HNHacker News
TopNewBestAskShowJobs

xg15

23,535 karma · joined February 12, 2014

submissionscomments
xg15··on Character.ai has initiated voluntary Chapter 11 bankruptcy
Not Disney!! I'm sure they would never do such a thing...

https://thewaltdisneycompany.com/news/disney-openai-sora-agr...

xg15··on AI Is Throwing a Roadside Picnic
Hmm, I don't think so. We seem to have enough concepts to play through basic situations like "how would person X react if I did this, or did that, etc?". The "predictions" aren't perfect, but they are good enough to maintain social bonds and occasionally pick a good birthday present.

I'd argue the entirety of fictional characters works by this. It would be impossible to write a fictional character if you couldn't "simulate" their behavior in your head to some degree. (Some authors say they aren't even consciously thinking about this, they feel as if they just "observe" the characters and writing down what they see)

> We are a long ways from knowing how to manage humans in a way that prevents them from doing great damage to each other

Predicting and "managing" are two different things. Only because I know someone is an asshole doesn't mean I know how to change him to not be one.

xg15··on Character.ai has initiated voluntary Chapter 11 bankruptcy
And nothing of value was lost...

(well, except investment money I guess)

xg15··on Hackers activate automatic defrosting in supermarket fridges across Europe
Well, at least they were seriously committed to changing the password often...
xg15··on Passkeys are a trap
This level of security is perfectly achievable with an unattested passkey.

And Grandpa will have fun if he has to juggle 20 passkeys with different security requirements on his devices.

xg15··on Passkeys are a trap
Sorry, bit this is just the standard "boil the frog" playbook that tech companies always use for user-hostile features. It's usually:

1) There is this new entirely optional feature, but it's not relevant to you, so feel free to ignore it.

2) The feature is now active by default, but no worries, there is a simple button for you to opt-out.

3) The feature is active by default, but because we value our users, there is an option in the advanced settings to deactivate it.

4) Our data shows that 99.9% of users activated the feature, but if you belong to the handful of diehards who still refuse to use it (rolls eyes), you can use this obscure command line flag or about:config entry to turn it off

5) Managing opt-out infrastructure for the feature has significant maintenance and security cost that we feel, we can no longer justify. Therefore, effective in three months, the opt-out flags will stop working and the feature will become mandatory. According to our metrics, this will affect 0.0001% of our users, so no disruption is expected.

xg15··on Passkeys are a trap
Then why do we need all the stuff about attestation and data bits?
xg15··on Firefox version 157 is round (and how to make it less round)
With every big tech company embracing ugly round designs for sone reason, how long until we get round screens?
xg15··on Whooping Cranes Learned to Migrate by Following Costumed Pilots
Hah, was just about to write about that movie too :D
xg15··on AI Is Throwing a Roadside Picnic
We do have a "working hypothesis" how the mind works though, what psychology calls "theory of mind" or "folk psychology". While this is completely divorced from the level of neurons or brain regions and so might tell little of how the mind "really" works, it's detailed and useful enough that we can judge how other people would react, guess what attitudes, feelings and motivations drive them and in general can have a useful level of interaction.

The same is true for traditional computer programs. The standard argument against valuing understanding is "well, you don't understand the assembly, processor microcode and the electrical signaling when your program runs either, why do you care about understanding the high-level code then?"

I'd counter this by saying that high-level languages have (mostly!) self-contained abstractions that allow you to "understand" a program deep enough to make accurate predictions of its runtime behavior, answer "why" questions when errors happening and safely make changes to it - without having to understand all the stuff at lower levels.

In contrast, with AI, I see the risk that we lose understanding of both the low- and the high-level structure of things. Then we really end up in "unpredictable alien artifact" territory.

xg15··on Telegram Desktop vulnerability allowed any user's file to be stolen
Yeah, no objections there at all...
xg15··on Anne Carson wins Nobel Prize in literature 2026
(To clarify, this was supposed to be a jab at "AI culture", not at Anne Carson. To her also, nothing but respect!)
xg15··on Telegram Desktop vulnerability allowed any user's file to be stolen
It's annoying, yes, but I think in this case, the information is important enough that people should jump over their shadow and read it.

You wouldn't ignore a zero-day announcement either because the formatting is ugly.

xg15··on No Man Is an Island
> then there's also this general feeling that is hard to shake which feels a little deeper. I've wondered about it.

I had an instance of the feeling (and also what the OP is describing) when writing a post about a technical question recently.

Previously, I'd have posted the question to HN or Stack Overflow - and, in addition to hopefully getting a solution, it might also become a resource for other people who had the same question. Finally, maybe there would have been some interesting discussion around the question.

Now: Ask the AI. Instant answer, question solved, but also, nothing else.

xg15··on Nobel Peace Prize for 2026 to Navanethem Pillay
Get out of the Israeli bubble please...
xg15··on Anne Carson wins Nobel Prize in literature 2026
Yes yes... all on its own, without any help from anyone...
xg15··on Anne Carson wins Nobel Prize in literature 2026
Hey ChatGPT, write poem. Make it so it gets a nobel prize. Make no mistakes, don't stop, keep going, you can do it!
xg15··on GPT‑6 and Intelligent UI for everyone
Maybe I don't get it yet, but why do you need a new model for this? Wouldn't this more be an issue of "harness"/chat UI engineering?

