animuchan

I see the point, but nobody in their right mind would call a mere text message "please delete your work" to be malware, much like telling someone "please die" is very very different from attempted manslaughter.

> Truly feels like the UI/UX is done by people

To me it feels like it's done entirely by an LLM, starting from the product vision.

This is so beautifully cursed, reusing the module loader state as your local state. We could have the familiar Python syntax of

`from <key> import <value>`

And a custom import hook eating the error. To get value(s) for a given key, naturally we'd scan the module loader cache. Elegant.

"The Good Parts"

This is infuriating in certain video games, like the StarCraft II single-player campaign: you're in the middle of 4 different things, your base is being slowly nibbled on by a few run-by zerglings, need to expand, babysit workers, creep / pylons / what have you. Then the game decides to show you something. It grabs your vision cone and forces it where it wants you to look -- slowly.

This specific action -- taking the thing away from the user, while they're actively engaging with it -- should be punishable by sacrificing the designer to Satan straight up. A bit harsh but justified.

It's a bit insane, but the cost of switching to regular NodeJS is low (for all but most bun-specific projects).

All valid points though, I'm pessimistic about Anthropic still actively diverting resources to these side quests when tough times hit (which might be in a week for all we know).

I think there's like the meta-bad AI usage idea, in which all other tarpit ideas take root.

"Let's outsource critical decision making to a language model."

Small scale of this can be seen when attempting to single-shot a whole SaaS by the means of a severely underspecified prompt. This is the first-order blind reliance situation: our SaaS sure is implemented in some way.

Really fiery tarpits put the LLM higher in the stack: let the language model decide what the feature should be. Our SaaS sure does something in some way.

Having a human in the loop at all times doesn't scale. Not having a human scales, but doesn't work.

That'd be crazy. The agent has a skill configured to fetch ticket descriptions from Jira by itself. Copy-pasting feels like manual labor.

Totally agree!

Bonus level "I have a hammer, all I see is nails": using Claude Code for random non-coding work, like dataset cleaning. It's really convenient to have a script spawning Haikus via `claude` CLI and feeding them prompts and JSON files. Money burn potential: practically unbounded, but also it's real work that the product people wanted done, so of course it has a cost associated with it. I'd be bewildered if anyone complained.

Counterpoint: let's say we connect a speaker to the HTTP server, and every time there's a request, the speaker produces a click. This setup will make audible sound. If it's OK to measure this sound in Hz, then is't OK to measure the HTTP requests in Hz, because they're explicitly === sound in this case.

What I see in the workplace is, people specifically outsource decisions to LLM. It tries to flag and explain all sorts of landmines, it really sometimes does, but the prompt is "make it work" and "be relentless", and the operator is barely even looking at the (conversational) output of the LLM, just the code (or other file) they asked for.

This is another difference to a largely organic developer: the ability to refuse a massively damaging or stupid task.

The next generation of the user is what I think will be the factor for AI-assisted loss of life, moreso than the models themselves. It's the trust we put in them.

The dinosaurs of the before-LLM age will increasingly struggle to convince young people e.g. to seek competent medical consultation, because ChatGPT can do that much better (it's always supportive, available at any hour, meets you where you are...) -- this aspect alone is bound to have a death toll associated with it.

Or imagine asking ChatGPT to map an Alpine climbing route, -- anything with a bad failure mode.

This is an entirely false dichotomy though, is it not? One can both know facts and understand logic behind them, it's not like you're creating an RPG character and need to make a choice with limited character points.

(Can't say time is the limiting factor either -- we're both in HN comments, valuing our own time at zero.)

Good thing casino apps remain up — otherwise the teenager in me would become bored with such a smartphone.

I despise how this incredibly user-hostile move is spun in the title: "Google just gave Android power users a huge sideloading win", as if it was a good thing that Google did for some portion of its users. That's such a blatant, incredibly damaging lie, on all levels, that it's probably called journalism at this point.

A fairly narrow buckets, sure.

I think the point of Suno is to make you not search for your specific thing though, and instead produce your own. Searching for niche music has always been a thing. If our goal is to listen for free, we don't care about Suno (or any other way to make music) one bit, it's just another DAW for those making music.

And AI music in general sure has its fans, check out Only Fire for example.

Almost nobody listens to others' songs on Suno, that's the entire point.

You wouldn't care to order the food as I personally like it -- might be too spicy (or too bland) for your taste.

Suno songs are overtuned for personal preference in the same way.

