wslh

The current state of AI is incredible, and useful and doesn't need to reach AGI to be revolutionary. For example, I uploaded a conversation between a few people and not only asked about translating the text but doing a psychological analysis on turn-turning and other conversational cues. Just around a decade ago, the speech-to-text Dragon Naturally Speaking[1] was not reliable with only one speaker without any background noise.

[1] https://en.wikipedia.org/wiki/Dragon_NaturallySpeaking

Combining information is certainly part of creative work, and LLMs are very strong at that. But creativity goes beyond aggregation. It is an elusive, open-ended concept, not something we can measure as cleanly as math or language ability.

Don't understimate the massive data you need to make those networks tick. Also, impracticable in slow training algorithms, beyond if they were in GPUs or CPUs.

Anthropic is astroscaling. We're essentially buying into a loop where speed and iteration take precedence over stability and support. If you view them as an experimental lab undergoing rapid atmospheric friction rather than a company, the "unreliability" is just the cost of being at the frontier. This is not an endorsement for Anthropic, just imagining their craziness on how you "can" grow in a fraction of time.

And, I also I think many of the mobile and web apps will end up in prompting in the next few years.

Related and insightful: "GPT-5.5: Mythos-Like Hacking, Open to All" [1].

[1] https://news.ycombinator.com/item?id=47879330

Just a tip: add [translated] subtitles to the top video.

Since the thread is full of coffee enthusiasts: I recently stepped outside my espresso-only routine and started appreciating V60 and Origami pour-overs. It's been great realizing how much depth there is outside of espresso. Where does the rest of HN crowd land on pour-overs?

Decades ago, I worked with my uncle in a family shop. Every single day, he sent me to the bank to ask for the balance. Then, they innovated: a person at the bank finally started giving him the balance over the phone.

I would be careful about attributing all of this to ordinary insider trading. Another plausible explanation is that some activity could come from state linked or intelligence adjacent actors, where the goal is not only profit but also testing or exploiting prediction markets as an auxiliary information channel.

Good analogy, but there's a key difference: mainframes were an institutional dependency, whereas the world's reliance on LLMs is consumer driven, ubiquitous, an uncapped (e.g. spend more on the same "loops"). Completely agree on the second point, though powerful local models are the inevitable next step, and they are arriving fast.

No, that's not my point. My point is that AI looks like something fairly unique in today's landscape: a resource that almost everyone is starting to depend on. It's a bit like the Internet, except usage is metered, and paying more can improve the quality of the result for the same underlying task, such as cybersecurity.

In this context I also imagine we will have greater and greater local models, and the (dependency) ending game is completely unclear.

But the unique thing about AI is that the "world" is depending on it like water, oil, gas, etc. Not just a specific use case.

> So intelligence has turned into a utility.

"Utility" is close, but "energy source" may be closer. When it becomes the thing powering the pace of work itself, raising prices is less about charging for access and more about taxing dependency.

I imagine that you will need to use a giant budget for covering those codebases, and you can always run a longer and more expensive loop.

I wonder if vibe coding is partly what happens when software engineering fails to converge on reusable abstractions. Instead, we got fragmented tools and endless reinvention of the same components, and LLMs arrived as an ad hoc abstraction layer on top.

There is a publication making a related point in the DeFi security context: as TVL rises, the incentive to attack rises too, and defenses do not (or cannot) automatically scale with it[1].

[1] https://web.archive.org/web/20240911103423/https://www.bittr...

It seems that sites as Wikipedia (including their faults) serves as a good checkpoint of LLMs results.

I understand that there was a problem finding a useful PoW from the computer science perspective, and that paper is showing a solution.

Now, it doesn't necessarily lower the cost of an attack because you can adjust the required output to a cost that is suitable.

There is a breakthrough on a more productive Po(useful)W: https://news.ycombinator.com/item?id=47430951

I'm starting to wonder whether what Glasswing really shows is that parts of security have already gone underground: black-hat teams and state actors may already know about many more bugs than the public record suggests, while many security professionals and clients still treat the relatively small set of disclosed bugs as the state of the art.

I hope they also help in improving battery management on Linux notebooks, even pressing vendors via regulations.

If you're following the Satoshi "archeology", you should definitely check out Sergio Lerner's analysis of the early mining patterns. He provides a fascinating forensic look at the actual machines Satoshi likely used, using data like endianness and nonce-incrementing behavior to reconstruct the original mining environment [1][2].

[1] https://www.youtube.com/watch?v=iwVXC1Lf00k

[2] https://bitslog.com/2020/08/22/the-patoshi-mining-machine/

That would be very risky for Iran because the top stablecoins could be freezed. They are centralized.

LLMs and human intelligence overlap, but they are not the same. What LLMs show is that we don't need AGI to be impressed. For example, LLMs are not good playing games such as Go [1].

[1] https://arxiv.org/abs/2601.16447

> Big opportunity for the other labs, if that's true.

It sounds like this is considered military grade technology as cryptography in the 90s. The big difference is it's very expensive to create, and run those models. It's not about the algorithm. If the story rhymes it could be a big opportunity to other regions in the world.

I imagine that some levels of patching would be improving as well, even as a separate endeavor. This is not to say that legacy systems could be completely rewritten.