Gary Marcus is certainly good. It’s not as if I think say, LeCun, or any of the many people who think that LLMs aren’t the way are morons. I don’t think anyone thinks all the problems are currently solved. And I think long time lines are still plausible, but, I think dismissing short time line out of hand is thoughtless.
My main gripe is how certain people are about things they know virtually nothing about. And how slap dashed their reasoning is. It seems to me most people’s reasoning goes something like “there is no little man in the box, it’s just math, and math can’t think.” Of course, they say it with a lot fancier words, like “it’s just gradient decent” as if human brains couldn’t have gradient decent baked in anywhere.
But, out of interest what is your take on the Stochastic Parrot? I find the arguments deeply implausible.
I’m not saying that we can’t ever build a machine that can think. You can do some remarkable things with math. I personally don’t think our brains have baked in gradient descent, and I don’t think neural networks are a lot like brains at all.
The stochastic parrot is a useful vehicle for criticism and I think there is some truth to it. But I also think LMMs display some super impressive emergent features. But I still think they are really far from AGI.
Gary Marcus is certainly good. It’s not as if I think say, LeCun, or any of the many people who think that LLMs aren’t the way are morons. I don’t think anyone thinks all the problems are currently solved. And I think long time lines are still plausible, but, I think dismissing short time line out of hand is thoughtless.
My main gripe is how certain people are about things they know virtually nothing about. And how slap dashed their reasoning is. It seems to me most people’s reasoning goes something like “there is no little man in the box, it’s just math, and math can’t think.” Of course, they say it with a lot fancier words, like “it’s just gradient decent” as if human brains couldn’t have gradient decent baked in anywhere.
But, out of interest what is your take on the Stochastic Parrot? I find the arguments deeply implausible.
I’m not saying that we can’t ever build a machine that can think. You can do some remarkable things with math. I personally don’t think our brains have baked in gradient descent, and I don’t think neural networks are a lot like brains at all.
The stochastic parrot is a useful vehicle for criticism and I think there is some truth to it. But I also think LMMs display some super impressive emergent features. But I still think they are really far from AGI.