alganet

52 ms·

811.
by alganet 1y ago
> a thing governed by a set of mathematical laws Again, you're mistaking the thing for the tool we use to describe the thing. > aside from precision limitations It's not only about precision. There are things we don't k
812.
by alganet 1y ago
We currently can't simulate the universe. Not only in capability, but also knowledge. For example, we don't know where or when life started. Can't "simulate forward" from an event we don't understand. Also, a s
813.
by alganet 1y ago
I'm not talking about soul. I'm just saying you're mistaking the thing for the the tool we use to describe the thing. I'm also not talking about simulations. Epistemologically, I'm talking about unknown unknowns. Th
814.
by alganet 1y ago
You're mistaking the thing for the tool we use to describe the thing. Physics gives us a way to answer questions about nature, but it is not nature itself. It is also, so far (and probably forever), incomplete. Math doesn't need t
815.
by alganet 1y ago
What makes you think that human intelligence is based on mathematics?
816.
by alganet 1y ago
Let's consider the other scenario: AI plateus, bottlenecked by the lack of quality training data. It becomes obvious that it can only improve as humans improve. For a while, AI companies trick humans that can come up with original thou
817.
by alganet 1y ago
I'm starting to rate AI rethoric by a new metric: how much it relies on the general audience imagination. This article features 8 "imagines" that are vague and feature common misconceptions about reality. These misconceptions
818.
by alganet 1y ago
Maybe AI is not nuclear power, or the printing press, but the optical mirror. A simple, unordinary, boring, mirror. Seemingly fantastical to those unfamiliar with it, it would have startled many who never saw one, with its apparent capacity
819.
by alganet 1y ago
That doesn't solve the problem. The person you are talking to face-to-face could have been targeted with disinformation as well. This is suggested in the paper: manufacture of narrative across communities. Those communities are not exc
820.
by alganet 1y ago
There's plenty we can do before any attribution is made. LLMs hallucinate. They're weak and we can induce that behavior. We don't do it because of peer pressure. Anyone doing it would sound insane. It's like a depth char
821.
by alganet 1y ago
If I understand what you're looking for, GTK. If it's a linux desktop, there's a super high chance that GTK is both installed and familiar to the user. However... > This would make vendors write more gui software for Linux
822.
by alganet 1y ago
Companies are swiftly moving from an "intelligence" narrative to an "augmentation" narrative (I think it's also doomed). I think the linguists already won, but the AI tale is so tall it hasn't had the time to h
823.
by alganet 1y ago
> engineering only has to deal with the complexity of the computer which is at least deterministic Engineers outside business structures can scramble themselves and produce value. It's messy and inefficient, but doable. Sometimes, i
824.
by alganet 1y ago
> request hangs, or fails completely, or gives a different response I try to avoid those, not celebrate them.
825.
by alganet 1y ago
https://en.wikipedia.org/wiki/Progressive_enhancement This. But it's hard. But it's worth it. Best of both worlds.
826.
by alganet 1y ago
Must... make... cathedral... at... all... costs... why... no... cathedral... halp... http://www.catb.org/~esr/writings/cathedral-bazaar/cathedral...
827.
by alganet 1y ago
I think AI companies overdid it, and they're struggling to get high-value collaborators to their ecosystem. Before, Google would launch a new Android version. Or Facebook would launch React, and developers would flock en masse to the e
828.
by alganet 1y ago
> If you’re still writing everything by hand, ask yourself why. Let me draw a comparison to the real world. Almost everything we can buy can be built in an automated production line. Some things, cannot. Some things we cannot even buy. W
829.
by alganet 1y ago
> Are they focusing on beginners? That's what their PR material seems to imply. > Have you tried building things with this stuff? Yes. > it's exciting, there's all kinds of new things to build Yes, and no. I don&#x27
830.
by alganet 1y ago
Ideally, you want more than one bootstrapped platform. Platforms eventually die, and you don't want to rely on an emulator for bootstrapping. Some time from now x86_64 will fade away, and there's a large chance rust will still be
831.
by alganet 1y ago
> Karpathy [...] I'm sure he's a great guy, but let's face it: He's there speaking those things because he made some popular videos in the past and they needed to renew their PR strategy. Speaking of PR strategy... &g
832.
by alganet 1y ago
I stand by it. You can definitely take a base LLM model then train it on existing, prepared root case analysis data. But that's very hard, expensive, and might not work, leaving the model brittle. Also, that's not what an "AI
833.
by alganet 1y ago
It's an extra tool, for sure. There is no silver bullet. No one likes huge inheritance chains, no one likes only tiny lego pieces, no one likes callback hell.
834.
by alganet 1y ago
I didn't. This also confused me: > LLMs being pattern matching machines LLMs are _not_ pattern matching. I'm not being pedantic. It is really hard and unreliable to approach them with a pattern matching mindset.
835.
by alganet 1y ago
> Use composition instead. Inheritance was a hack in the first place. Composition is great. It's the preferred way. However, if you take it to the extreme (composition only, never inheritance) you sometimes end up with really weird
836.
by alganet 1y ago
Fuck this. All engineers are normal engineers. Everyone can be an engineer.
837.
by alganet 1y ago
> we wrote an LLM Excuse me, what?
838.
by alganet 1y ago
I can see it just fine, it just doesn't appeal to me. I want to build things, not "launch startups". If someone asks me about a minimal thing of what I build, I want to be able to speak for hours about it, in detail. I want t
839.
by alganet 1y ago
I think you're wrong, and copyright law is the unexpected ally of the people in this matter. Big studios already got their death sentence. Smaller independent productions, even individuals, will get the larger slice of the pie eventual
840.
by alganet 1y ago
> Clickbait trash. Is this really the quality of a normal HN post? Hi, it's 2025. The whole internet has been like this for more than a decade now, and it's getting worse by the year!

More ›