dosinga

Mozilla exists because one company tried to own the front door to the web, and an open community rose up to make sure it never could.

I'd say that the front door to the web is pretty much owned by Google and Apple at this point given Firefox current marketshare. And maybe that's enough, maybe a future where a low percentage of open models keep the rest of the system honest but that doesn't seem the argument of this article

...and it's rebuked by Maynard Keynes: "We're all dead in the long run".

Keynes said that when somebody complained that his theories didn't work in the long run. And the true rebuke is, we're now in the long run and Keynes is dead

second paragraph

A model can generate a login system in seconds. It will not ask if emails must be unique. That single missing requirement can collapse the system. LLMs cannot see the category of questions that keep software safe.

really? I am assuming here that the author means that the emails used when registering across users need to be unique. even mediocre models will do that for you.

even more to the point though, if you tell an LLM, hey I want to build this system, let's talk about the architecture, it will ask you all the relevant questions. this can really help with the engineering. sure, there's a rubber ducking aspect to it, but even so. telling an LLM to immediately implement an idea does not lead to great engineering. starting to code yourself has the same problem. doing some brainstorming with an LLM (like in the olden days with humans) is where they help with the engineering part.

What's up with this glass? Excuse me? Excuse me? This is my glass? I don't think so. My glass was full! And it was a bigger glass! Who's been pinching my beer?

Does anybody else find this article reads really AI written? The way the argument is built, the emdashes, the dramatic one sentence paragraph:

"That assumption is wrong."

"The human is the equation. AI is what makes that equation run faster." -- what?

The company I work for pays for my tokens, but I find enough joy in doing weekend projects supercharged by AI to chip in myself too

Yes! It's a great addition to the argument

The counter argument (not mine): Software Engineers are willing to spend their own money on AI. The same people that wouldn't pay 10 dollars for code if there was a workaround that took hours.

Austerity does not create fascism. The examples in the article are just all wrong. Even if you call Trump and Orban fascists, they did not follow austerity. The reaction to the financial crisis was the largest fiscal and monetary interventions in history. Orban consolidated power by spending and state expansion, not austerity. True austerity in Greece, Spain, Portugal of recent times led to new parties, but those countries are politically doing surprisingly well.

I ended up using https://eu.switch-bot.com/products/switchbot-bot -- just have a finger robot push the button

Employers and landlords do that sort of thing all the time. Rent goes up, job descriptions change, return to office is suddenly required. And yeah, you can get a different job or a different home if you don't like it.

if you exclude the enslaved, the south had a higher GDP per capita than the north.

In other words, if you remove the people that earned the least (close to nothing) the overall income per capita goes up? If you exclude the non nobles I am sure the middle ages had a very high GDP too

you can run https://block.github.io/goose/ in headless mode (I work on goose)

ah, but you can always just ask the LLM questions about how it works. it's much easier to understand complex code these days than before. and also much easier to not take the time to do it and just race to the next feature

I have looked but I can't find out if it actually means something. Does 89 seconds before midnight mean we have a 50% chance to survive the next N years somehow?

this is cool. not sure it is the first claude code style coding agent that runs against Ollama models though. goose, opencode and others have been able to do that a while no?

Who hasn’t turned in a paper on a broken floppy disk, with the excuse ready that the floppy must have broken when the teacher asks a few days later?

I feel seen

One thing that seems missing from a lot of these comparisons is the base rate of success for dieting itself.

Most people who “start a diet” never meaningfully lose weight in the first place, or lose a small amount and plateau quickly. The cohort of “dieters who regain weight” is already heavily filtered toward the minority who were unusually successful at dieting to begin with. That selection bias matters a lot when you then compare regain rates.

GLP-1s change that denominator. A much larger fraction of people who start the intervention actually lose substantial weight. So even if regain after stopping is faster conditional on having lost weight, the overall success rate (people who lose and keep off a clinically meaningful amount) may still be higher than dieting alone.

In other words: “people who regain weight after stopping GLP-1s” vs “people who regain weight after dieting” ignores the much larger group of dieters who never lost anything to regain. From a population perspective, that’s a pretty important omission.

Google-as-the-new-Microsoft feels about right. Windows 1 was a curiosity, 2 was “ok”, and 3.x is where it started to really win. Same story with IE: early versions were a joke, then it became “good enough” + distribution did the rest.

Gemini 3 feels like Google’s “Windows 3 / IE4 moment”: not necessarily everyone’s favorite yet, but finally solid enough that the default placement starts to matter.

If you are the incumbent you don't need to be all that much better. Just good enough and you win by default. We'll all end up with Gemini 6 (IE 6, Windows XP) and then we'll have something to complain about.

eh, maybe. Sure Google bought Youtube, but the whole making it social came later. Apple spending a lot of time refining them is exactly the point. They did go up in the stack (given that they started out as home computer builder quite a bit). Word first came out for MS DOS, so definitely going up in the stack.

This doesn’t sound very convincing, mostly because the examples don’t really line up with the claim. Apple supposedly struggles “up the stack,” yet many of the best and most-used iPhone apps are built by Apple itself. Google is held up as failing at social, but YouTube is arguably the largest social network in the world. Oracle is described as struggling in apps, yet it’s clearly doing just fine as a massive, profitable enterprise software company. And the IBM example is backwards: IBM didn’t accidentally hand Microsoft the OS layer, it already had its own operating systems. In fact, Microsoft is the clearest counterexample here, it got big by owning the OS and then very successfully moved up the stack to dominate applications with Office.

I don't know. The World Happiness Report relies on one simple question, which is easy to criticise but at least it applies a clear and consistent method. The paper referred to does not. It uses a special US dataset for states and a much smaller global dataset for every other country, then treats the results as if they measure the same thing. This setup almost guarantees that US states look unusually good. The authors present this as evidence, but it mostly reflects differences in survey design rather than real differences in wellbeing. In that sense the methodological problems here are more serious than the ones they point to in the World Happiness Report.

If there are no free alternatives, yes. 100 USD a month for ChatGPT seems great value

Also: Recency & survival bias. We remember what became the classics from the 90s and also what came out this year and was available on our last flight

At first sight I read Minimal Viable Army - would be a good read for an aspiring evil mastermind

900k is a large plane flying from Dallas to Sydney and back

This advise against quitting you find everywhere is just wrong. Sure you should give it a fair shake, but if you are on a dead end, never quitting means never winning. If something doesn't work, it's possible you should just stop doing it and try something else.

Ideally you have a document in place saying this is how we handle vibe coding, something like: if you have the AI write the first version, it is your responsibility to make it reviewable.

The you can say (and this is hard), this looks like it is vibe code and misses that first human pass we want to see in these situations (link), please review and afterwards feel free to (re)submit.

In my experience they'll go away. Or they come back with something that isn't cleaned up and you point out just one thing. Or sometimes! they actually come back with the right thing.

I mean, sure we're in a bubble, but the trick is to call it, with that old Keynes quote about the market staying irrational longer than you can stay liquid.

But also: > "So, you're saying a third of the stock market is tied up in seven AI companies that have no way to become profitable and that this is a bubble that's going to burst and take the whole economy with it?"

I said, "Yes, that's right."

that is something different in this case isn't it? those seven companies making up a third of the market do not need to become profitable, they are insanely profitable. Mostly they invest a lot in AI but if that doesn't pay out, all but NVidia have their day job to go back to.

"In German, oder Englisch, I know how to count down Und I'm learning Chinese!" - as relevant as ever. Maybe less so German and more matrix multiplication