Of course I disagree because this generalized statement is quite obviously false. Have you used electronics made by hand? Versus made by a machine? You can't possibly claim that a PCB assembly by hand has higher quality than done by a machine. That the transmission of a message by hand from Berlin to Paris is higher quality than if communictations electronics was involved.
Of course you didn't mean that (but you wrote it). You probably mean a creative process. The process I descibed is still having the creative human in the loop. Just that the machine is filling in the gaps that are qualite mechanically fillable. Nothing strange with that.
Your configuration structs are not "everything". You probably have dozens of things in them. You have hundreds to thousands of different parameters to different functions. By the time your configuration structs have hundreds of things in them you have thousands to tens of thousands of parameters to functions. Only a fairly small fraction of things make it all the way up to main.
This is exactly what I was talking about with this being a "cognitively available" operation. You notice when you have to propagate something all the way up to the top precisely because it is not something you do very often and it stands out as an exception when you do. If you are literally propagating everything up to main, all the time, every time you change any function anywhere in your program, you are doing something very, very wrong.
It's counter-culture and I'm here for it. I had a fun convo with my daughter about why rail cars are so frequently tagged. More fascinating are the hidden pieces in tunnels and the like, where only few people will ever see it.
People are seriously confused what is being created here. The claim here we will have misaligned AGI, not a mindless LLM. If you don't know and you don't agree what you are even creating, how can you defend against it?
If this LLMs are statistical parrots, then the risk is someone using it as a weapon, to hack etc, but that is different risk category that why this person quit his job with different diffence mechanisms.
I joined a company using Notes back in '00. Performed the 4 to 5 upgrade, skipped 6 and went to 7.
We had our "email blast" code in the founder of the companies mail database. When we launched new features or had various announcements, you would log on to Mark's email database and run those features in the drop down. Of course there were various permissions in check that only certain people could do (or view) certain things in the database, but that's where it ended up...
I always thought the data/design replication was cool as well. I had to bake in what I called "relational fields" into the documents to be able to perform some standard relational database type of work, but it was really easy to work with overall. Maybe too easy - as one person mentioned in the thread, there were hundreds of little custom apps built by anyone/everyone in the company, because it was just that easy.
I played around with codegraph. I did a quick experiment where I gave claude code a feature to implement, and I had it perform the same task with and without codegraph. This wasn't scientific, so take it all with a grain of salt.
The codegraph implementation used fewer tokens. However, I performed an automated code review on each implementation, and there were more issues with the codegraph version. These weren't major issues, more like it re-implemented a function that already existed, or used styling that wasn't consistent with the rest of the codebase. My theory is that there's useful information an LLM gleans while grepping through code that is missed when it uses codegraph instead.
What if you have 10-20% certainty the AI being built will kill everyone? Go on a rampage and land yourself in prison just so Lab B or China can will the arms race and their AI can kill everyone? Probably not. Quit your job? Sure.
My initial testing on a very small dataset, please note I am not a music professional but I have generated my fair share of musics with Suno for a couple of years :
v6 has the same downside of "mainstream beat" as v5. v4.5+ is still the best for a "real organic" song vibe.
v6-wild is ultra-sensible to the weirdness factor. Where v5 at 70% weirdness would produce something interesting, v6-wild made garbage.
v6 generations are more bass heavy. Listening with a headset, I feel if I were to play this on a proper speaker the walls would shake.
> But... after more than 50 years in tech, I still use Thunderbird as my email client. If I could figure out a way to get it to turn off HTML mail I would.
This is unreasonable because nobody can know 100% certain that the AI built at Anthropic will kill everyone. It is also a big-ass strawman because the author never claimed he was 100% certain that AI will kill all humans. You're arguing from an irrelevant hypothetical (in which, I might add, most people would still not automatically change into murderous psychopaths as you apparently think is normal).
> Horse Shit. Here's what was written
It's a whole thread, my man. And nowhere does it say "omg I’m quitting they’re going to kill everyone". Read the entire thread again and then say with a straight face that the gist of it is "omg I’m quitting they’re going to kill everyone".
Reading these replies and you'd think HN'ers have no fun (or feelings) at all. The movie is funny, clever, clean (good for families/kids), emotional, and has important lessons and metacommentary that adults will enjoy. Not to mention the irony of the plot with what happened to the movie in real life. Go see the movie.
That requires abolishing capitalism, and since these AIs are going to hyper-empower the already powerful capitalist ruling class, that utopian outcome seems unfortunately extremely unlikely.
Feeding ever more amazing technology and science into a broken social/political/economic system does not result in a better world unfortunately.
Usually when stuff like this happens, it means that Tailwind's developers are moving to work on Spotify instead, which means Tailwind itself goes into maintenance mode. That's what usually happens when a dependency is picked up by a company with a huge, "more interesting" product -- the dependency starts to depend on the more interesting product for attention, and stops receiving attention on its own merits. It grows only to support the more interesting product.
Furthermore, I already have specific experience with Tailwind Labs neglecting their other libraries that aren't Tailwind itself, probably because they were only meant to support their paid products (which now will no longer exist! so probably those other libraries will stop getting updates too). Headless UI was something I wanted to rely on heavily, but it's not built to be flexible enough and basic issues from 2021 or earlier haven't been resolved or seemingly even reviewed.
Now that they have Spotify, I'll be surprised if Tailwind continues to receive any more attention than Headless UI did all those years ago.
So if I'm understanding this right, it's a meta harness, but for the current "cowork" class of products (ChatGPT Work, Copilot Cowork, Claude Cowork) as well as collaboration space agents like Claude Tag?
