CobrastanJorji

My friend, there are people are out there telling their AI personal assistants to go on Amazon and purchase a surprise present for themselves based on their known preferences. And then they intentionally do not inspect what the AI purchases because it would ruin the surprise.

People have wildly different risk calculators.

That's interesting. If you're usually rained on when there's a 10% chance of rain, that suggests that either you live in a particularly rain section of Portland, or that you move around to a large percentage of the Portland area each day, or that somebody's using a different definition for "chance of rain" than I do. I wonder which one is going on.

I think they were supporting you -- you had free time because you carefully evaluated whether the random story you came across was AI. The 'author' did not have free time because they asked an AI to generate a story instead of writing one theirself.

> Will it rain over forty percent of the city? Will it rain for forty percent of the day? Or does the speaker just not know and is too shy to say so?

Unrelated, but to answer his question, it's the probability that a random spot in the covered area will get at least a smidge of rain at some point in the covered time period. A 10% chance of rain can mean either that the forecaster is 100% sure that exactly 10% of the area will get rain, or it can mean that the forecaster sees a large rainstorm nearby which would cover the whole area if it arrived, but which only has a 10% chance of arriving. Or anything in between. If there's a rain cloud meandering over the city, covering 1% of the city at a time, and over the whole day it randomly walks over 10% of the city, that's also a 10% chance.

I agree. What I expected was a place for very personal little stories and journals about specific, hyperlocal spots on Earth. What I got was an empty site for public polls on restaurants in large cities.

Mate, it wasn't wise to believe that in 1952 either. It's gotten worse, sure, especially at the top, but there are still thousands of "politicians" out there running local board meetings and trying to do their best to keep their communities going.

I wonder. Say I wanted to prevent something stupid like this from happening in the future. What would be a sensible check? The higher level authority doesn't see the private key itself, right? Just the certificate request. And I wouldn't want someone to manually inspect the keys to see that it looks random-ish. Perhaps there could be a small check script added to the generation of certificate requests that says "hey, your key a.) is mostly zeroes, b.) appears to be ASCII text, or c.) seems otherwise extremely non-random, please use -f to continue anyway".

I dunno, it seems like decisions about what to build in a given spot should be made as locally as possible. If I want to build a house in Sonoma, the voters in LA should not get a say.

Major projects like power plants or dams should be made at a state or even national level, but rules for which house goes where, less so. And if your town's rules are overly burdensome and stupid, that's an issue for your town's government to solve.

As an aside, I do not like unnecessarily misleading graphs. Own your strengths and weaknesses.

In the "Results" section, they choose to compare Ultramarine only to models that it's better than on every metric. They ignore models like Fable 5.1, GPT-6 Astra, Claude Opus 5.5, etc.

Then, in the "Score against cost per task" graph, which is focused on price, suddenly all of those models that perform better than Ultramarine come roaring back because they are also more expensive.

Just list yourselves as the second best sometimes. Your thing is impressive. Don't let the idiot marketing guy hide that behind stupid tricks.

What's been driving me nuts recently is when coworkers send me clearly AI-generated blocks of text in otherwise-human conversations. Like I'll ask in a chat window "hey, Jim, what's the difference between these two deployment pipelines," and I'll get back a block of text with markdown and headings in the style of our team's AI, and I get this sad feeling where I thought I was going to be having a conversation with a colleague, but instead my colleague was just playing as a chinese room.

You're way overthinking this. The administration told Anthropic to let them do something, Anthropic said no, and then they got pissy and started throwing rocks and wouldn't back down because the defining trait of this administration is that they walk into stupid positions and then refuse to back down because backing down would make them look weak and wrong.

You're ascribing complex motives to toddlers.

Nobody's immigration fears are data-driven, nor have they ever been.

What is a possible use case for a legal reasoning tool that is wrong 46% of the time?

> Astra for Law passed the evaluation’s overall correctness check on 54.0% of questions.

How is this a product that you are selling?

> I don’t like the idea that anything I say in a public setting...could end up in a transcript summary on someone’s personal device.

My friend, if there's a reporter standing behind you, it could end up on the nightly news. And it's been that way forever. Certainly, a plethora of terrifyingly broad data collection greatly exacerbates the problem, but it's not a new problem.

> bla bla bla...Amazon Echo...listening in on us without our knowledge or consent. I've lost track of how many times I've experienced or heard others...discussing a particular topic...and then as soon as they get on Facebook...they see posts and ads featuring the very topic they were discussing.

