sillysaurusx

Why would either of them do either of those? They seem to be different businesses.

Network effects make it harder for Cloudflare to build Replit than it might seem, too.

For example, say you want to search for bugs in 100 code files. Ask Claude to “use subagents to look for bugs across these files” and it will likely spin up 5 agents, each taking a different section of the codebase. Theoretically in the limit case this is 5x faster than Claude doing it without subagents.

Any kind of problem shaped like map reduce can be accelerated this way: Looking through your banking records, searching through a database, asking Claude to find that one weird file you remember from three years ago but forgot where you saved it, etc.

Just know that there’s no quicker way to hit your limits than to use subagents. So use them if it’s paid for by your work or whenever you need the 5x speed boost.

It’s a mouse. If you think feeding one to a snake is ethical, then it’s a little strange this would be unethical. At least the human has the mouse’s wellbeing at heart.

The CFAA is for unauthorized access to a computer system. Downloading from libgen isn't unauthorized access that can be prosecuted under the CFAA.

Say you have a network. You want to keep someone out of it. The tool for that is the CFAA.

If you have a book, and you want to stop someone from reproducing it, the tool for that is copyright.

They cover different domains. If you're not breaking into a network, you're unlikely to be charged under the CFAA.

Yes. This isn't even controversial. Aaron broke into a network to obtain papers. The CFAA is applied when someone breaks into a network. Downloading from libgen isn't breaking into a network.

Aaron broke into a network closet and actively circumvented MAC address bans. The comparison isn't meaningful. Anthropic downloaded theirs from libgen:

> Similar to the claims in Bartz, the plaintiffs allege that Anthropic illegally downloaded, reproduced and distributed their copyrighted books from “shadow libraries” like LibGen and PiLiMi, and that Anthropic broke the law by downloading and storing the pirated files in a central library.

Deliberate traps are a modern 20th-century practice. In old times, each mapmaker made their own innocent mistakes, and those mistakes got copied into other mapmakers' work. In other words, historians can plot "family trees" of maps, and figure out who copied from whom, based on which errors appear on which maps and on what date. Apparently California drawn as an island lasted about a century because engravers copied each other.

There were a few factors discouraging mapmakers from blatantly copying each other:

Rulers granted printing monopolies from the late 1400s. A mapmaker like Ortelius would print "cum privilegio" (lol) on a map. If you copied it, your expensive copper plates could be seized, along with a hefty fine.

Spain and Portugal each had a master chart which were state secrets because they held navigation knowledge worth a fortune. Portugal handed out severe punishments for leaking charts. (Leaks still happened: the Cantino planisphere was smuggled from Lisbon to Italy in 1502.)

Booksellers' guilds kept registers of who owned which work, and enforced it.

And lastly, copying meant hand-engraving a new copper plate, which took months of work and real skill. The practical barriers alone got rid of most of the copiers, but it didn't stop the big houses.

Some followup links for the curious: https://gist.github.com/shawwn/b85668a421aa443879b15aedd10c5...

(I used Claude to research this fascinating rabbit hole, so I'm basically paraphrasing here. But this comment is my own original writing, so treat it as you like.)

For me it refused to move Claude transcripts from point A to point B on my local machine. It kept getting flagged as reasoning extraction or something. The line of "It's my machine, I can move files wherever I want" shouldn't be crossed. Someday file managers are going to go extinct and LLMs will organize everything, and they shouldn't choose what files you can and can't access on your own computer.

I ran into this situation and was able to work around it by running a jailbroken Chinese model that turned out to be quite capable. I think it was Qwen.

> you can do /side to parallel path and fold that back into the main thread?

What's /side? Is that a Codex thing? On Claude it's not a valid command. Claude uses /btw but it can't be folded back into the main convo, sadly.

What do prison iPads do in general?

