hnhn34

Nope,they use Agent SDK now https://ampcode.com/news/use-your-claude-plan

this is no longer the case. If your harness handles login through the Agents SDK, you can use your sub now. That's why your Claude sub can be used in Amp and T3 Code for free now, but not in Pi (at least not without workarounds)

> If you have to write LAWS.bend which is pure code describing the laws then it isn't basically like those old days of writing unit tests and that too tests first hence the TDD?

There is a key difference: the laws are formally verified, as in Lean or Rocq (but much faster). So it's like writing a unit test or property-based test, but when it passes, you have a mathematical proof that you will get the expected output given ANY input in the infinite space of possible inputs. In traditional TDD, you make up some test case, write some asserts, and it passes if you get the expected outputs from those inputs and only those inputs. So you have to make multiple test cases for the same thing, and you still don't have any formal guarantee of your code's correctness.

> Also, why the compiler is not written against and with LAWS.md so that no audit is required at all?

Because it is mathematically impossible due to to Gödel’s second incompleteness theorem, which states: any consistent formal mathematical system strong enough to harbor basic arithmetic cannot prove its own consistency

Is it usable on Pro plan? Seems like it would burn my rate limits immediately

This is a false equivalence. There is no known correlation between amphetamines and a decline in IQ. If anything, they are nootropics (they don't make you smarter but improve your short-term memory, focus, etc.). Psilocybin stimulates neuroplasticity and adult neurogenesis. Cannabis, as far as we know, does no such thing.

> What's even worse, it definitely fucks up your neuromodulators

fwiw despite decades of research into nicotine, we have very little data on how bad it actually is at low dosages and via delivery methods other than tobbacco. The vast majority of research on nicotine by itself is in the context of smoking cessation. A lot of the risks people warn about wrt nicotine (dopamine spikes, vasoconstriction, dependence, etc.) is also present in any stimulant, including ADHD medication and even coffee. Also, again, very limited research on this, but some of it suggests nicotine could be helpful for ADHD.

That said, none of this really detract from your point about risks in children, and the trendy delivery methods like vapes and pouches seem to be extremely addictive.

> would that solve the problem for you?

not the person you replied to, but it mostly would for me. Factory farms are among the closest things to hell on earth.

If you have an LLM that doesn't make errors ever, then you have an ASI, at which point the conversation is meaningless. In the meantime, having a lower error rate but more uncaught errors is less important than making incorrect code impossible to compile, and/or flagged by strict linters.

Looks like there's a bug in WebKit-based browsers where clicking on something doesn't always work.

Nothing to add other than I really, really like this. Keep us posted on future updates!

slop is not, by definition, AI generated. The word slop is from the mid 16th century, and its modern colloquial/meme use originated in 4chan in 2016. That's why we call AI slop "AI slop", and not just "slop".

> But since if you're so lazy/undisciplined to disregard skilling up (Leetcode, Coursera, Udemy, Pluralsight, etc)

What's with the hostility in these comments? OP straight up said he just doesn't like the damn job. Grinding leetcode is not gonna change that. Is everyone just choosing to ignore this?

> But there is a difference between Google, which is arguably a mixed bag, and the AI companies, which are unquestionably cancer

Google's DeepMind has been at the forefront of AI research for the past 11+ years. Even before that, Google Brain was making incredible contributions to the field since 2011, only two years after the release of Go.

OpenAI was founded in response to Google's AI dominance. The transformer architecture is a Google invention. It's not an exaggeration to claim Google is one of the main contributors to the insanely fast-paced advancements of LLMs.

With all due respect, you need some insane mental gymnastics to claim AI companies are "unquestionably cancer" while an adtech/analytics borderline monopoly giant is merely a "mixed bag".

> It's in Thiel's nature to distance himself from Altman

This is way off base. Altman is basically a mentee of his, like Zuckerberg. They go way back, and they've both publicly praised each other quite recently. They're very close friends.

