schmudde

In the Berry style, I can say that I don't need to use AI because my wife can do all the statistical inference I need.

> OpenThropic IPOs, vocal shareholders from Ferenginar quite insistently ask where they money at, token prices go up singificantly: Only a problem if you’re dependent on AI.

Exactly. People are going to be surprised when they can't afford to maintain their own code.

I actually agree with Collison - although I have to admit I never thought about Brothers K so deeply. It never engendered that response from me.

The Idiot is my favorite Dostoyevsky novel. The Prince's good nature were also his flaws. This leads to so many profound scenes and scenarios. And it gives greater texture to Christian philosophy and the image of Christ than any other analysis I have ever read.

"The House of the Dead" changed the way I looked at prisons, "Crime and Punishment" about ego and justice, "Notes From Underground" about the lone genius - I can't say that Brothers K. gave me any of this.

> Closer to reality, I can’t experiment with cooking without understanding the underlying chemistry. Generally speaking, I can’t use something I don’t understand completely – if it doesn’t click, I can’t work with it.

But there is always more detail to learn?

The article only mentions tuition once - which is the original sin of US universities.

I have generally found that US Americans don't really understand the incredible value that these institutions bring to the overall culture and economy. Thus the broader population has a will to critique but no will to improve the situation.

The pace at which sports betting has changed in the USA is incredible. I grew up with Pete Rose not getting into the baseball hall of fame. Not it's impossible to imagine that athletes are not betting on games. And more importantly, there is the social cost:

"One UCLA study published earlier this year found that in states that legalized sports gambling online, bankruptcies and credit card delinquencies increased about 25%. A survey by the National Council on Problem Gambling found 8% of American adults, almost 20 million people, reported experiencing at least one indicator of problematic gambling behavior "many times" in the past year. "

"The Annotated Turing" - I didn't know you could meaningfully intertwine biography and technical analysis like that.

I've always had more projects than I do time. And it's easy for me to be selective about where I might use an LLM. So no changes here.

Been getting into Rudy Rucker, who was funded by the AutoCAD folks to do cellular automata research: https://www.fourmilab.ch/cellab/.

Autodesk funded some really interesting projects.

Trevor Paglen had a similar project for US military SIGINT. It's interesting how much more "serious" the Russian insignias are than the US patches. I have a few pictures here: https://schmud.de/books/i-could-tell-you-but-then.html

The web browser app paradigm changed everything. The Windows API isn't nearly as important as it once was.

I don't understand why this would be different than the Microsoft Office/LibreOffice situation. Adoption has been really slow in Europe (though it is happening - to the credit of European bureaucracies) even though the software packages essentially have feature parity.

This is also what landed with me.

The issues we experiment with are usually quite broad - improvements to logging, routing - but the implementation is always more narrow. Still, it seems like there is still more room for more refinement and focus now that you mention it.

Yeah - we're doing something like one-on-ones that drive these explorations. Maybe we need to make the "20%" more formal.

The real missing part is discussed in the "knowledge building" link. I'm on board with Peter Naur's idea of building software, but it's very difficult to share individual knowledge with a team as it is being formed.

For example, reading someone's implementation as a work in progress is a different than reading an implementation that is a complete idea.

A work in progress is a lot easier to read with a design document in hand, but that design document that is full of assumptions that might be better explored in short feedback loops on production.

I haven't enjoyed the ecosystem of text editors on the web as much as I have local apps. HackMD was really hitting the sweet spot of simplicity and shareability.

I'll give it a shot. At first glance it seems like it has the simplicity of a text editor and all the sophistication has gone into collaboration. Kinda what web apps should be about.

Ah nice tip. Will put it on the reading list!

Sharing this seems like an appropriate way of commemorating David Kahn's passing (https://news.ycombinator.com/item?id=39233855). <3

It's astonishing (and sad) how few opportunities exist for basic research.

Before folks jump on the business model aspect, one should keep in mind that many (all?) online communities go through a similar growth curve. Wikipedia included.

But this article goes into the details of what makes this feel different. Quora is aesthetically much worse than it ever was.

⃞ I have a salt deficiency.

The moderation is quite arbitrary. Here are three specific examples I've noted:

1. "Don't hate Jira, hate your manager." https://lobste.rs/s/n4v6a8/you_don_t_hate_jira_you_hate_your... is okay but "Making time for planning" https://lobste.rs/messages/h9z5ee is not (Management is off-topic.)

2. "A normal week (in tech)" https://blog.ignaciobrasca.com/work/2023/05/01/a-normal-week... is not okay (Article does not relate to computing.) but "What are you working on this week?" is okay.

3. "What Punch Cards Teach Us About AI Risk" https://bcantrill.dtrace.org/2023/11/26/what-punch-cards-tea... is not okay (Business history is off-topic) even though specific implementations of punch card machines are discussed

There will always be sharp edges to moderation but I've generally found more permissive policies to be more fair.

Prefect combination with Article 45 (https://www.eff.org/deeplinks/2023/11/article-45-will-roll-b...). You can roam around EU versions of websites that rely on government own certificate authorities and allow you to log in with your eIDAS id - which will be a requirement for all "very large online platforms."

The EU Parliament giveth and taketh away.

You might try Brave New Europe? They don't look at tech specifically like EFF, but tech touches their stories on economics, regulation, and the media: https://braveneweurope.com/?s=tech

The context of China destroying Tibet should not be overlooked here. Nepal shares much with Tibet culturally. Perhaps China isn't an immediate threat to their sovereignty, but it's certainly an existential threat.

It's nice to see, but it's backing Abelton into feature parity with Cubase on the Atari ST in 1989.

The data has many different ways to tell the story. For example, 50 years of fires: https://observablehq.com/@ktrudeau/2023-fire-weather-days

> As always, this will be made open source and undergo continuous improvement.

Wait - the recommender algorithms will be made open source? I find that hard to believe.

For just $4 more you can buy the Atari ST version!

I really dig your portrait.