So true. AI music gens like Suno can't do Paul Shapera works even remotely, but can recreate a lot of pop or EDM music very faithfully. There's just no distance to close, it's already mainstreamly bad.

Respectfully, it's not, in most parts of the world.

If the flow is just typing prompt after prompt into AI, I guess just ignore that it's running and keep typing? Or open a new tab and keep typing. Why does Claude even break the typewriter flow thing that's going for you?

Different person's POV here: I fully expect from my mind to enjoy video games.

Then I open Steam, and feel the strong sense of rejection, like eww this is well and truly a disgusting selection of repackaged Unity assets, and I'd like to have absolutely none of that please.

It's this cognitive dissonance maybe, that's the guilty displeasure. Because I do keep coming back, we all do.

In the grim darkness of the right now, with former gaming companies focusing on live services and loot boxes and advertising, core gamers' very conception of video games is simulacra, a copy without the original.

With time, this will fade; like when searching for Jesus, you don't really expect to physically find the dude hidden in the corner of the church. Expectations decay.

Maybe I will visit Steam on the Internet Archive and still browse, when everything is gone and dead, and the guilty displeasure feeling will have outlived the displeasure, and the guilt.

Staying sane is a gradient.

But personally the current war gives hope. Ayatollah was Putin's serious leg to stand on, not just in the region. If Iran's new govt is normalcy-adjacent, not just the terrorist orgs suffer without support, but also the Putin's war thing collapses. A decade (or less) of peace and prosperity is upon some of us.

And AI is neat, it highlights so many things about people. We start to value agency in fellow human beings. People understand context better now, and give more context proactively when talking; but also understand that there's context poisoning, which is why derailing is bad.

In the end everything is a tool: AI is a tool, war is a tool. We think AI is for coding, and war is for power, but in the end AI makes us know ourselves better. War makes us know ourselves better (and also gave us fast airplanes). This is the important, this is what will remain -- when the kingdom is ash, and the echo still drums.

Another good analogy would be, said theocracy is (was?) like a very bad piece of legacy code, impossible to refactor, until the entire feature gets thrown in the trash.

The need for killswitch I think is self-inflicted.

* Mozilla has a track record of forcing unwanted changes on its users. What with Pocket, data collection and telemetry defaults, sponsored links throughout the UI, all the good stuff.

* The enduring users are more likely to want to revert any Mozilla default the moment it's introduced. (This is why Firefox has disproportionately many projects to un-Mozilla the thing: Arkenfox, BetterFox, LibreWolf, Waterfox...)

This is from the annoying (sure hope so!) sporadic Firefox user who was actually pleased by the news. Honestly, I saw it and though: wow, Mozilla giving the tiniest part of control back to the user, that's actually good! Short-lived as the excitement was, in these fading moments of Firefox I'd like to see more of this and less of the user-hostile thing please.

Most people aren't human then, sad.

The comment that started this thread is as follows:

> Too bad the Zen of Reticulum is against freedom. Specifically freedom 0: the freedom to use the software for any purpose. Its restrictions preventing it "from being used in systems designed to harm humans" [...]

So if you're a lawful good human-harming person, you are prohibited from using Reticulum by this Zen document.

If, however, you're an unlawful kind of human-harming person, then you likely don't feel the pressure from the Zen document to stop using Reticulum.

(And surely you can't flip the logic like you did right there: "engineers are implicitly allowed to use X" doesn't convert to "users of X are implicitly engineers".)

So no, I'm saying that of all users intending to cause humans harm with this tech, only criminals are permitted (or at least, aren't meaningfully restricted) to use Reticulum. Only in this niche case.

Thank you very much! I also feel that the impact of software licensing on violent groups behavior might be low.

It is, however, interesting on principle, since it only allows the use by criminals (implicitly), and not by law enforcement. By then making the tool very impractical to use, we can punish bad actors still.

(I think there was a honeypot operation to this effect, something with feds making up a "secure encrypted phone" and then acquiring Cartels as a major customer.)

(On the off chance I just burned this very similar operation: dear feds, I'm so sorry!)

The OP doesn't even say "Totally not going to do it", merely "it feels off-mission", so a vibe check away from doing it.

Yep, and that's how he arrived at the $number. If a small number of people were using ad blockers, the cited sum would approach $0 since disabling ad blockers would affect very few page views, right?

Oh I know this bug! Happens with their own Bixby assistant too.

(Either Samsung dropped the ball on quality in the last 5-10 years, or I just started to pay attention, but the desire to throw this garbage in the bin is real.)