I’m not quite sure what you mean. If you took a picture of an orange and sampled a pixel from it, it wouldn’t look like an orange (fruit) any more, but it would look orange (color). Likewise sampling a pixel from a picture of a piece of gold wouldn’t look like gold (metal), but it would still be gold (color).
Gold is – it’s an orangeish yellow. I think what you’re meaning is that when people say “gold”, they usually are referring to the material properties of the metal as well. But the actual color does have a spectral hue. Magenta on the other hand, (including shades of pink that fall under magenta) is not a spectral color, and is just how our brain interprets the combination of signals from our red and blue cones.
To be fair, it’s the same for humans.
Yes, you can run ollama via termux.
Gemma 3 4b is probably a good model to use. 1b if you can’t run it or it’s too slow.
I wouldn’t rely on it for therapy though. Maybe it could be useful as a tool, but LLMs are not people, and they’re not even really intelligent, which I think is necessary for therapy.
And nothing is okay for people who are just using it for web browsing and streaming.
I want local music, and to be able to take pictures without worrying about storage, etc. so ~20GB isn’t enough for me, but for some people it really is fine.
My guess is $30, could be $40 although I doubt it after the response they got.
Any general purpose consumer device should probably have 64GB or more.
But I don’t see the point in disallowing <32GB, as that can still be enough for using tablets for lots of uses like e-readers, smart home displays, kiosks, etc.
In practice, this just means that low end devices will stay on older versions of Android even more than they do already.
I think they’re joking about being single and unemployed.
Well, it falls apart pretty easily. LLMs are notoriously bad at math. And even if it was accurate consistently, it’s not exactly efficient, when a calculator from the 80s can do the same thing.
We have setups where LLMs can call external functions, but I think it would be cool and useful to be able to replace certain internal processes.
As a side note though, while I don’t think that it’s a “true” thought process, I do think there’s a lot of similarity with LLMs and the human subconscious. A lot of LLM behaviour reminds me of split brain patients.
And as for the math aspect, it does seem like it does math very similarly to us. Studies show that we think of small numbers as discrete quantities, but big numbers in terms of relative size, which seems like exactly what this model is doing.
I just don’t think it’s a particularly good way of doing mental math. Natural intuition in humans and gradient descent in LLMs both seem to create layered heuristics that can become pretty much arbitrarily complex, but it still makes more sense to follow an exact algorithm for some things.
I considered this, and I think it depends mostly on ownership and means of production.
Even in the scenario where everyone has access to superhuman models, that would still lead to labor being devalued. When combined with robotics and other forms of automation, the capitalist class will no longer need workers, and large parts of the economy would disappear. That would create a two tiered society, where those with resources become incredibly wealthy and powerful, and those without have no ability to do much of anything, and would likely revert to an agricultural society (assuming access to land), or just propped up with something like UBI.
Basically, I don’t see how it would lead to any form of communism on its own. It would still require a revolution. That being said, I do think AGI could absolutely be a pillar of a post capitalist utopia, I just don’t think it will do much to get us there.
The math example in particular is very interesting, and makes me wonder if we could splice a calculator into the model, basically doing “brain surgery” to short circuit the learned arithmetic process and replace it.
Depends on what we mean by “AI”.
Machine learning? It’s already had a huge effect, drug discovery alone is transformative.
LLMs and the like? Yeah I’m not sure how positive these are. I don’t think they’ve actually been all that impactful so far.
Once we have true machine intelligence, then we have the potential for great improvements in daily life and society, but that entirely depends on how it will be used.
It could be a bridge to post-scarcity, but under capitalism it’s much more likely it will erode the working class further and exacerbate inequality.
The US isn’t the only country, and I bet people using the fediverse skew heavily toward android in the US anyway.
While it’s true that we don’t understand consciousness, LLMs don’t have the hallmarks of consciousness that humans and other animals do.
Modern LLMs are essentially just guessing the next word in a sentence. There’s no continuous experience of the world, and there’s no self, no agency.
Don’t me wrong, it’s fascinating tech, and if we do one day create machine consciousness, it might incorporate parts of our current technology. But right now it’s pretty safe to say that LLMs aren’t conscious, unless you believe in panpsychism / animism.
No modern “AI” has anything resembling consciousness.
The flip style ones are pretty useless other than getting a preview on the main cameras.
The fold style, especially a trifold, seems pretty useful, and lets you basically carry around a decent size tablet in your pocket.
I definitely wouldn’t call it a gimmick, although the price and durability tradeoffs are nowhere near worth it for 99% of people.
coffin_dance.mkv
Same, it’s really easy to keep them separate mentally.
I think QWERTY may even be better for typing on a phone, because common letters are more spread out, leading to fewer errors.
I honestly don’t think this looks very good. The lighting and assets look good on their own, but this remaster completely changes the art direction of the original, and it ends up losing a lot of the atmosphere that the original had.
I hope they make some last minute changes, or we at least get a mod to restore it to the original game’s aesthetics.
I think a simple color grade, and tweaking/removing some of the lights could basically fix it, so I’m a bit surprised they took this approach. It looks like they prioritized tech demo over faithful recreation.
The article is paywalled, but from the part I could read, it sounds like they’re just making hyperbolic colors, which is a pretty well known phenomenon which you can experience at home: https://en.wikipedia.org/wiki/Impossible_color#Chimerical_colors
They’ve done it using a new technique which is cool, but seeing super saturated hyperbolic colors is not a new thing.