Recent submissions

Comment · 28 days ago
LeCun's argument wasn't about the definition of learning though. He stated that they would never get these common sense things correct because they weren't sufficiently part of the training data. A statement that we can hopefully all agree has been thoroughly refuted.
Comment · 1 month ago
Working on https://increader.com an incremental reading platform that combines read it later functionality with annotations and spaced repetition. Supports web articles, epubs, pdfs and rss feeds with a lot of little things to remove distractions and keep you focused.

I built it because I wanted to see if there's merit to the idea of incremental reading[1] and it worked so well that it turned into the main way I read things now.

[1] (https://en.wikipedia.org/wiki/Incremental_reading)

Comment · 3 months ago
> unless it uses a harness

Sorry for skipping over your actual argument, but if it hinges on that assumption then it's probably moot. I'd assume almost all ai generated code that makes it into codebases is produced using a harness.

Comment · 3 months ago
Yup, I do this regularly now. Whenever I think that something takes longer than it should I just tell it to set up a testing harness to measure each part of the process.

Once I have the baseline I'll just set a goal to improve performance by 10x. Works better than it has any right to. Have to be careful to exclude changes that massively increase complexity for tiny gains after its done, but the big improvements are often sensible choices that I'd also make as an engineer (i.e. more efficient data representation, caching and memoization where it matters, parallel processing, etc.)

Comment · 3 months ago
Can you elaborate? I don't know much about Yu Deng or lesbian fanfiction, but it's not a crossover I would have expected.
Comment · 3 months ago
Both benchmarks are made by the same people.
Comment · 4 months ago
Surely there's something missing from your argument. It shouldn't be slowed down because it's growing exponentially?

Presumably you think that the end result of extreme and rapid ai growth is beneficial to most and that is why it shouldn't be slowed down? That arriving earlier at whatever end-point you have in mind will provide so much benefit that it's worth disregarding the pains to get there?

Or is there something else to your argument? Because if there isn't, you are staking an awful lot on your expectation coming true. Especially that going slower doesn't provide any worthwhile benefits to the outcome.

Comment · 4 months ago
Your comment implies that if Apple Configurator was available not just for MacOS that it would see a many more people using it to manage their iPhones.

To me this seems so far down the list of reasons more people don’t use it that I’m wondering whether you really believe that.

Comment · 4 months ago
Oh no, is this why my laces hold tight all day on some days and untie themselves regularly on other days?

Guess I have some experimenting to do.

Comment · 5 months ago
Yup. Basically the same story here.

There’s lots of stuff not to like about macOS, but there’s something about it that makes it click much more for average people.

For my partner in particular I noticed that they use way more features than they ever did with windows. Boring stuff like spotlight search to find files, space bar to preview documents, airdrop to send stuff to an iPhone, etc.

With windows I got called over constantly for questions like “how do I find x, I forgot where I put it”, “can you help me get this on my phone, I want to send it to <friend>”.

It’s not like good solutions for this don’t exist elsewhere, but something about MacBooks make people better at discovering those features.

I have secret pet theory that part of it is actually just that people know about Apple stuff that it’s supposed to be intuitive and so they have more confidence in trying to figure it out, which makes them have more success at it and that turns into a positive cycle.

Comment · 5 months ago
I thought so too, but I went in with the mentality that most things I just quickly look up are probably not all that important and nothing of value would be lost if I couldn’t do that anymore. The few things for which that isn’t true can probably wait until I am sitting in front of a computer.

That part honestly worked out pretty great. The first few days were excruciatingly boring, but I quickly adjusted and learned to spend more time with my thoughts.

I ended up reinstalling a browser because there were too many establishments that expected me to have a phone with a browser.

With mdm you can really control the phone top to bottom. Whitelist domains, global http proxy, allowed Wi-Fi connections, fully disable cameras, airdrop, the list goes on. Most of it isn’t super interesting to manage doomscrolling habits.

What drew me to it is that you can’t change the setup without connecting to the Mac, a solution I find much more comfortable than having a friend type in a pin, as well as easily restricting domains and apps including system apps.

For example something I still do is disabling all mail clients including the system one. I don’t need email on my phone. It’s an inherently asynchronous communication medium and it can definitely wait until I’m home.

Comment · 5 months ago
Yes
Comment · 5 months ago
I think more people should set up their iPhone using Apple Configurator, a Mac app used to control apples mdm solution. You do have to factory reset your phone for this once, but after that you have extremely granular control over what you can and can’t do. It’s much more powerful than the parental controls system and much harder to circumvent.

I use it to straight up disallow a bunch of apps and websites (tiktok, Reddit, YouTube, etc.)

For a while I even uninstalled safari which you can just do with this. Not having a browser at all on your phone is a neat experiment and really changed how I interact with tech on the go.

I did eventually install safari back, but overall I prefer the Apple Configurator setup a lot over any of these kinds of apps.

Comment · 5 months ago
I’ve had the same experience. The ui is a little unclear about this, because it says you have 5 scans, but 1 scan is just the continuous monitoring of the default branch of a repo.

The high impact findings have almost all been bang on for me. I was especially surprised by the high-quality documentation it produces as well as how narrow the proposed fixes are.

I’m used to codex producing quite a but more code than it needs to, but the security model proposed fixes that are frequently <10 loc, targeting exactly the correct place.

It’s really quite good. I’m assuming it’ll be pretty expensive once out of beta, but as a business I’d be jumping on this.

Comment · 8 months ago
The author clearly disagrees with that statement.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton