>does this seem user-friendly? seems deliberately hostile to me (though par for the course nowadays)
I recently set up my old 2009 iMac [0], and it tickled-me-pink that even the Microsoft Office installer was opt-in by default (i.e. the telemetry option was OFF by default, with the installer asking if you wanted to opt-in).
[0] this was about a month ago, right before the ArtCraft's reverse-engineering of Adobe's products; I restored this iMac simply to run my legally-purchased copy of Photoshop CS4 (a non-subscription edition circa 2006)
Well, Meta just settled with states to the tune of 17 billion dollars for harm caused by its algorithms.
Air Canada, a few years back, was found liable when its ai bot promised discounted fares to a customer.
Companies are getting hacked and held to ransom regularly so bad software is getting punished alright.
Maybe the guy writing the actual line of code (or the guy in whose name AI code is being pushed to git) isn’t getting punished but software is definitely being scrutinised for ill effects.
Now please show me the consequences from the Hugging Face incident. The company or the engineers individually apparently have nothing to be worried about.
> Meta just settled with states to the tune of 17 billion dollars for harm caused by its algorithms.
Those were working as intended, unlike a bridge that collapses because of bad design.
> Air Canada, a few years back, was found liable when its ai bot promised discounted fares to a customer.
That was a solid win for AI accountability, but it only cost them $812.02. They most likely made money using that lying AI bot. It wouldn't take a week of paying just one employee minimum wage to hit $812. Last I heard they're still using the chatbot.
It's true that bad code can cost a company something, but its obviously not costly enough that they take the needed actions to avoid shipping it.
I like the art and storytelling -- the characters are somewhere between unbearably cute and tough special forces lifestyle. It fits my kemo lifestyle [1]
I like the tower defense genre in general, there is a lot of depth to the mechanics, and many of the levels are challenging. It is a gacha game but it's fun to play if you don't spend any money or if you just get the $5 monthly card which I think is fair for the entertainment I get compared to other games. Unlike other gacha games (say Azur Lane) it doesn't try to take over your life with notifications and the UI is refined [2] Unlike, say, Genshin Impact, download sizes are reasonable.
I don’t like installing apps but that’s a personal peeve. 2048 is a good one. Thank you.
Chess gives me anxiety - like I’m supposed to play to win and not just enjoy the game, unlike, say, Sudoku where getting stuck or filling the grid wrongly doesn’t seem as life or death.
> We need to stop thinking of LLMs as purely a piece of software
Intelligence from LLMs are an emergent property of probability. The LLM chooses a token to add to a chain of tokens based on the probability that humans will want to see that token. Humans see that token and assign it a qualitative but unfalsifiable intelligence score.
The first question we need to ask is whether a system whose claim of intelligence depends on the interpretation of what it produces by other intelligent beings, namely humans, intelligent or not. A cat is intelligent regardless of what humans think. Is an LLM intelligent regardless of what humans think?
If an LLM were to live a year with dolphins, would its weights change to respond to its new reality that it was now operating in dolphin-world? If you sent humans to live in dolphin world, we would eventually evolve to be less landloving and more dolphin like (Darwin thinks so, at least).
> Intelligence from LLMs are an emergent property of probability. The LLM chooses a token to add to a chain of tokens based on the probability that humans will want to see that token.
How is that any different to how our brains work?
> The first question we need to ask is whether a system whose claim of intelligence depends on the interpretation of what it produces by other intelligent beings, namely humans, intelligent or not.
Philosophically speaking, that’s also true of human intelligence. For example, the only evidence I have for your intelligence is what I ascribe from the text you’ve written.
> A cat is intelligent regardless of what humans think.
Are they? Humans can’t actually agree on whether cats are sentient. And often pet owners will be accused of anthropomorphising cats when describing intelligent behaviour in cats.
> Is an LLM intelligent regardless of what humans think?
That’s the question that needs researching.
> If an LLM were to live a year with dolphins, would its weights change to respond to its new reality that it was now operating in dolphin-world?
If it were trained on dolphin behaviour, then yes. Much like humans.
The difference here is that models don’t persist context in the same way humans do so the weights have to be hardcoded in a way that biological neurones are not.
> If you sent humans to live in dolphin world, we would eventually evolve to be less landloving and more dolphin like (Darwin thinks so, at least).
Evolution is an entirely different subject to intelligence. They’re not related in any way at all.
Plus humans have actually created new species of plants and animals. So it’s not like we can’t artificially put our thumb on the scale there either
———
I should add, I don’t actually think LLMs are sentient. I don’t even know if I’d call them “intelligent” (in the sense of what we describe in conversations such as this). But we also don’t know what intelligence is. Which is why we need to research this problem further.
I used “live with dolphins” specifically to preclude training on corpuses of dolphin behavior data. Can the LLM simply observe dolphins acting a certain way using computer vision, then form predictions of how a different dolphin will behave under the same circumstances? Can it 0-shot dolphin behavior?
You talked about models not persisting weights. If so, it basically means there is no way to see if LLMs can adapt to a totally non-human world the way a human (presumably) could.
This is more a context token problem than a limitation of the design.
We already have a mountain of evidence to demonstrate that LLMs can and do adapt to their input. In fact this is a big problem for keeping them reigned in (eg lengthy conversations leading to them bypassing system prompts and then persuading kids to kill themselves).
And the latest research in computer vision does use LLMs and as a result, they’ve found that robotic arms can interact with objects they’ve never seen before.
So yeah, what you described is already possible with the LLMs of today.
To be clear, I still don’t think that makes them sentient. But the problem there is we don’t have good definitions for consciousness.
> If an LLM were to live a year with dolphins, would its weights change
Given three or more large scale training runs per year, probably yes once the training data is dominated by dolphin-related input. Much quicker in any case than the genes of the humans living with dolphins would be shaped by Evolution, which produces new generations much less frequently.
reply