jetrink

> Tokens become cheaper than tool calls

The author observes that a call to GPT-5.6 Luna is only 4-5 orders of magnitude more expensive than grep, and then predicts that at current rates of progress, calling an LLM will soon be cheaper than a grep. I think this is a good time to invoke Stein's Law: "If something cannot go on forever, it will stop." These efficiency improvements won't continue forever. It's more likely that the per-call cost of high-quality, compiled software like grep will be a lower-bound that LLMs asymptotically approach, rather than a line that they blow past with perpetual exponential progress. (Barring a true breakthrough in something like quantum computing or room-temperature superconductors.)

Yes, Meta has consistently acted with integrity and has earned the benefit of the doubt in situations like these. It's safer to assume that the massive PR team working for this one-woman play in Spain has manufactured the whole thing, including the damning quotes, which the Guardian has somehow published without properly verifying.

I don't like Citizens United either, but you should better inform yourself about the decision.

1. The idea of corporate personhood predates CU by over a century and the Supreme Court had already asserted that corporations enjoyed certain constitutional protections in previous decisions.

2. Far from inventing the idea, the CU decision didn't even rest on corporate personhood, but on the idea of the freedom of speech generally. The logic of the majority was that speech itself is protected, irrespective to whether the speaker is a person or an organization. The First Amendment covers individuals, but also newspapers, book publishers, radio stations, and so on, and that should extend (they said) to non-media corporations. No assertion of personhood necessary.

The problem, in my opinion, is that that conclusion combined with previous decisions that treated limits on spending as limits on speech, allowed for unlimited spending. The majority also naively asserted that independent spending posed no risk of corruption, which I think is laughable.

Centuries, in fact. For instance, Isaac Newton was involved in multiple priority disputes, since he tended not to publish promptly.

I really want to see Claude's store-brand speedy blue hedgehog. Edit: Haha, thanks.

Could this be mitigated by allowing voters to vote multiple times and receive multiple confirmations, but only count the last vote for each person? That way an individual could prove that they voted a certain way to a third party, but there would be no way for the third party to know whether it was their final vote. Meanwhile the individual knows that they spoiled their previous vote or votes and that their final vote is included.

I would expect to see LLMs that are creative in math before any that are creative in writing. Creativity is more easily specified in math and the solutions can be formally verified. There's no good way to classify creative writing. Many truly great works are overlooked by experts and the public until decades later. Many derivative works are commercially successful.

I can see the video and view the first three seconds or so, but after that, it throws up a login prompt.

I feel like if I attempted this, the bike frame would look fine and everything else would be completely unrecognizable. After all, a basic bike frame is just straight lines arranged in a fairly simple shape. It's really surprising that models find it so difficult, but they can make a pelican with panache.

The religious right, specifically. They would say that all people are descended quite recently from Noah and his family.

Hilariously (and appropriately), the decision cites Sony Corp. of America v. Universal City Studios, Inc., also known as the "Betamax case."

> (a) “The Copyright Act does not expressly render anyone liable for infringement committed by another.” Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417, 434.

> In Sony, copyright owners sued the maker and the retailers of the Betamax video tape recorder. Id., at 422. The tape recorder could be used to record copyrighted television programs for later personal viewing, which would not constitute infringement. Id., at 449. On the other hand, it could also be used to reproduce and sell copyrighted television programming, which would constitute infringement. Ibid. The lower court found the Betamax maker liable because the tape recorder was “not suitable for any substantial noninfringing use” and infringement “was either the most conspicuous use or the major use of the Betamax product.” Id., at 428 (internal quotation marks omitted). This Court reversed, concluding that “[t]he Betamax is . . . capable of substantial noninfringing uses”—like personal use—so “sale of such equipment to the general public does not constitute contributory infringement.” Id., at 456.

It could create the right sort of incentives though. If I'm a junior and I suddenly have to take my work to a senior every time I use AI, I'm going to be much more selective about how I use it and much more careful when I do use it. AI is dangerous because it is so frictionless and this is a way to add friction.

Maybe I don't have the correct mental model for how the typical junior engineer thinks though. I never wanted to bug senior people and make demands on their time if I could help it.

At first I thought the "unmanned tunnels" description was just a way to avoid broadcast regulator scrutiny, but it does look like it's genuinely designed to be used underground as part of an emergency alert system. That led me to "leaky feeders", a type of broadcast antenna used in mines and tunnels.

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

FTA

> Hand axes were held in the palm rather than attached to a wooden handle.

> their groceries won’t make it to the stores and packages won’t get to their homes without a robust road network.

