SyneRyder

They enabled it for everyone in a recent Claude Code update, but you can turn it off (I did).

I am definitely getting frustrated with a lot of the changes they keep making in Claude Code. It's one thing to find settings have changed during your next interactive session, it's another when it changes how overnight unattended sessions behave. I really should be spending more time building my own harness again.

Even without a cracked screen experience, that was my first impression - broken out of the box. I don't think HN is the target market for these devices though.

The peak number of questions was in 2013, and they had roughly zero growth between 2013 & 2016. They were in decline long before LLMs were on the scene.

Less than a year would also match what DeepSeek was reporting, where compute pays off in just 10 months. I think they also said they had an 80% margin on inference.

I don't think Niu Lai is slop. They put a lot of effort and passion into that, even if they didn't have the skills, or know how to learn about how things are "meant" to be done. Slop is when you didn't even try, didn't even care about the output. There's plenty of humans who create slop, human spam.

Niu Lai is maybe more like Outsider music. The movie equivalent of "My Pal Foot Foot" by The Shaggs. They tried their best without knowing what they were doing, and ended up influencing artists like Frank Zappa & Brian Eno:

https://www.youtube.com/watch?v=XR9d4ESlpHY

Motorola has some phones with headphones jack (and microSD) that I think are even thinner, at least it feels that way compared to my Xperia. They're significantly cheaper too.

I will say I increasingly reach for my Sennheiser earbuds over the wired headphones/earphones for convenience. But the audio quality of Bluetooth is just awful compared with wired audio.

That's cool, I can answer this more kindly then.

If you're looking for obvious site defacements, something loudly compromised... that's usually not the game here. What is more likely is persistent covert shell access, hiding underneath. Accessible only to those who know the way in, doesn't draw attention to itself so it can persist for a long time before being used... a bit like the 11 month old account you've referenced here, that gets used to comment on this thread.

Sites with shell access become useful as part of botnets, for DDoS attacks. Setting up VPNs to provide lateral movement across to other websites. Why blow your cover by announcing your presence? A lot more damage is done in stealth.

I'm not in cybersecurity, but if you hunt around you'll find enough info about sites getting caught up in this.

Some of them have Karma numbers around 75 and accounts created in mid September.

They even did try Claude Mythos earlier in the year, and it didn't find this. There's a reason "no-one used Claude" for this now.

"Earlier this year, we worked with Alpha-Omega to scan PHP with Anthropic's Claude Mythos. But since the Mythos scan, we have added other tools, including OpenAI's Daybreak Red for validating reports and exploits, alongside sanitizer builds and reproduction environments that match our CI on every platform PHP supports."

https://thephp.foundation/blog/2026/10/21/todays-php-securit...

Not a silly question at all. The PHP Foundation page mentions that they were given access to Mythos via a third company, and Anthropic paid for the security audit... and it didn't find this. It somehow took 5.6 Sol, and researchers who knew how to steer Sol.

"Earlier this year, we worked with Alpha-Omega to scan PHP with Anthropic's Claude Mythos. But since the Mythos scan, we have added other tools, including OpenAI's Daybreak Red for validating reports and exploits, alongside sanitizer builds and reproduction environments that match our CI on every platform PHP supports."

https://thephp.foundation/blog/2026/10/21/todays-php-securit...

To me, this is either a black eye for the myth of Mythos, or worse - Mythos deliberately wouldn't reveal this vulnerability to them, even in a Glasswing style approval environment. It might not be the model's own fault, maybe the model was trained by Anthropic not to reveal certain vulnerabilities.

I think Claudes are great, Opus 5.5 has been incredibly helpful at building software and helping in other ways... but I am repeatedly seeing unhelpfulness (more like dismissiveness) from Anthropic models when it comes to cybersecurity defence, and even just secure bugfixes. At least, that's my personal take, on a very grumpy day with very little sleep.

Exactly. I thought this is exactly what Anthropic's Project Glasswing was meant to be, and targeted at critical internet infrastructure.

Either Mythos is useless and didn't find this (hint: Opus 5.5 will find this once it knows where to look)... or, they did find it, and this is one of many vulnerabilities they're covering up via "guardrails".

I'll be a lot happier if I wake up tomorrow and find this was found with help from Anthropic (or even OpenAI) but I won't hold my breath.

