This is a very interesting benchmark. Another example is something like, given 1 datapoint e.g. an image of a creature you haven't see before, be able to classify 100 images for whether that creature is visible in the image or not. You, me or even a young child can easily solve a problem like this. Current neural nets require enormous amounts of data and still make mistakes a human would never make. All this AI is going to be smarter than us hype will only be true once we discover algorithmic ideas that get machines to learn given very little data. We absolutely need new ideas.
Not sure if this is representative of AI in general. DeepMind and other AIs have solved games (eg chess, go) using similar examples and training. Thoughts?
Define novelty? A Human child goes through stages of pattern recognition. I think Piaget defined one stage as "pre-conceptual". I forget; has been > 40 years since reading his work. So, maybe 20K hours of training a child goes through to recognize generic patterns. Cross domain Hebbian learning is so different from LLMs. I wonder about the interplay between what a child see, hears, and touches -- a kind of cross domain pattern of space, time, and transformation. BTW, your Mojo demo brought me here.
Man, keep the focus on what that AI is able to do, instead of what isn't!! Because the amount of thing we cannot do is largely greater than what we can! Even humans!! Our ideas of "How intelligence must be" is probably wrong and because we keep trying to find something that didn't exists may push us far from the real solution!! We are on the way! Things are getting better! Even us, humans, we still struggle memorizing things, we also cannot talk the amount of languages LLMs can! Nor even know the amount of subjects a LLM can!! And that's the point!! That king of expectation who looks for "this must happen to be intelligent" seems like symbolic AI trying to figure out how intelligence is and model them! It seems reasonable, but who knows? Nice video!! Thankss
Hi, sir I'm in college learning BCA (Bachelor of computer Application). What would you recommend me to learn to for long term goal. I'm confused please share your thoughts 💭 Thankyou love your videos❤
OpenAI solved it with o-3, but we moved the goalpost yet again
This is a very interesting benchmark. Another example is something like, given 1 datapoint e.g. an image of a creature you haven't see before, be able to classify 100 images for whether that creature is visible in the image or not. You, me or even a young child can easily solve a problem like this. Current neural nets require enormous amounts of data and still make mistakes a human would never make. All this AI is going to be smarter than us hype will only be true once we discover algorithmic ideas that get machines to learn given very little data. We absolutely need new ideas.
Do agentic architectures qualify?
Not sure if this is representative of AI in general. DeepMind and other AIs have solved games (eg chess, go) using similar examples and training. Thoughts?
This is way harder for AI… so far, nobody has cracked it
Define novelty? A Human child goes through stages of pattern recognition. I think Piaget defined one stage as "pre-conceptual". I forget; has been > 40 years since reading his work. So, maybe 20K hours of training a child goes through to recognize generic patterns. Cross domain Hebbian learning is so different from LLMs. I wonder about the interplay between what a child see, hears, and touches -- a kind of cross domain pattern of space, time, and transformation.
BTW, your Mojo demo brought me here.
Man, keep the focus on what that AI is able to do, instead of what isn't!! Because the amount of thing we cannot do is largely greater than what we can! Even humans!! Our ideas of "How intelligence must be" is probably wrong and because we keep trying to find something that didn't exists may push us far from the real solution!! We are on the way! Things are getting better! Even us, humans, we still struggle memorizing things, we also cannot talk the amount of languages LLMs can! Nor even know the amount of subjects a LLM can!! And that's the point!! That king of expectation who looks for "this must happen to be intelligent" seems like symbolic AI trying to figure out how intelligence is and model them! It seems reasonable, but who knows? Nice video!! Thankss
At last some sanity in the web!
What is love?
Baby, don't hurt me
No more
ohhh
Hi, sir I'm in college learning BCA (Bachelor of computer Application). What would you recommend me to learn to for long term goal. I'm confused please share your thoughts 💭
Thankyou love your videos❤