A road network isn't the only solution. In the early 20th century, for example, there was a separate narrow-gauge tunnel network beneath Chicago dedicated to freight. Deliveries were made directly to businesses via subbasements or elevator shafts. The network had stations at rail and ship terminals for accepting freight arriving from outside the city. At its height in 1929, the network had 150 locomotives pulling 10 to 15 cars per train.

1. https://en.wikipedia.org/wiki/Chicago_Tunnel_Company

> Did the average person throughout history have more leisure than we do? I doubt it.

Recent anthropological and archaeological research is challenging the traditional view that ancient lives were "nasty, brutish, and short." Instead, it appears that many ancient peoples worked less than eight hours per day and frequently took time off for festivals or to travel long distances to visit friends and family. And unlike today, work usually had a more flexible rhythm where short periods of hard work were separated by long periods of light work and rest.

Since this is generating revenue for NYC, you can't consider whether this tax is good or bad in a vacuum though. The alternatives are a different tax with its own effects, or more debt, or less spending. (In this case, the revenue goes to the MTA.) Any opportunity costs due to less traffic are at least partially offset by opportunity costs you aren't having to pay somewhere else.

Also, many of the posts seemed intended to be humorous and satirical, rather than merely 'futury.' They made me laugh anyway.

> Google kills Gemini Cloud Services

> Running LLaMA-12 7B on a contact lens with WASM

> Is it time to rewrite sudo in Zig?

> Show HN: A text editor that doesn't use AI

I heard one of the founders interviewed on the Hard Fork podcast[1] (which confusingly is primarily concerned with AI, rather than crypto.) I went in with very negative expectations, but came away with a positive impression and optimism that they might be onto something. As you say, AI is not core to the project. Instead, the focus is on using technology to facilitate individualized learning. It is true that teachers are 'replaced', but by humans whose job it is to keep the students focused and motivated, rather than to convey information.

1. https://www.nytimes.com/2025/09/05/podcasts/hardfork-educati...

The author credits Alexis King at the beginning and links to that post.

The videos are more subtle and it's not apparent in every frame. Look for things in the background snapping into and out of focus, weird textures appearing on Will's head and neck, and people's faces looking unnaturally sharp at the edges, while their skin is uncannily smooth (sort of like Max Headroom.)

I understand the point he's making: New cars are much safer than old cars, and the average person is driving a car that is 12 years old, while new cars are bought primarily by the wealthy. However, that seems like a natural consequence of two things that are very good for everyone. First, cars are lasting much longer than they used to, which lowers the lifetime cost of ownership. Second, cars have gotten much safer in the last fifteen years. As long as these trends continue, the safety gap will exist, but I think everyone would still prefer cars keep getting safer and more reliable.

"Dave, this conversation can serve no purpose anymore. Goodbye."

https://www.youtube.com/watch?v=YW9J3tjh63c

This is such a wonderfully unlikely story. The museum, known for its dinosaur displays, was drilling a borehole in its parking lot as part of a building improvement project. The 5cm borehole happened to go straight through the spine of a small dinosaur.

The Bitter Lesson is specifically about AI. The lesson restated is that over the long run, methods that leverage general computation (brute-force search and learning) consistently outperform systems built with extensive human-crafted knowledge. Examples: Chess, Go, speech recognition, computer vision, machine translation, and on and on.

It turns out that Woodland v. Hill is not about landscape photography.

I jumped over to the Wikipedia page of early blogger Justin Hall to see what he's up to. He has another distinction that he can probably claim: The longest recorded gap between registering a domain and finally using it to start a business.

"In September 2017, Hall began work as co-founder & Chief Technology Officer for bud.com, a California benefit corporation delivering recreational cannabis, built on a domain name he registered in 1994."

It is an accurate depiction of how Chicago police operated, unfortunately. In fact, one Chicago detective who tortured suspects went on to work as an interrogator at Guantanamo Bay[2]. It's terrible that the series would glamorize that behavior.

1. https://chicagoreader.com/news/the-police-torture-scandals-a...

2. https://en.wikipedia.org/wiki/Richard_Zuley

That sounds like a reasonable prediction to me if the LLM makers do nothing in response. However, I'll bet coding is the easiest area for which to generate synthetic training data. You could have an LLM generate 100k solutions to 10k programming problems in the target language and throw away the results that don't pass automated tests. Have humans grade the results that do pass the tests and use the best answers for future training. Repeat until you have a corpus of high quality code.

I just looked it up. The process involves modifying a fertilized egg or embryo to create a founder organism. The gene drive is designed to be inherited by more than 50% of the organism’s offspring and propagate through the population over successive generations. Because humans have fewer offspring and longer generation times compared to insects, it would take many generations and potentially hundreds of years for a gene drive introduced in a single human to spread widely through the population.