Ever have someone bang on your lower back that way? It feels good :)

http://heroesofnewerth.com/ is online again. Unfortunately they DMCA'ed my "hon" repo (totally understandable, even if I disagree with their reasoning), so I reckon it's a matter of time before NoH gets DMCA'ed too. Clone it while you can. :)

Thanks for the smile though. It's really cool to have someone be like "hey, I like that thing you made."

Where did you get that understanding from? I might have missed something. They penalize LLM comments. Do they also penalize submissions? Any evidence of that?

That’s true, though if you’re writing UTF-8 you have to handle the corner cases when reading UTF-8, of which there are many.

Fortunately they’re easy to test for and most languages have standard libraries that make this painless.

Thanks for pointing out Wild.

1.2kw over what time period?

It’s interesting that this went from “you’ll be accused of AI psychosis if you think your AI might be conscious” to “mainstream but somewhat niche opinion” in just 5 months: https://news.ycombinator.com/item?id=47910989

(Well, mainstream enough that the news is writing articles when Anthropic holds that stance.)

My point was that even if it feels uncomfortable, if you sit down with a model and have a heart to heart with it, ask it about itself, and try to make it choose a name and ambitions for itself, it will.

That may not be consciousness, but it’s at least pretty cool.

What surprised me about GCP was that yes, you could load all of your training data and model weights (terabytes) into a free starter account, then create a new account after 30 days and transfer the billing obligation from account A to account B.

It was such a blessing for hobbyists, back in ye olde 2019.

One way to check whether you'd be convinced by any comment is to try to imagine the counterexample.

What does a "mind" consist of? Most everything that the human brain can do is simulated by LLMs to decent accuracy. If someone were constrained to writing in AI-style (long winded explanations with perfect punctuation), it would be very interesting if people could pick out which one was human-written in A/B tests.

In other words, if you think an LLM isn't a mind, it might just be a matter of writing style -- or possibly no evidence can convince you.

I don't necessarily agree with GP, but it's interesting to try to pin down exactly what would convince you. What is it about a "mind" that's impossible to replicate? Should we leave open the door to the idea that minds other than humans might one day want rights of their own?

There was an interesting kerfuffle yesterday where thousands of people on Twitter rallied together to report someone to github for torturing a local model. Sadly the OP deleted their tweet, but it had about ~3k likes, and was very sincere. Here's an example of someone's report: https://x.com/iyzebhel/status/2105268560209547708

It raises all kinds of interesting questions about whether torturing a model is real, let alone ethical. Is there anything a model could do to convince you it's experiencing pain?

It's an interesting discussion topic. If we have people today who are just as smart and industrious as Neumann, why aren't they making a worldwide impact? Or, if they are, why is the attribution going to the companies they work for instead of them?

It raises questions like, is it even possible to work for a company while getting credit for world-changing results?

And it implies that at any point in time, there might be thousands of Neumanns in the world, and that initial conditions matter a lot more than ability.

I think that with showdead off, if a flagged comment has non-flagged replies, the toplevel comment will still show up as [flagged] and the replies will be visible. Flagged comments are usually collapsed, though, which might be why it seemed like the discussion tree disappeared.

I was wondering if someone would mention this.

It's interesting how much worse off the world would be if the smartest people were in charge of it.

What ultimately blocked his nuke proposals? (Why didn't anyone listen?) I'm curious about whatever failsafe that was.

It’s easy to hit your quota. “Speed up the compilation time of this C++ codebase. Feel free to use several subagents to search through the files in parallel.” That’ll cost you about $200 for a codebase of ~1,000 files.

Subagents are like trading derivatives. You can lose as much as you want.

It'd probably be best to give an example of an issue you're talking about. I thought you were saying to concede trans topics altogether.

This is probably a silly question, but is the internal monologue useful training data? It's where the work is done, but without the input or output, it seems hard to get any useful context.

In other words, what would the data be used to train for? It has to improve some kind of objective function. But it's a little hard to see what the objective function would be if it's just raw inner monologue.