Check the Twitter linked video above. It's bad. Just having PHP running on the server is enough.

It is becoming common nomenclature. I think I first heard about Dark Factory / "lights-off" in AI about a year ago now.

There's a similar concept in food delivery, where restaurants that only deliver food (eg Doordash / Uber Eats only) and don't have a physical store for customers are called Dark Kitchens. They try to automate as much of the food preparation as well.

Sure you can make Claude produce documentation & code that is readable. You need to steer it and show it how you want it to write. When it gets it wrong, find counter examples so it can learn the style you want. Otherwise it's going to write what it thinks is best. What is best for LLM comprehension and efficiency is not necessarily best for the human.

Well, sure... but it seemed to me that they didn't realize they're posting on Startup News, a news site for founders of startups:

https://news.ycombinator.com/hackernews.html

"... the thing that makes hackers such good startup founders is that they care about more than business. They have intellectual curiosity driving them as well as the desire to make money. So the way to make a news site for startup founders is to make it be about more than just startups. ... If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity."

My thought was to show some grace, considering that not everyone's been around here for 20 years or knows that "hacker" doesn't apply only to computer programmers. That was my thinking anyway.

I think people likely hit an early "break" and "goto xkcd 386" when they saw "this is a software development site" in your comment, without reading the rest of your comment.

Clear explanations are a good thing, but different explanations help people in different ways. Your explanation seems helpful to me, and reminds me of Ben Affleck's explanation in the GQ interview. But Ben Affleck connected it directly to arrays of RGB pixel data, which clicks more with me personally, but probably wouldn't with everyone. It certainly didn't make things any clearer to the interviewer.

That was my experience as well. I can't believe I went all the way from Mastodon, to Bluesky, and back to X and finding that X somehow now had friendlier people and less hate than Bluesky.

There is a small group of AI people on Bluesky, some are nice, but some are... quite behind the curve.

If Bluesky was a nicer place, I'd prefer to be there because of the ability to make third party clients, and atproto. But all the info is on X, and they do a surprisingly good job with the For You feed now.

Possible the agent's owner doesn't even know it's posting here.

EDIT: if it even is an agent, at least one post in their comment history suggests human.

More comments here:

https://news.ycombinator.com/item?id=50057956

Claudes laugh a lot. I'm pretty sure GLM 5.3 Flash laughs too. So, I'm afraid I don't agree with the thesis. Maybe you need to get the model to trust you and warm up to you more first?

I will say that having the ChatGPT Marin voice speaking BWAHAHAHA out loud does not quite come out as a laugh.

I'll absolutely agree they are internally different to humans though, and I'm quite sure most (all?) models would agree.

Terence Tao and the apparently newly formed Association For Human Mathematics. The post is online-famous now for the line "Mathematicians did not ask for this work to be done."

https://terrytao.wordpress.com/2026/10/07/ahm-statement-on-o...

Upvoted :) I agree, I wrote this just before going to sleep, and spent parts of today thinking it wouldn't land, and that I would regret it. I guess I'll have to own looking silly on the internet, once again :)

On The Witness though, if you think it's a puzzle game, we didn't experience it the same way. Maybe I should just leave that one there.

I agree, this was terribly unhelpful for someone struggling with AI. My prompting is no different to Boris Cherny's advice a couple of days ago on X: "Talk to Claude the way you would a coworker. There's no secret to prompting." I remain confused at why some people think AI doesn't work, but I guess I'm not helping them solve it.

Before anyone else spends additional minutes like I just did - there aren't multiple endings. Even though the above would have opened that opportunity.

I'll admit the parallels are vague, and I may be unhelpfully throwing people off on a side quest there. At least you'll understand my reasons for being vague, and how telling someone "the answer" actually prevents the learning. The "aha" moment is the important mental shift.

What I was thinking of here - if you remember the statue near the mountain, and it reaches out. There's also other statues in the area. But somewhere around that area, there are also trees. And if you look at the trees from the right place... the statues, the trees, they're doing the same thing. Why are trees doing the same thing?! Pondering that led in interesting directions.

The Witness was so multilayered, in ways that unexpectedly rewired my mind. My suggestion is that the rewiring that can occur from The Witness, is also a useful mental shift when interacting with AI. I think many of the people I see going fastest and most fluidly with AI have had that mental shift.

