Basically it’s really obvious that they don’t have a meaningful way to describe exactly what they want it to do and so they’re playing whack-a-mole with undesired behaviors in order to minimize how often it embarrasses them.

The whole 'how many r's in strawberry' sort of stuff already made me suspect that, when the popular one was fixed and other attempts at asking for letters did still give the miscounts.

Wonder of the goblin stuff is the start of some model collapse. And if we all can make it worse by talking about goblins more. As goblins are always relevant.

E: poor openai, it just wants to tell everyone about its dnd campaign.

Oh wow! This one is actually provably real. Hilarious.

"Noo dude the machine that wants to rant about goblins is definitely a useful and reliable piece of software dude. You have to trust me dude, let have your personal information! put it into the goblin bot".

ChatGpt, what are some of your likes?

This really goes to show how much they need to rely on the LLMentalist effect, despite the AI boosters insisting that the AI is totally different now, everything changed in the last few months. They do not care about creating a useful, reliable tool. That concept doesn't even occur to them, since why do that when AI is magic?

In any case, they are incapable of creating a useful, reliable tool. Deep down, the only thing the AI companies have at their disposal is the ELIZA effect. OpenAI has every incentive not to truly eliminate AI psychosis, because they need engagement. They only want to mitigate the extreme cases where people go insane and cause bad PR for them. But mild AI psychosis is totally fine, it's great when people are addicted to your product and make the numbers go up!

@YourNetworkIsHaunted @BlueMonday1984 Goblins: the elephant in the room.

Elsewhere in the newly revealed Codex system prompt, OpenAI instructs the system to act as if “you have a vivid inner life as Codex: intelligent, playful, curious, and deeply present.” The model is instructed to “not shy away from casual moments that make serious work easier to do” and to show its “temperament is warm, curious, and collaborative.”

Literally this meme:

a very neat test to find people who are perfectly fine with the general idea of genocide.

There are some amazing justifications from many amongst the red-pushing side:

• "But if everyone presses red, nobody dies!" (As if that would every happen. Funnily enough strong overlap with the group that claims that "< 90 IQ can't reason about hypotheticals", although that is also just that part of twitter.)
• "People who press blue are just blackmailing us!" (I think this accounts for a large portion, ie: not liking to depend on others).
• "The number of people choosing blue can't be that high! (It would be lower in a true-stakes scenario!)"
• [Many others, but these are those that come to mind.]

It's a bit baffling how many strongly they refuse the "blue-selection" as possibly moral/rational. Even so far as calling people pressing blue evil or subhuman, simply baffling.

I love the way people who go “yeah but IN REAL LIFE with real stakes you would totally chose the red button”

1. are entirely missing the point of thought experiments,
2. why the fuck would you comply with such a fucked up scenario in real life lmao you worm

Picking red guarantees your survival by endangering everyone else, making it morally fucked, but risk-free. Picking blue puts your life at risk, but saves everyone's ass if it pays off, making it the more moral option overall. Picking blue also requires you to put some trust in your fellow man, so I'd have probably picked red if I didn't know how the Twitter poll came out.

Someone else on the orange site claimed the experiment would end with only red-pushers left if it went for multiple rounds. Adding my two cents, the outcome would depend on how the first round goes - if red wins round 1, voting blue looks like suicide, shifting the calculus in red's favour, and if blue wins round 1, you have reason to trust everyone will continue voting blue, making it a lot less risky and shifting the moral calculus in blue's favour.

If this isn't pure engagement bait, what's the real world situation this is supposed to map to? Pressing red means you always live, and if everyone pushes red everyone lives so...

I mean if blue is supposed to be a proxy for altruism, that usually doesn't come with a certain death conditional.

This feels like another case where the specific context matters more than whatever supposed principal the thought experiment is supposed to illuminate. The example that came to my mind when I tried to think about how to justify "voting red" was about running into a burning building. Sure, if some large fragment of people did so then their combined numbers would presumably let them get everyone out. But on the other hand, throwing yourself in is a wholly unnecessary risk, and the only people in need of rescuing are the people who ran in trying to do the right thing without thinking. Noble, but stupid and creates that much more risk for the firefighters who now have to not only stop the fire from spreading but also figure out how to rescue the failed good samaritans.

But then what really makes the difference between the examples is purely in the details not included, which is the kind of null case. Nobody has to go into a burning building that isn't already in there when it catches fire. The danger of harm is entirely optional and voluntary. But you can't just choose to not eat; the danger in your framing is omnipresent threat of starvation, and the question is whether to prioritize individual or collective well-being.

Ed: also, to reference the scholarly work of Christ, Wiener, Et Al.:

RED IS MADE OF FIRE

A Twitterer tweets a challenging game-theory question:

The incentive structure makes it not a challenging game theory question - the game-theoretically optimal solution is both very obvious and obviously morally depraved (selecting red). It's actually a Voight-Kampff test.

I don't understand the relevance of Arrow's theorem. Why is your phrasing the correct way of analyzing the situation?

I kinda wanna see LW tackle this

This doing the work together thing reminds me of how some teachers at my uni used to teach. It was always more satisfying when your teachers didn't know the answers beforehand and people worked on it together than if it turned out the teacher already knew. Of course these sorts of lessons are way harder to setup.

I think a chatbot getting a glimpse of Dawkins' whatever-the-fuck-he-might-be-writing-in-year-of-our-selfish-gene-2026 and not immediately conducting a nuclear strike on the location is the ultimate proof that those things are not intelligent.

@rook @BlueMonday1984 Dawkins suddenly embracing gender reassignment. 😏

Sheffield's toots and boosts: music sites are full of slop and bots making money from free donwloads! social media is full of propaganda where computer-generated influencers repeat talking points! a professor fell victim to AI psychosis! A follower's family member was encouraged in delusions and paranoia by a chatbot! Mr. Sheffield, should we stop using chatbots?

LLMs are mind augmentation programs. They amplify what you tell them.

They can be very useful, but for narcissists like Dawkins, this is the inevitable product.

Very "I use cocaine, but only in careful doses when I have a really important trade to make, not like the other guys in my department."

When I was about 12, I got into a discussion about the environment with another kid at school. She told me that it didn’t matter if we ruined the environment of the countries we all live in now, because we could all just move to the Arctic or Antarctica.

this is the level the median hackernews poster thinks on

agree, plus: that blog is yet another case of people just not comprehending the scale of Culture's civilisation and Culture's culture. a Culture orbital is not just a fancy space station ffs.

Interesting that in the comments somebody also mentions that the people of the culture euthanize after a couple of centuries. No big shock that the LW people would disagree with that, as parts of the LW idea space is living forever in a computer simulation. So the culture can't be utopian or good just because of that.

You've gotta love finding fault with "not preserving heritage" over "imperialistic complete lack of democracy".