeepSeek V4 Pro can be considered a frontier model because there's no cheaper model that is as intelligent.

For any solution in the Pareto Frontier, there no "no-brainer" alternative, in the sense that there's no other option that is better in some way without giving up something else. It's the best of its "weight-class".

I liked this comment from that thread :)

I think you should spin up a whole bunch more instances, and try to cause an integer overflow so they they owe you $978 Trillion.

You can use tokio::spawn_blocking to separate compute-heavy stuff, it gets handled in a separate thread pool. You shouldn’t need two tokio runtimes, that sounds problematic.

How do you properly mix IO/compute (in any language)?

Good question, it’s a really hard problem.

In principle you are meant to use threads, the OS will automatically switch to other work while you are waiting for IO, like with async. But there are difficulties with threads too: memory overhead, switching overhead, thread limits… Although I am getting the feeling that they have a much worse reputation than they deserve. More projects should try switching to threads and actually measuring the difference.

That does clarify some things, but still. Say you have an Isolate pinned to a core and it's doing some long compute. In the meantime some background IO finished and an Isolate in the same core needs to handle the result. That will not happen until the long compute is done.

Isn't it just like in async and any other cooperative concurrency model? At least in multi-core async, that message can be handled in a different core so it's not completely stuck. But sure the author doesn't like that work-stealing cost happening automatically outside of their control, fair enough.

Yes I do agree. It's not that Tokio has taken over just due to community momentum. There's a certain subtle lock-in that happens due to how the Rust type system and dependency system work together. It's a hard problem to address.

Regarding mixed IO/compute and preemptive scheduling, well that's what threads are. They are not as lightweight, but they are there and they are quite ergonomic to use in Rust.

I could even imagine an async executor that simply starts a thread per task, so that you can keep the nice syntax, but that's obviously not ideal from a performance standpoint. Tokio already dispatches tasks to a thread pool, it is not completely cooperative, it's the sensible way to do this. And there's always tokio::spawn_blocking too.

PS: Actually, it is underrated how well designed classic threading is in Rust. It was just going out of vogue when they did it, but they addressed so much of the complexity that was around for years in Java, C++ and the like. They did actually mostly achieve the holy grail of fearless concurrency and with a more ergonomic design.

I was looking forward to checking out Project Tina until I realised it was a completely different language. Classic story. Surely you can build a thread-per-core message passing concurrency framework in Rust, the language is designed to allow such alternatives.

Is Project Tina a bit like the Actor Model, but having actors pinned to cores?

And I don't understand how Tina deals better with the problem of compute-heavy tasks blocking the thread. It looks to me like it is also cooperative concurrency per core, and if one Isolate runs for a long time the other Isolates in that core will not be able to handle their messages.

I agree that a request-response model can be quite ergonomic for concurrency. I've used this pattern in quite a few Rust projects in a somewhat ad-hoc manner.

It's a bit like having a system composed of microservices (nanoservices?) that communicate via function calls.

It sounds a lot like the actor model, but I always found the classic architecture too limiting: requiring every actor to be a single-threaded message processor, instead of being able to handle requests concurrently. It's not too different from classic object-oriented design either, with singleton services.

In some project I've called my concurrent services Gods just to have a bit of fun with it :)

To be fair, the Rust async model itself was intentionally designed not to be prescriptive in the way you describe. You can build, and there exists, different task executors that can handle things like priority and many other execution models.

Async is just a way to describe a tree of concurrent tasks that may depend on (wait on) each other at certain points. It is mostly declarative.

Tokio has taken over as the default choice, but there's a reason why it's not part of the standard library, it is not meant to be the only choice.

We are creating an AI for science and engineering: https://vicena.ai

It's connected to all papers of course, and all kinds of scientific simulators and specialised models. But I'm currently in Shanghai talking to labs to join a CloudLab (and hopefully setting up our own robotic labs), so that AI can actually order real physical experiments that are executed cheaply, efficiently and seamlessly as tool calls.

Through experiments like autoresearch we have seen that AI is already, if not always smarter, at least more systematic than humans at following the scientific method relentlessly (hypothesis-experiment loop). Let's see what we can do by connecting it to the real-world :)

I initially quite liked it, particularly when the references are a bit more obscure, generally from classic scifi/fantasy. You can end with a nice sounding name and a small wink to fellow nerds.

But it is such a shame that it has started to become the brand of dark-side (militaristic/authoritarian) Silicon Valley: Palantir, Anduril, this… Tolkien would be so very sad.

PS: Palantir is at least rather fitting and honest, it’s literally an evil crystal ball (at least the one shown in the movies).

As a founder who has gone through scaling processes, I am quite impressed at Zuckerberg's ability to remain relatively competent and relevant at the top role throughout the whole journey. The skillset required changes quite a lot in each phase.

There are very few cases where the archetypical young garage-founder has stayed on top in the long-run.

Yes I think it’s a linguistic confusion more than anything else.

To me, consciousness is not generally that you can be aware (conscious) of things around you and can react to them, lots of things can do that.

Consciousness is a shorthand for saying that something is conscious of itself, or conscious of its own consciousness. It is the meta-ability to observe its own perceptions and thoughts. And a sense of self, a sense that the observer is the same over time.

But frankly, it’s a terrible concept and my definition is plenty flawed too. In practice it is more of a moving goalpost to denote the specialness and superiority of humans over all. That thing we can’t quite put a finger on that makes us different. It is a secular euphemism for the soul. It is not very scientific.

And that is quite problematic because the privileges we ascribe to those on the wrong side of the line fall off a cliff. We rely on that line as a foundation for so much of our morals. We have seen the catastrophes that happen when a group has a different idea about the line.

I don’t have a solution to propose, it’s hard.

I’m sure there will be a ton of unexpected complexities that arise only when you are trying to push the limits, like in all engineering domains. And it’s all a highly interconnected system, you cannot expect to dramatically change the water flow without impacting others aspects.

I know it is quite distant, but from my experience in large-scale data engineering, 90% of the time goes in addressing subtle issues that can only be observed hours into a job, the rest of the issues are quickly resolved earlier. I am assuming that such complexities will be so much harder in physical systems, and even more so in biological systems.

Sounds cool but for such experimentation you would want relatively fast experimental iterations to get anywhere, and this would take literal ages. You can play around with growth speed of course but that’s a different question and might be in some ways opposed to achieving height.

Well the point of the article is that water usage is much worse in the power plant capacity necessary to feed datacenters.

I feel like it’s never made clear in what way the water is used up in these cases.

It’s not like it’s consumed like fuel. And it is not absorbed like in agriculture. But I understand it is not trivially recyclable either, the heat of the water alone can be harmful if released as-is. Does cooling happen via evaporation and is that how the water is “lost”? And I am not sure if it is contaminated in other ways.

What is the actual impact or opportunity cost of using the water in datacenter (or energy plant) cooling versus other uses?

What do you mean? Did you read the article? There’s so much evidence showing that it was the opposite.

Their profits shot up 3x in 2023 and 5x in 2024. They had 70%-140% profit margins. They publicly said that the end of the flu was a risk for their profits. There’s plenty of messaging recording explicit price collusion.

How is that a natural supply shortage?

The underlying economic factor is simply that monopolies or cartels will always try to manipulate prices in their favor if they can.

The chart is meant to show how absurd the conspiracy theory is, but it turns out it’s literally what happened this time around at least. Well they didn’t forget their greed of course, they just temporarily lost the ability to exercise it.

I assume that the PowerPoints can be absolute slop because they do not matter at all, the power is elsewhere.

It all comes to names here. The world puts waaay too much faith in successful people being successful again, and too much of the success is attributed to their inherent talent. It’s a classic fallacy of authority.

Ironically, because everyone thinks that everyone thinks that they will succeed, investing on them does yield good returns, because others are likely to invest after. And they do indeed continue to succeed!

It’s as arbitrary as tulips. Bubbles of names. That’s why it’s weirdos, because it’s quite random who is chosen, and I suppose weirdos have a more pronounced brand.

Of course it’s not completely arbitrary, there’s real merit in cracking the initial chicken-and-egg, depending on where you start in life. And there are plenty of ways of squandering the opportunities that come after. But still.

I would be glad to learn if you are willing to explain, this what I found from trusted sources, but it would be great to know if there’s additional nuance.

To save others a search:

The akritai (singular akrites) is a term used in the Byzantine Empire in the 9th–11th centuries to denote the frontier soldiers guarding the Empire's eastern border, facing the Muslim states of the Middle East. (Wikipedia)

Akron means edge or border, so "frontiersman" or "those of the border".

EDIT: Commenters seem upset about the Muslim part, I didn’t mean to imply anything, you cannot just copy-paste contemporary disputes and prejudices a thousand years ago. In the historical context it’s just like most borders between different civilizations. The point is that they were a collective organization getting together to defend their land.

That sounds incredibly harsh. What kind of university was it? Is there more to the story?

Kicking someone out and effectively voiding all progress just for failing one challenging final year class twice is completely mad.

Was it rather that you lost a scholarship and couldn’t afford to finish or something like that?

Sorry to dwell, I’m sure it must be painful, I wish you the best luck.

If it is forced it is not what you would rather do.

What have the Romans ever done for us?

Such a cool article chock-full of cool facts!

Nearly every species of scorpion intensely fluoresces under UV light. […] Scorpions have photoreceptors in their tails, separate from their eyes. […] It is hypothesized that a scorpion uses this fluorescence to tell whether any bit of its body is left exposed from its hiding place. Its tail “looks” down at its body, and if it sees its own fluorescence, it knows it is exposed to light, and in danger.

And a special call-out to the “Andean Cock-on-a-Rock” :), see a photo in the article.

As far as I remember, from when I was closer to academia, in NL postdocs had to be offered a permanent contract after 3 temporary contracts, with a maximum of 1 year per temporary contract, or something like that. I believe this wasn’t exclusive to postdocs and it is a general law for most professions.

In recent years in Spain they aggressively decreased that threshold to the point where most employees need to have permanent contracts. Interestingly it has led to significant growth because, among others factors, it has increased consumer confidence, and it has been a much smaller burden on companies than expected.

Perhaps the term “permanent” contract is confusing to some. It’s not in the sense of a functionary or tenure, where you virtually have a job for life unless there are extreme circumstances. A permanent contract is an indefinite contract, one without a specific end, where firing you needs to be properly justified, but you can be fired, certainly.