seri4l

In less than half a century, we've created 3 different time standards (TAI, GPS, and now UTC) that tick identically and differ only by arbitrary constant offsets (+37, +18 and +0 seconds). If we haven't come up with an elegant solution by then it's likely 65000 years from now most computing power will be spent running time-conversion libraries.

I wouldn't put AI on the same level as the inventions of the industrial revolution. Automation replaced specific tasks. AI is replacing humans themselves. What are humans supposed to do in a world where everything a human can do, a machine can do too?

Paper [0]

[0] https://link.springer.com/article/10.1007/s44338-026-00232-9

>5-10% loss at the register

Where did you get this number from?

Your psychiatrist looked at you like you were crazy?

So what software could be making these ENUM queries? Any theories?

It's the kind of feature that would rub people the wrong way, whether it's a cop, a friend, or a stranger you want to show a funny video to.

To give a couple of examples, using this would be a bit like locking the glove box in front of traffic cops, or going around the house locking rooms when you have a guest.

The only way to make it socially acceptable would be to make it some sort of default (for example, requiring face id when switching apps, with pin/fingerprint as a backup), which will probably be an option as the lag and failure rate of face id or other owner-detection technologies improve.

For erudites Word is a much better name. We seldom use that verbum.

Deepseek is, with difference, the most "Western" of Chinese models, so it's a bit perplexing that it was chosen to test this hypothesis.

I didn't run any benchmarks but I played around a little, and after getting around the API-level filter Deepseek V4's answers about "China-sensitive content" aren't any different from what I get from Claude and ChatGPT.

Probably a random Roman. People used to walk a lot, often barefoot. The random modern Westerner would get blisters within the first week.

Yes, distance is about the same. 1488Km in Google and 1075 Roman miles in Omnesviae [0], that is ~1590 Km.

[0] https://www.omnesviae.org/#OVPlace337_TPPlace1048

>This problem is solved in China

It isn't. China is the best example that draconian identity verification / KYC processes don't stop scammers.

I agree. In case you haven't heard of it already https://tldr.sh is pretty good for this.

>We can say a lot of things about what may happen next, but the main one is that AI is going to get better with scale.

Could anyone elaborate on this? Further down he talks about the necessity of bringing the cost of computing down. Is that really the bottleneck?

>it means that it's impossible to download public data from a web app without proxying it through your own server.

Maybe with a subdomain that points to the target website IP? If it supports HTTP and doesn't check the host header it should be fine. That's pretty exceptional these days though.

Edit: I just checked and a subdomain won't do it, it needs CORS as well.

I'm surprised to read that "the scientific wisdom of the time (...) held that there were no significant elemental differences between the Sun and Earth", because I distinctly remember looking up why helium was named after Helios (the Greek name for the Sun). It turns out that it was discovered by analyzing the spectrum of the Sun, only years later being isolated from Earth minerals.

I don't think I've ever bitten into a pear or plum with "unappetizingly soft or gristly bits".

Instead of collecting more data why don't do something with the data we already have? A quick look at the Fedora Bugzilla or the GNOME GitLab issues tab suggests the bottleneck doesn't lie in data collection, but in processing.

I get results for the Bible book when I search for "book of megadrive". Genesis and Megadrive are different names for the same SEGA console. Master System is a different console altogether.

What would you say is the primary issue? Personally I see vectorization and similar fuzzy techniques as the things that have most affected search results in the last <10 years, both positively and negatively. On one hand it allows me to search really dumb queries and still get sort of relevant results, on the other it makes well thought query tuning for more specific results seem futile. I don't know what goes on internally, but it's as if Google just discards as noise most of the words in the search query. "Verbatim" mode seems to tune that down somewhat.

Post-WW2 Germany comes to mind.

Many phishing lists don't treat github.io as a pseudo-TLD which leads to the entirety of *.github.io being blacklisted when someone uses Github Pages to host phishing or malware.

I'm sure many people do. There are tons of apps to scan and parse invoices. Makes sense as the current mid range phone cameras work just as well as mid range scanners but don't take any space and do that job an order of magnitude faster.

The ones with fish are great.

There are some experiments that suggest that human actions are not the result of conscious thought. Maybe you've heard about Michael Gazzaniga's experiments on split brain subjects. The most radical interpretation of the results is that our conscious thought merely constructs a narrative about actions it has no direct control of.

In other words, our brains act like a bunch of interconnected neural networks and one of them, the language processing part, tries to make sense of what all the others are doing, serving the only purpose of communicating with others.

I apologize for the double negative. What I meant is that hashing doesn't improve privacy because if you know the hash and the hashing function it's easy to build a hashmap of all the possible IPv4s (around 3.5B). Unless the hash uses some sort of expensive key derivation function, but that doesn't scale.

>hash the IP-address

How would that work? I can't think of any approach where getting the original IP back from the hash isn't trivial.

I think it may be Noto Sans Mono.

>I tried whatthefont

Has that website ever worked for anyone?

I'd say it's the last one, with a "subcommand based position dependant syntax" like git or zfs. It's a matter of taste but personally I prefer it to the traditional one. As long as the behavior is consistent remembering the order of the arguments is often easier than remembering the exact keywords.

>But start bringing in cars and trucks and see it crumble in a week

There are some Roman bridges that still get automobile traffic (Römerbrücke, Puente Alcántara) or were only recently pedestrianized (Puente Romano). There are more probably, that's just what I found after skimming some Wikipedia pages.

https://en.wikipedia.org/wiki/Roman_bridge

I always thought these types of services didn't go after account sharing as a form of flexible pricing. The users who are OK paying $10/mo pay that. The users who wouldn't pay $10 share an account and get a slightly less convenient service.

Are their profit margins so tight that a 1:2 paying non-paying user loss ratio is a net win?