When will there be AI superintelligence?
-
When will there be AI superintelligence?
@evan not in our lifetime
-
When will there be AI superintelligence?
No
-
When will there be AI superintelligence?
@evan
It's already there. GPT 6 Astra is smarter than any human. -
When will there be AI superintelligence?
@evan I will go with something a little bit arbitrary, but fun: the early 1970s with the invention of the first pocket-sized electronic calculators. These AI had extraordinary intelligence and could do very difficult calculations at lightning speed. Almost whatever multiplication or division with many decimals you threw at it, these AI would give you the answer in an instant. And the answer was always right.
-
@evan I will go with something a little bit arbitrary, but fun: the early 1970s with the invention of the first pocket-sized electronic calculators. These AI had extraordinary intelligence and could do very difficult calculations at lightning speed. Almost whatever multiplication or division with many decimals you threw at it, these AI would give you the answer in an instant. And the answer was always right.
@evan These AI super intelligent machines couldn't read my mind, couldn't design a beautiful house, touch my shoulder in just the right way to comfort me or facilitate a consensus-meeting with culturally different people. It had a very specific kind of artificial intelligence and was super good at that. I still rely on this kind of AI superintelligence on a regular basis!
-
@evan I don't think that it's possible for a glorified pile of linear algebra to think at all.
@me can any software think?
-
@evan How about never?
@davep why not?
-
@simonzerafa why not?
-
@davep why not?
@evan Because matrix multiplication and activation functions will never be superintelligent, no matter how much deterministic code they use in harnesses to make it good at specific things.
-
@evan Because matrix multiplication and activation functions will never be superintelligent, no matter how much deterministic code they use in harnesses to make it good at specific things.
@davep how are humans intelligent? Magic?
-
@davep how are humans intelligent? Magic?
@evan Not by variations on matrix multiplication. Unless they find some new type of maths I just can't see it not having really stupid non-human-like failure modes.
-
@davep how are humans intelligent? Magic?
@evan@cosocial.ca @davep@infosec.exchange not magic but thanks to processes we don’t completely understand, let alone replicate
-
@evan@cosocial.ca @davep@infosec.exchange not magic but thanks to processes we don’t completely understand, let alone replicate
-
-
When will there be AI superintelligence?
@evan You're smuggling in a premise that it will happen. When people say "na-hah" you shift the burden of proof onto them. If you think it's going to happen (including in the lifetime of some readers here) then the onus is on you to explain why you think that.
-
@evan
It's already there. GPT 6 Astra is smarter than any human. -
The point is that current LLM tech doesn't have the capacity for understanding truth or accuracy in the way we do. It's a statistical probability process. It's getting refined in some pretty cool ways, but that fundamental inability to actually understand is just dealt with crudely by harnesses in order to try to attain some form of deterministic approach, and that will never scale computationally to encompass anything approaching the human experience and breadth of understanding/intuition etc.
-
@davep how are humans intelligent? Magic?
In other words, yes, quite possibly some form of "magic" compared to what we think today. There's so much we don't know about how the brain works, and so much has evolved in tangled complex ways over millions of years of evolution. There could be quantum effects involved in sentience for all we know. It could be much more difficult.
-
@evan@cosocial.ca @davep@infosec.exchange
To be honest the second is not a good case at all, reproducing is not something we control, we kinda just go with the flow, maybe we can twitch some little things when we’re lucky but it’s more biology making us than the other way around isn’t it?
You’re right by saying we are be able to make things work even if we don’t truly understand what’s going on under the hood. I guess cooking is a good mundane example. And I feel this is what they’re trying to to with AI, like if they were trying to make AGI like if it was a loaf of bread “just a couple of petabytes more and AGI will leaven, let me cook bro”.
I just think this specific field of knowledge needs something more than limitless injection of capital providing infinite computing power to process all stolen data humans can put together, maybe a philosophical approach less similar to a 60s-fashion space race or Third Reich experiments on human -
The point is that current LLM tech doesn't have the capacity for understanding truth or accuracy in the way we do. It's a statistical probability process. It's getting refined in some pretty cool ways, but that fundamental inability to actually understand is just dealt with crudely by harnesses in order to try to attain some form of deterministic approach, and that will never scale computationally to encompass anything approaching the human experience and breadth of understanding/intuition etc.
️