vdupras

This is what I mean by "organizational problem". I understand that sometimes -- in fact most of the times -- you need to onboard mediocre programmers. But on the other hand, it's really great when you don't have to.

I don't know, I'm seduced by the elitist approach: code with a high cognitive load keeps mediocre developers away.

Case in point: Forth. It generally has a heavy cognitive load. However, Forth also enables a radical kind of simplicity. You need to be able to handle the load to access it.

The mind can train to a high cognitive load. It's a nice "muscle" to train.

Should we care about cognitive load? Absolutely. It's a finite budget. But I also think that there are legitimate reasons to accept a high cognitive load for a piece of code.

One might ask "what if you need to onboard mediocre developers into your project?". Hum, yeah, sure. In that case, this article is correct. But being forced to onboard mediocre developers highlights an organizational problem.

If I try to sum up all the irrational rules I've seen in my professional life, sure, I've seen some coming from unions, but my general impression is that most of the "wtf" I've seen come from management, or, in the case of the public sector, the political elite. "most" is probably an understatement.

I guess you could settle the question by trying it!

The amount of offline documentation required to do so is gargantuan. Try any kind of "repurposing" of any phone -- go for something trivial, not as hard as controlling an automated greenhouse circuit -- and try to do so without the internet -- or let's take the difficulty down a notch, without AI or search engines, wikipedia allowed. The operating system on that phone is likely way too complicated for you to succeed. It's also likely to be locked.

From https://nx.dev/:

> 2.5 million developers use Nx every day

> Over 70% of Fortune 500 companies use Nx to ship their products

To quote Fargo: Whoa, daddy...

Now that's what I call a rapidly degrading situation we weren't ready for. The second order fallout from this is going to be huge!

Some people are going to be pretty glad they steered clear of AI stuff.

It's alright, I'm not French :)

But yeah, I have to be colored surprised. I was certain I was replying to arrogance, rather than (possibly misplaced) admiration. I guess that puts me on the wrong side of the moderation fence. Sorry.

I have trouble understanding how that guideline applies here. The original article shows how it's possible that we're about to see an AI bubble pop, the parent comment show generic american arrogance[1], and I come up with a historical example of such a mix of hubris and arrogance.

If my comment can be characterized as flamebait, it has to be to a lesser degree than the parent, right?

And I'm not even claiming that the situation applies. If you take the strongest plausible interpretation of my comment, it says that if indeed this whole AI bubble is hubris, if indeed there's a huge fallout, then the leaders of this merry adventure, right now, must feel like Napoleon entering Moscow.

But well, anyways, cheers dang, it's a tough job.

[1]: the strongest possible interpretation of "This is how America ends up being ahead of the rest of world with every new technology breakthrough" is arrogance, right?

Certainly, and they lost the other half on the way home. On the way to Moscow, there was still something to look forward to. In Moscow, with no supply, dying like flies, all they had to look forward to was a long march through the Almighty Intense Winter, during which they'd be dying like flies too.

[flagged]

> And the prices of these new and used are out of this world even compared to baseline inflation.

How many such cases do we need to have to be able to put out the hypothesis that the baseline inflation figures we're being fed are wrong?

Common accepted knowledge around here is that greed is good and that capitalism works because of it.

Fine, let's accept it.

But here's my pet theory and here why I think we see cracks forming up in this "greed is good" mantra (which, frankly, hurts common sense of anyone living in a judeo-christian culture, because, well, isn't greed a sin?): this only works in a world of expanding EROEI (Energy Return On Energy Investment).

This cheap energy is basically a giganormous pool of energy slaves. As long as there are slaves to capture, greed is good because it's good for everyone to capture and distribute as many slaves as possible.

When that pool dwindles, greed stops working as well as it did for the advancement of public good and it begins cannibalizing on it (the public good). It still gets accepted as a truism today, but it will soon go back to sin territory.

Of course, it goes without saying. But on the other side, not many people would have what it take to not let that kind of wealth go to their head. This désinvolture has to be praised.