> Moonshot AI is such an overused generic name that I had to ask an LLM which company this is

I just googled "Moonshot AI" and got the information right away. Not sure what's confusing about it, the only other "Moonshot" I know of is Alphabet's Moonshot Factory.

> This company is far from "open source", it's had over $1B USD in funding.

Since when does open source mean you can't make any money? Mozilla has a total of $1.2B in assets. The company isn't open source nor claiming to be.

This model was released under a "modified MIT-license" [0]:

> Our only modification part is that, if the Software (or any derivative works thereof) is used for any of your commercial products or services that have more than 100 million monthly active users, or more than 20 million US dollars (or equivalent in other currencies) in monthly revenue, you shall prominently display "Kimi K2" on the user interface of such product or service.

Which sounds pretty fair to me.

[0] - https://huggingface.co/moonshotai/Kimi-K2-Thinking/blob/main...

>but after Deepseek stunned the world with the R1 model, subsequent models got heavily censored and languished in relative obscurity.

I'm pretty sure this isn't what happened. DeepSeek just hasn't released a big model upate. But in the meantime, Qwen, Bytedance, Ziphu and Moonshot AI have released extremely impressive models, some of which are SOTA or close to it. The open source/open weight world is still in love with Chinese labs as they keep releasing cool stuff and filling the void left by Meta and Mistral.

I think you had a kneejerk reaction from taking GP's comment too literally. Happens to the best of us. They're saying the LLM is simulating an anxious human and needlessly burning tokens ($$$) as a result, not that the LLM literally has anxiety and needs digital Xanax.

Include me in the screenshot.

Same! I guess HNers have the same tendency to hoard bookmarks so we all ran into the same pain points.

Just in case you didn't know, they raised the rate limit from ~50/week to ~50/day a while ago

> Whatever people say about AI, they're incapable of doing quality investigative reporting any time soon

the latest Gemini models have a very low hallucination rate in benchmarks [0]. Feel free to try it yourself. Just go to gemini.google.com, choose Deep Research, and ask it to write a report about a topic that you're intimately familiar with.

[0] - https://github.com/vectara/hallucination-leaderboard

I still don't get why .NET barely ever gets mentioned in these threads. Even new or niche frameworks like Phoenix, loco.rs and others get mentioned, but almost never .NET. It's as "convention over configuration" as it gets.

I haven't used RAG yet, but OpenAI's embeddings are pretty far behind AFAIK. Gemini has the SOTA in embeddings right now.

Check out the MTEB leaderboard:

https://huggingface.co/spaces/mteb/leaderboard

I'm Puerto Rican and this hurt to read because I wish the noise problem was limited to beaches. Multiple cases of noisy, inconsiderate neighbors (in different apartments and cities) drove me insane, to the point where I left and don't plan on ever coming back. The way it affected my sleep during my formative years probably did irreparable damage.

Feeling like you don't belong in your own country is maddening and difficult to accept.

Throughout all stages of my schooling from kindergarten to college, there were always artists, tinkerers and builders, but they were never the majority.

I can't think of any time in history, under any economic system, where humans who create tools and art were the majority. Most people just want to have families and enjoy life.

>Think about your browsing history over the last year. You’ve probably consumed an obscene amount of React components, maybe millions.

Yeah, most of which are rehashes of the same thing, and most of them on the same ~10 websites.

Been thinking about something like this for so long, but didn't want to bother with building a VSC extension. Thanks for sharing this!

I suspect it would be useful as a fast "context collector" for LLMs as well. Maybe framing it like that could catch HN's attention a bit more.

My (biological, born female) wife has androgenic alopecia and not even the only woman I know who suffers from it. It's not that rare in women, just less common than in men.

News of this exact research has been reposted a couple of times and previous threads weren't nearly as bad as this one. Something changed.

Half of these are literally an import and 2-5 lines away in .NET, Rails, Spring, etc.

Most of the others are a 2-3 npm installs away in the frontend, and a few are not used by extremely popular sites like archive.org, Wikipedia nor this very site.

The good news is Orpheus-3B just made Sesame essentially obsolete.