A user says trying AI for slides and coherent plans makes them feel AGI will take a while.
A user says using AI models beyond coding and machine learning—for slides, coherent plans and longish content design—makes them feel AGI is going to take a while.
72.1K
5 Sources, first seen ago
—
Not ranked yet
—
Not ranked yet
@natolambertTrying to use the models for things outside of coding / ml, like making slides or coherent plans or longish content design, makes me feel the AGI-is-going-to-take-a-while.
@SamuelAlbaniecurious what evidence would change your mind? (maybe it's effortful to articulate) fwiw, my sense is that even though slide creation & coherent plans are still bad, the rate of improvement is high (particularly w.r.t. how well they can replicate my voice/style from previous longform examples)
@dhadfieldmenellLonger thoughts to share here, but this is right wrt the dynamics of our current moment. Labs are collectively biased to overestimate progress because anything they focus on improves measurably, but they overestimate their ability to focus on everything that matters.@natolambertTrying to use the models for things outside of coding / ml, like making slides or coherent plans or longish content design, makes me feel the AGI-is-going-to-take-a-while.@SamuelAlbaniecurious what evidence would change your mind? (maybe it's effortful to articulate) fwiw, my sense is that even though slide creation & coherent plans are still bad, the rate of improvement is high (particularly w.r.t. how well they can replicate my voice/style from previous longform examples)@dhadfieldmenellLonger thoughts to share here, but this is right wrt the dynamics of our current moment. Labs are collectively biased to overestimate progress because anything they focus on improves measurably, but they overestimate their ability to focus on everything that matters.@natolambertTrying to use the models for things outside of coding / ml, like making slides or coherent plans or longish content design, makes me feel the AGI-is-going-to-take-a-while.@SamuelAlbaniecurious what evidence would change your mind? (maybe it's effortful to articulate) fwiw, my sense is that even though slide creation & coherent plans are still bad, the rate of improvement is high (particularly w.r.t. how well they can replicate my voice/style from previous longform examples)@dhadfieldmenellLonger thoughts to share here, but this is right wrt the dynamics of our current moment. Labs are collectively biased to overestimate progress because anything they focus on improves measurably, but they overestimate their ability to focus on everything that matters.@natolambertTrying to use the models for things outside of coding / ml, like making slides or coherent plans or longish content design, makes me feel the AGI-is-going-to-take-a-while.@SamuelAlbaniecurious what evidence would change your mind? (maybe it's effortful to articulate) fwiw, my sense is that even though slide creation & coherent plans are still bad, the rate of improvement is high (particularly w.r.t. how well they can replicate my voice/style from previous longform examples)@dhadfieldmenellLonger thoughts to share here, but this is right wrt the dynamics of our current moment. Labs are collectively biased to overestimate progress because anything they focus on improves measurably, but they overestimate their ability to focus on everything that matters.
