dimal

I had been worried about what would happen as the American republic goes down, whether other countries would follow suit. So those two examples are cause for hope.

As for the second term, we can’t forget that the Democrats basically threw the election. This election was a layup. An easy win. And they did absolutely everything wrong. I almost blame them as much as Trump for this mess.

There’s always been this draw in software engineering to find the silver bullet that will allow you to turn off your brain and just vibe your way to a solution. It might be OOP or TDD or pair programming or BDD or any number of other “best practices”. This is just an unusual situation where someone really can turn off their brain and get a solution that compiles and solves the problem, and so for the type of person that doesn’t want to think, it feels like they found what they’re looking for. But there’s still no silver bullet for complexity. I guess there’s nothing to do but reject the PR and say “Explain this code to me, then I’ll review it.”

I’m autistic and I agree with this. Pretty much any time someone says, “autism is…” they’re wrong. So, saying this disorder is the opposite of a poorly defined cluster of traits is just absurd.

That’s the opposite of an incorrect stereotype. The idea that autistic people don’t have empathy is false. I’m autistic. Of all the other autistic people I know (about a dozen) all are highly empathetic. To neurotypical people we often don’t appear empathetic but that’s very different.

This disorder actually sounds somewhat adjacent to autism. Many autistic people have intense affective empathy and get overwhelmed by other’s emotions, especially in groups. But often we just shut down, which appears like nothing is registering and we have no empathy.

I sympathize with this point of view but I disagree. I’m not sure if I’m ADHD, but I’m autistic. I’ve been diagnosed as ADHD in the past, I’m just not sure how useful it is as a diagnosis for me. I have enormous problems with focus and executive functioning. For large portions of my life I’ve been seriously disabled, unable to work or support myself. There are times where I’m unable to talk. And I still don’t see autism as inherently dysfunctional.

If you put me in a village in Europe 5000 years ago, I’d be fine. I’d be better than fine. I’d be the guy in the hunting part who could smell the fresh scat from 50 yards away. I’d be the guy who could remember all the fucking barks and plants and mushrooms that are good for what ails ya. I’d be the guy who knew the story of every god and goddess and why they’re important. Most social situations would involve people I knew very well or people in the same culture, where I could depend on knowing the rules of the culture.

The modern world is full of random noise and stifling bureaucracy. I love being autistic. But it’s awful, truly awful to have this nervous system in this society. The endless stress breaks you down day after day, year after year, and system teaches you to see yourself as inherently broken, when it’s the system that has broken you.

Maybe you’re disabled, but maybe it’s the system that did it to you.

Hmm. That hasn’t been my experience. Getting adderall or Ritalin was as easy as saying, “I’m having trouble concentrating.”

We shouldn’t call it a health care system. That’s an aspirational term that doesn’t match reality. We have a medical industry.

Like all industries, the goal is to return value to shareholders. If healthcare is provided as a side effect of profit maximization, that’s nice to have, but that’s not the purpose of the system.

For example, UnitedHealth is the number three company in the US by revenue. Only Apple and Amazon make more. Their entire business model is to collect money from people, then not give it back when they need health care.

The system isn’t broken. It’s working exactly as intended.

Seriously. I have never seen this, even once. I had been wondering if it was impossible. If a model can really say “I don’t know” when it doesn’t know, that could change everything. How many pointless, dumb rabbit holes could be avoided?

Even with large contexts there's diminishing returns. Just having the ability to stuff more tokens in context doesn't mean the model can effectively use it. As far as I can tell, they always reach a point in which more information makes things worse.

“Trust the science” is a terrible slogan. It almost turns science into a religion. Most people that use it seem to think that science is whatever a scientist says. We should be saying “Trust the scientific method”.

With the dynamics of publish or perish being what they are, what’s the way out? As long as there is high demand for papers (not knowledge) then some market will pop up to feed that demand.

I hate to say it, but cutting off the money spigot of government funding for papers seems like a good start.

It feels like our society has been optimized to game a few metrics like this (government wants to raise GDP, CEOs want to increase shareholder value, university deans want to increase funding to write papers), and all of them have toxic second order effects that make society worse.

Setting up linting with noExplicitAny is essential. But that won’t stop them from disabling it when they can’t figure something out. They’re sneaky little bastards.

Technically you’re not vibe coding. You’re using AI to do software engineering. Vibe coding is specifically the process of having AI produce code and plowing ahead without understanding it.

I know I’m being pedantic, but people mean very different things when they talk about this stuff, and I don’t think any credence should be given to vibe coding.

They need to fix bugs first. For fucks sake, predictive text on the iOS keyboard regularly predicts non-grammatical words, and dictation is terrible. It’s ridiculous that AI is good enough to write sonnets and coherent code, yet Apple can’t even do autocomplete. Whisper has been available for forever, yet it’s still painful to enter text on iOS without a third party app.

Mozilla is over. They put all their eggs in the Google basket, and soon they’ll lose that. They have no viable path forward.

Servo and Ladybird are the future. I’m astounded by how quickly Ladybird is proceeding, with far fewer people than Mozilla. It’s been inspiring to see what that project is doing.

Hmm. I am autistic, so as far as humans go, I'm robot-adjacent.