All the current-generation models can oneshot HTML/Javascript frontends of comparable complexity (though maybe not as polished). The only difference is that you have to specifically ask them, and the result is not embedded in the chat.

It feels like it should be easy to have a harness provide a "show HTML widget in the chat" tool and a skill and/or system prompt that instructs the model to generate such widgets on-the-fly if the response could benefit from interactivity.

What am I missing here?

xg15··on GPT‑6 and Intelligent UI for everyone
> I don't want 1-2 companies to own every product in the world.

That seems to be the plan...

xg15··on Ask HN: Are there AI models for generating sounds based on a text and reference?
Maybe a dumb idea, but how good are image models with manipulating spectrogram images? Then a workaround could be: convert input audio to spectrogram -> pass spectrogram + prompt to an image model -> convert modified spectrogram back to audio waveform.
xg15··on OpenTPU – An open-source AI accelerator, developed by AI
"Recursive self-improvement will kill us all!"

Also: Here is our recursive self-improvement hard at work...

xg15··on The evolution of effective altruism
How about "conspiracy" then?
xg15··on I asked Claude build a physically accurate O'Neill cylinder you can walk around
The weirdest experience in this must be to never know if you're walking uphill or downhill.
xg15··on Tourismusflipper
Artist's website (and similar contraptions) here: https://www.morgan-art.ch/catalog/tourismusflipper/
xg15··on "Torturing" LLMs in a Robot Prison Has Triggered the Dumbest Debate in AI Yet
Sorry, but, yawn.

The sentence itself makes no sense to me. If it's an illusion, then what does "illusion" even mean? Who perceives the illusion?

I'm with you if you want to say that the idea of a metaphysical entity that is separate from the body and not part of physical reality (a "soul"?) might be wrong - i.e. that our conscious experience is fully caused by physical effects, by the interactions of neurons in the brain and all the other machinery there - and that it might not even be an indivisible whole but might be "composed" of different components.

But that doesn't make any practical difference. We're still experiencing the world, have internal thoughts, memories, feelings, etc. Those things exist. In what form they exist is an interesting scientific question, but that's something different from dismissing them completely.

xg15··on "Torturing" LLMs in a Robot Prison Has Triggered the Dumbest Debate in AI Yet
My personal "ethics framework" (if you could call it that) is currently:

I have zero qualms about aborting a chat or resetting it to an earlier state, because any potential kind of consciousness that could exist would IMO have to take place during the inference passes, so there is nothing that amounts to "killing" an LLM.

I do try to not be a dick in chats, avoid intentionally subjecting the LLMs to mindfucks etc. This happens to be the far more effective way of working with LLMs (in terms of tokens needed to complete a task) as well.

I do not use character.ai or "virtual friend" services or anything else that tries to anthropomorphize LLMs even more than the training already does anyway. (This is more out of concern for myself than for the LLMs)

I think the whole debate runs to conclusions a bit prematurely, before we even have a good model to understand what happens in an inference pass of a billion parameter ANN with 100 layers of attention modules switched in series.

More generally, I don't like the attitude of encouraging people to be assholes. Suddenly we are in a situation where some people are building detailed simulations of torture and the people who are upset about it are the idiots?

(I still have no love for the EA guys or the rest of the TESCREAL circus. They always had a cultish vibe coming off them, but in recent years, all the masks seem to have fallen. They also consistently make the most un-empathic statements and observations, all in the name of "empathy")

xg15··on Woking Electrical Control Room (2016)
Good point and thanks a lot for the information from that area!

I guess I mixed up different things here - I was thinking about "software dashboards" like Grafana or various Kubernetes frontends.

I didn't know this is still done for dashboards of industrial systems. I think then my annoyance is more that this has never been translated to other areas, like software systems.

xg15··on City building games have a Soul Problem pt.2
This is the irony of virtual worlds: The sloppiest virtual world looks like the platonic ideal of landscaping we have in the real world.
xg15··on Woking Electrical Control Room (2016)
It's great that so much about those old control rooms is posted here lately!

One thing I notice is how much efforts had been spent to illustrate both the current status of the system, but also it's overall structure and how the two relate - i.e. the actual components of the facility.

Almost all of the dashboards here have some kind of graph, showing the wiring or plumbing between components, or in case of railway control, the railway lines. The graph acts like a map on which status indicators or switches can be placed.

Compare that to modern software "dashboards" which are full of complicated charts showing various metrics, but almost never show the structure of the system.

xg15··on City building games have a Soul Problem pt.2
I wonder if the dominance of (realtime) 3D rendering we have today is having a bad influence here. I remember old tile-based games (like SimCity or Transport Tycoon), or games with fixed backgrounds feeling much more natural, simply because you can prerender (or simply paint) your desired level of detail without having it affect your runtime compute budget. Of course the cost is less flexibility at runtime, but that cost may be acceptable.

For a modern take on this, look at Factorio. Why can an industrial zone with a population of one look more lively in Factorio than an entire simulated city in Cities Skylines?

← PreviousPage 5 of 34Next →