You mean that in the sense "is it too soon to say we'll have a 10% inflation year?". Certainly, and I don't mean that. I just want to counter the general "annualized inflation figure numbness". But this is not noise. Prices are a real signal.

News outlets have been feeding us with annualized rates, which is making this 0.9 figure sound pale, but it's pretty big because it's for July only. Times 12, it means inflation over 10%. The 3.3% annualized figure in the article is because they've been averaging the past 12 months.

Something in July made the prices jack up in a really serious way.

My pet theory, if we indulge in a bit of conspiracy theory about the oligarchs, is that they're in fact benevolent and that the final outcome with be a net good for humanity in relation to the climate challenge.

I mean, Trump is so blatantly destroying american influence that it would be hilarious if it wasn't so tragic. When you think about the alternative, that is, a "regular" american elite saying that climate action is important with their regular hypocrisy, pretending to be doing anything at all, but in fact going the other way, then things don't look so bad in the medium term. Sure, a bit of short term pain, but otherwise, might end up being better off.

I mean, try to put yourself in the shoes of one of those oligarchs a few years ago and ask yourself how you'd actually solve the climate problem, given current cultures around the world. Maybe you'd come up with the "let's prop up Trump".

That being said, it's just a pet theory. I actually have doubts that such smart oligarchs exist.

A few new facts we'll see pop up:

1. There has always been town-destroying wildfires every summer. It's always been like that.

2. It has always been the case that AC units were necessary for a human to survive an american summer. I mean, look at native americans, they always had AC units!

3. Los Angeles always flash flooded 10 feet up across every summer.

4. America has always been at war with Canada.

5. Wildfire smoke across the sky for the whole damn summer? Always been thus. What do you mean, blue?

Pacino? De Niro maybe?

> Brazil also imagines a hyper-efficient future ...

I don't think that's how I'd call it. The future in Brazil looks deeply dysfunctional to me. About as efficient as a car with no wheel revving up to its maximum capacity. It's just that everyone is extremely busy.

Maybe that a better way to say this would be that the society as a whole has a fixation on a certain idea of efficiency?

Let them eat cake, right?

https://www.ers.usda.gov/topics/food-nutrition-assistance/fo...

Opportunity also is a precious resource. Redirecting resume to /dev/null is wasting it. I have a hard time believing that LLMs, with all their sophistication, aren't ideally suited for this task.

Aren't they an AI company? Couldn't they sort it out? If Anthropic, of all companies, can't sort out incoming job applications, what exactly are their tools for?

Maybe it's a documentation problem? It seems to me that for a piece of software to be considered good, one has to be able to grok how it works internally without having written it.

> Everyone keeps saying it these days: treat your AI like a (brilliant) new junior dev.

Really? You mean that person that drains senior developer time, but in which we still invest time because we're confident they're going to turn out great after the onboarding?

Until then, they're a net time sink, aren't they? Unless the projects you've got to deliver are fantastically boring.

So is this really what it's all about? Perpetual onboarding?

It's because programing thus is akin to Asimov's tech-men, vibing with the fabric of the universe through their religious incantations.

I'm surprised that this isn't a bigger story. Geo-strategic implications of this gradual escalation seem significant to me. And do they really don't bother pretending to do it for ransom already? Seems important to me.

That's an interesting point of view. If that is actually the goal of Israel right now, what if the Hamas executed all hostages on the spot, right now? Would Israel say "we have no more military objective, time to go home" or would they leash out in retaliation?

I think we know what the answer would be. Because we know that, how can what you say possibly be true?

I have no idea, it hasn't been me for years.

What a blast from the past, I created that library, what more than a decade ago? How simpler the world was back then. This was used by nobody except us for our little shitty use case. How noisy this project has become!

It has to be something about familiarity, because for me it's the opposite. I find that nested parens in Lisp make the code pretty hard to read. It's difficult to pinpoint where the processing begins. I also find myself constantly scanning left and right for parens boundaries. It's exhausting.

In Forth, you read your word from left to right, "playing" stack effect in your head, and you always know where you're at processing wise.

Pretty simple, but I guess one must have inclinations towards thinking about the generated native code underneath rather than thinking in terms of symbols.