There are lots of real problems with surveillance tech, but the stuff about smart speakers specifically is still a dumb conspiracy theory. People are CONSTANTLY testing the hardware in these things and watching network activity. They really are not listening to you, and blindly repeating this claim greatly harms your credibility.

> bla bla bla...George Orwell envisioned in his dystopian novel 1984

Okay, new rule, these articles need to pick at least one other dystopian novel. There are so many. What's wrong with Fahrenheigh 451, where the government had everyone look out their doors in a mass manhunt? That's a good one! Use it to talk about Ring cameras or something.

I love it.

No, but take other applications with a grain of salt. Everybody gets excited about the idea of having a dozen large virtual monitors floating around your desk, but even really great resolution has just not been good enough to make those displays look crisp. Also, a lot of folks still get pretty bad headaches when they use VR or AR. So MOSTLY it's for games and movies, but people remain convinced that there's a big killer app here somewhere that we just need to figure out.

The NYT watches an event like Trump making a crowd of supporters swear to "cheat like hell" and to vote even if they're not registered, and they turn it into headlines like "Trump leads supporters in a pledge to vote." They are no longer a reliable source of accurate information.

Plausible. Figure 100 GB each of search index for Stack Overflow, Wikipedia, and GitHub issues, then add a dozen more for docs of all your favorite techs. So maybe half a terabyte. Download and build updated dumps of those once every week or two, and it'd work pretty well. Impractical, but possible.

What's wrong with my interpretation?

That's just it, though. The people who buy electric cars because they want to help the environment are wrong. They're only doing a good thing because they're self righteous pricks who are full of themselves. They're just as bad as the polluters.

The people who want to stop smoking are wrong because they're self righteous pricks who are full of themselves. They're worse than the cigarette companies.

The guy who want to help the environment is wrong because he's a self righteous prick who is full of himself.

The folks raising awareness about North Korea are wrong because raising awareness isn't actually helping people directly. It's just a bunch of self righteous pricks who are full of themselves.

Any one of those episodes has a perfectly valid point to make. But taken as a whole, the overarching message is that both sides are always bad, and the one that sounds like it wants to make the world a little better is probably worse than the evil they're opposing.

See the word "rare" in the headline?

They didn't go after Gore because he was wrong. They went after him because he committed the cardinal sin of really, earnestly taking a stand on a big issue. That is NEVER okay in South Park. There is never a correct side. Both sides must always be bad, on every issue. The "correct" view in South Park is always to not really take a stance on anything, and any character who is strongly on one side or the other on any issue will always be portrayed as wrong to do so.

That's not to say that South Park doesn't have some great episodes poking fun at some very deserving targets. They absolutely do. Some real gems in recent seasons.

As soon as I thought "man, this sounds like some evil sociopath shit," my second thought was "oh, Sam Altman must have been personally involved."

Appendix C is blasphemy against cuberule.

Thanks, I was trying to work it out in my head. I knew brazilian jiu jitsu was 20th century, and I was like "but wait, is jiu jitsu itself not that much older, that seems very surprising."

Tangentially related, but has anyone else noticed a big increase in fraud attempts over the last year or two? I hardly remember any a while back, but I've gotten probably a dozen calls this year that go "hey this is AmEx please confirm your credit card number" or some such.

Merlin's story is very interesting: https://www.birds.cornell.edu/home/the-magic-of-merlin/

The journey to a magical, free bird sound ID app involves some obvious steps (the NSF gave them money), but the amount of timeline offramps is really high. The NSF didn't think this app was educational ("it's just telling them the answer, they don't even have to work for it!"), the funding ran out so they were thinking of charging money, eBird happened to have five years of bird sound recordings they could use, a computer vision guy on the team happened to learn that spectrograms are basically the same as pictures...

And the result is tremendously useful! They get great signal on where birds are, and they're making people so much more interested in birds and nature and science. And I think charging $1 for the app or, shudder, some sort of subscription model might've ruined the whole thing.

This would maybe be a pretty good opportunity to use some of that "upscaling" AI tech: given this weird results of spatial scanning, produce something plausible and much higher definition. There are some dangers of doing so, of course. People will believe details that show up in it because it's taken from real life, and AI loves adding fake details, but for a little game it seems like detail beats ground truth.

Maybe it would, but I didn't say it was a good idea: I said I think people would vote for it.

> Why don’t the poor people vote to take all the money away from the rich people and redistribute it amongst themselves?

It's because the Powers That Be don't put it on the ballot. If you put "take 99% of the money from the 1000 richest billionaires (about $10 trillion, I think), and nobody has to pay any income taxes for the next four years" on some sort of national ballot, I think there's a very good chance that it would pass.