As I get older, I can see a future where I’m cut off from parts of the web because of captchas. This one, where you just have to click a button, is passable, but I’ve had some of the puzzle ones force me to answer up to ten questions before I got through. I don’t know if it was a glitch or if I was getting the answers wrong. But it was really frustrating and if that continues, at some point I’ll just say fuck it and give up.

I have to guess that there are people in this boat right now, being disabled by these things.

But the future is vibe coding and shipping fast even if we don’t understand the code the LLM wrote!! The only speed metric worth caring about is time to market! Just fucking ship! /s

Have you seen the tests that LLMs write if you just vibe code it and don't pay attention? They're generally worse than worthless. They test pointless stuff and they miss important stuff. If you have them on a tight leash, then you can get some useful tests, but then that's not vibe coding.

> But watch out, because sometimes the LLMs try to delete tests that are hard to pass. If vibe-coding means reading no code, not even the tests, then, yeah, you're gonna have a bad time. But if it means "only reading the tests and not reading the application code," then vibe coding works pretty well, actually.

Yeah, vibe coding means plowing forward with whatever the LLM does without understanding it. That's the definition. If you're using your brain, you're not vibe coding.

> developers fall into two camps:

> Craft-focused developers enjoy the process of writing code…

> Delivery-focused developers care about shipping products.

What the hell am I, then? I’m a craft-focused developer who cares about shipping products, and I like AI.

I find that AI doesn’t reduce my enjoyment of the craft. It reduces my need to type code with my fingers. I’m able to spend more time thinking above the code and put more time into getting a good design instead of getting something that just works. I’m not doing 100% AI generated code, though. I still find that there are lots of fiddly refactors that AI is too clumsy for.

I get the sense that most programmers don’t like to refactor. They enjoy the process of getting something working, but don’t like the process of making it good. To me, refactoring is the craft. That’s the part that I love. AI gives me more time for that.

Weird that it doesn’t mention Trump. He and Bezos were at odds until Bezos gave him $1M for the inauguration and decided to neuter the Post’s opinion section. Then all threats of antitrust against Amazon magically went away. Bezos probably had to trash the Post to save Amazon. The loss is a drop in the bucket to him.

Vibe coding pushes errors rightward, but using AI to speed up typing or summarizing documentation doesn’t. Vibe coding will fail, but that doesn’t mean using AI to code will fail. You’re looking at one (admittedly stupid) use case and generalizing too hastily.

If I have an LLM fix a bug where it gets the feedback from the type checker, linter and tests in realtime, no errors were pushed rightward.

It’s not a free lunch though. I still have to refactor afterwards or else I’ll be adding tech debt. To do that, I need to have an accurate mental model of the problem. I think this is where most people will go wrong. Most people have a mindset of “if it compiles and works, it ships.” This will lead to a tangled mess.

Basically, if people treat AI as a silver bullet for dealing with complexity, they’re going to have a bad time. There still is no silver bullet.

Did you actually watch it? He talks A LOT more about Simula and C++ than Smalltalk. He goes back to the original sources: Kristen Nygaard and Bjarne Stroustrup. Seems odd to focus on Smalltalk when that’s not what the talk was about.

I still do a lot of refactoring by hand. With vim bindings it’s often quicker than trying to explain to a clumsy LLM how to do it.

For me, refactoring is really the essence of coding. Getting the initial version of a solution that barely works —- that’s necessary but less interesting to me. What’s interesting is the process of shaping that v1 into something that’s elegant and fits into the existing architecture. Sanding down the rough edges, reducing misfit, etc. It’s often too nitpicky for an LLM to get right.

That is completely bizarre. I’ve been wondering what’s happening to programmer interviews now that AI makes all the standard formats pointless. I never imagined that they would ADD coding to the process for other roles. Having PMs vibe coding in an interview? That’s idiotic.

Interesting how supposed fraud from lower class people is a high priority that must be punished, but fraud from upper class people is almost always protected by the corporate veil.

You chose oil for your example, but what about natural gas? If Europe needs natural gas, they can just buy it… and give money directly to their enemy, Russia. Just buying what you need isn’t without second order effects. The second order effects of solar and energy diversification are more palatable than directly funding an enemy.

“Look at Japan”. Ok, let’s look. They attacked the US in 1941 because of the US oil embargo. Their current situation is predicated on the US continuing to be the world’s policeman, ensuring that shipments get from point A to B. There will come a time when that assumption will not hold.

Things change.

All that money wasted is the whole point of Mozilla. It’s been captured by Google. The role of Mozilla is to provide a fig leaf of “competition” while using Google money to pay executives who fulfill that duty. Anything that is successful or innovative, like Rust, Servo, or FakeSpot need to be jettisoned because they are a distraction from Mozilla’s true mission.

Won’t make much of a difference? To what?You’re only talking about whether to offshore or not. Not whether to HIRE or not.

Many companies simply won’t offshore core functions because doing product development on your core product with a team in a different time zone or from a very different culture often doesn’t work. But this will matter to companies that have laid off US engineers or avoided hiring and now won’t have that extra tax burden.

This is a shot in the dark, but there is a lot of excitement right now about treating schizophrenia and bipolar disorder as a metabolic issue. Some people are seeing complete remission, without medication. This isn’t a quick easy recommendation to just go off meds and go keto, but for some people that apparently has worked. Chris Palmer M.D.[0] is one of the main people driving it.

[0] https://www.chrispalmermd.com/chris-palmer/