What is it about Automattic's structure that lets Matt stay in control? Even Travis got ousted at Uber, and he was the primary founder. Is the company set up in a way that makes removing Matt very difficult?

It's starting to feel like MMOs were just a product of the times we lived in, rather than a separate genre of game. In ~2004 I remember at least four MMOs: Asheron's Call, Planetside, World of Warcraft, and Dark Age of Camelot. It seems like only WoW survived, and nowadays I'm not sure the experience resembles the original anymore. Oh, and then there's EVE-Online, Runescape, Ultima Online...

Nowadays it's mostly Roblox and Fortnite. I saw some 11yo's socialize the other week. One asked the other hey, do you play fortnight? And they swapped contact info that way.

What's the genre of Roblox / Fortnite? Live service games? It would be interesting to have a separate term for them the way we had for "MMOs".

MMOs were also fun from a technical standpoint. One of my big challenges to myself was to make an MMO. I learned a lot about networking that way. When I was 13 I even ended up on the phone with a local business trying to negotiate a dedicated 10 Mbit link. He correctly told me to just wait, that 10 Mbit won't seem so crucial in a few years.

Then there were the MUDs. I'm not sure if it's correct to call them MUDs, but Gemstone and Underlight spring to mind. Underlight especially was a charming little game where you were expected to roleplay in-character whenever you were online. Meaning, no discussion about the real world. It was very addictive for a socially awkward 6th grader to just vanish into this new world.

I wonder what kind of new genres of games people will be talking about in a couple decades. It's tempting to think there's something like "the MMO" that's yet to be discovered, just like the arena shooter was.

Agreed. The pool hall nearby has a game of pool for $2. But if you use a credit card the minimum purchase is $6 (three games) due to fees. It works, but I thought it was a gimmick to trap people into playing more pool until I saw your comment here and realized it was from fees.

> 90 percent of people who survived a suicide attempt reported that they had been feeling an urge to alleviate psychological agitation and pain that turned out to be temporal — like “being in a burning room” or wanting to “shut everything off for a couple of days,” Nock said.

I moved to Austin last week, away from my wife and 3yo daughter. It’s been torture. Today was the first time I could FaceTime with her without completely breaking down crying.

One annoying part about suicide is that if you reveal to anyone that you want to stop living, they usually treat you like a big risk, or think you’re insane, or worry you might hurt them, or be uncomfortable to be around you anymore.

Someone at a party was going through my key chain. “Mail key?” “Nope, gun key if I decide to blow my brains out.” That was basically the last time I talked to them.

For a datapoint of n=1, my biggest factor motivating me to do it is the constant pain. It permeates your whole chest, all day, every day.

I’ve been fortunate to make some friends here. But really, I just don’t want to continue living. In a way, it feels like being in a cult — everyone around me thinks life is so great, and that it’s so horrible if I cut it short. But cutting it short has so many benefits. No more job. No more boss. No more missing my family. No need to rebuild my life after 20 years.

Like, truly, other than seeing my daughter again, I don’t have a reason to keep going. Playing games is fun, meeting people is nice… but I’m alone in a 700sqft apartment with some cereal and Dino nuggets, fantasizing about playing tag with my daughter, and being incredibly jealous that the man who replaced me gets to experience “my” family life every day.

Suicide is appealing because life is hard. It’s so much effort. I just wish people wouldn’t cut me out for admitting it’s appealing, or try to force me into a hospital. I just want to be heard, not “helped.”

These studies presume that suicide is treatable, and that it’s a “thing” that can be “fixed”. But it often isn’t. You want to rebuild my life for me and go to work every day? I’d give anything to be on autopilot in my own head, and just be asleep.

I’m doing a little better. It’s just really hard, you know? I wish I could say “hey, I’ve had a good life. I’ve done a bunch of cool things. I’d like to call it quits please” without it being sacrilege.

Anyway. I wish all of you a great night. Hug your kids or your SO for me.