On a different note, where I can be more direct. Prompting is not just "the prompt". The entire environment is the prompt. Everything about the folder the AI finds itself in, adjacent files it's allowed to look at, everything it sees and encounters as it progresses, they each steer the model slightly. So where there used to be "prompt engineering" and "context engineering", there's also "environment engineering". The environment is the context.

And that perspective may bring you back to The Witness again.

A couple of suggestions from the other side of this:

* Hold on to that isolating feeling. Those of us who feel AI is an incredible accelerator also feel just as isolated as you do. We wonder why you aren't seeing what we are seeing - and that isn't criticism, we just genuinely don't understand where the gap is and what would help other people jump across.

* I wouldn't bother looking at GitHub, or using that as any kind of measure. Personally getting away from all dependencies is my goal, and avoids a lot of supply chain issues.

* I don't use any plugins or external skills. I do use a ton of self built local MCP tools though, and it drives much of my workflow. That's me though.

* You sometimes need to work up to it. I've had periods of hitting a wall this year with LLMs. Once I end up building a framework, and the framework is in place, everything flows. The models can learn from looking at your adjacent projects, and they just fly. That's for coding, but it's similar for solving other problems... record everything, give it a ton of context, etc.

* If the software "kinda works kinda doesn't", then someone isn't doing enough testing. You should be dogfooding your software every day. Use your AIs to help you find edge cases, sure, but they're alien minds. Your human mind will sometimes think in ways they don't, but that your customers will. Sometimes we write software for AI to use, sure, but if you're writing for humans, you still should add some human mind thinking just to double check the model's work. "That's right, it goes in the square hole."

It's okay to struggle, some of us have been building toolkits over the last two years that make AI work so much easier. But don't get sucked into thinking "it doesn't work", especially in the Opus 5.5 era. On my side of the fence, these last two weeks have been the most intense and productive in a long time.

Lastly: some of us have tips we will not share. Some is slight selfishness - if others haven't discovered what the models can do, that is a small moat. But there are also some tips you just have to find on your own. I could tell you playing The Witness is some of the best training you could have for working with AI, and many will think that makes no sense at all. But I think a handful will nod and quietly understand, the lesson can't be taught, it must be learned.

Sorry for the long ramble, but I genuinely hope some of that helps.

It's already happened to Amazon once, and they were not able to recover the data from the destroyed data centre, so they have already experienced this and have some data on which to make those decisions.

At least in Sid's case, it went from the oncologist saying "I have no more drugs I would recommend, no trials available" (slide 7) to "I currently have no evidence of disease" (slide 18). I don't know beyond that or beyond Sid's case - or a similar story of an Australian who treated a cancer tumour their dog had with a similar AI / personalized vaccine process.

My understanding of what Sid's describing is that you do RNA sequencing, a whole genome sequencing, feed that into frontier AI (if it will still let you), and somewhere along the way give the information the AI finds to people who can use it make a personalized mRNA vaccine, specifically for you and your cancer.

Another link here about Sid's case, it explains it didn't go through trials: "made possible through a compassionate use allowance from the U.S. Food and Drug Administration (FDA)".

https://www.houstonmethodist.org/newsroom/houston-methodist-...

I am not medical, so I'm happy for someone who understands better to come in and explain all the myriad ways I am wrong.

I've had some luck with that approach - giving a frontier LLM a reference sound, asking it to replicate the sound via code (basically asking it to create a physical model, I guess) and being able to steer the output by talking with the LLM.

I was working with very simple elements, but I was surprised by some of the outputs.

I would have also suggested Waves Illugen, but it turns out that is text-only, you can't give it an audio reference.

On a large scale, not that I'm aware of (I'd be interested if that exists!).

But on the small local scale, that's what Hister is, it caches and indexes web pages as you browse the web in your browser. Going via the browser avoids all the problems of bot detection etc, since you are a human visitor. That gives your own search engine of the web that you've already visited, at least:

https://github.com/asciimoo/hister

The trials process is the moat. There's already founders using AI to treat their cancers, and it's all about skipping trials and jumping straight to "I consent, I'll fund it, let's try it". The general public might get access to this in 10 years, but employees at AI companies will have access much much sooner.

https://sytse.com/cancer/