Manny’s review of What to Think About Machines That Think: Today's Leading Thinkers on the Age of Machine Intelligence – Essays on Artificial Intelligence and Human Destiny (Edge Question Series) > Likes and Comments
8 likes · Like
Cliff-hanger.
Yes, will it be the unification of General Relativity and quantum mechanics or have he and his AIs already decided that's too small-scale?
I did my PhD in machine learning in the 80s. I never doubted AGI will happen one day soon, but I was surprised by the progress in the last 5 years. I still do not believe the current technology alone can produce AGI simply through scaling. We have to wait for a few fundamental breakthroughs, but understanding of the biological brain, and even consciousness, is progressing rapidly. Therefore, it can certainly happen in the next decade.
I read the book in 2017 but for some reason did not review it and I only remember that I was not overly impressed. You are right people seem to be not very good in predicting the future.
But there are exceptions. Do you know The Great C by Philip K. Dick? I read that a couple of days ago and was excited. Written in 1953 he describes a LLM that in a post-apocalyptic world must answer three questions every year. And is obviously bored by the questions. "What keeps the sun moving through the sky? Why doesn‘t it fall to the ground?" – "You will be astonished at the answer. The sun does not move. At least, what you see as motion is not motion at all. What you see is the motion of the earth as it resolves around the sun...“ And so on. No computer language, but fluent language and the Great C knows exactly what the guy asking the question knows and will comprehend. Quite amazing.
Kunal wrote: "I did my PhD in machine learning in the 80s. I never doubted AGI will happen one day soon, but I was surprised by the progress in the last 5 years. I still do not believe the current technology alo..."
Perhaps we need at least one more breakthrough, but I don't think it will be in the area of understanding the human brain or consciousness. To me, it seems more likely to be a technical improvement to machine learning architectures, like transformers. I am sure we will soon find out :)
Liedzeit wrote: "I read the book in 2017 but for some reason did not review it and I only remember that I was not overly impressed. You are right people seem to be not very good in predicting the future.
But there..."
What is surprising is that so many of these clever people appear to believe they are good at predicting the future.
back to top
date
newest »
newest »
message 1:
by
notgettingenough
(new)
10 hours, 26 min ago
Cliff-hanger.
reply
|
flag
Yes, will it be the unification of General Relativity and quantum mechanics or have he and his AIs already decided that's too small-scale?
I did my PhD in machine learning in the 80s. I never doubted AGI will happen one day soon, but I was surprised by the progress in the last 5 years. I still do not believe the current technology alone can produce AGI simply through scaling. We have to wait for a few fundamental breakthroughs, but understanding of the biological brain, and even consciousness, is progressing rapidly. Therefore, it can certainly happen in the next decade.
I read the book in 2017 but for some reason did not review it and I only remember that I was not overly impressed. You are right people seem to be not very good in predicting the future.But there are exceptions. Do you know The Great C by Philip K. Dick? I read that a couple of days ago and was excited. Written in 1953 he describes a LLM that in a post-apocalyptic world must answer three questions every year. And is obviously bored by the questions. "What keeps the sun moving through the sky? Why doesn‘t it fall to the ground?" – "You will be astonished at the answer. The sun does not move. At least, what you see as motion is not motion at all. What you see is the motion of the earth as it resolves around the sun...“ And so on. No computer language, but fluent language and the Great C knows exactly what the guy asking the question knows and will comprehend. Quite amazing.
Kunal wrote: "I did my PhD in machine learning in the 80s. I never doubted AGI will happen one day soon, but I was surprised by the progress in the last 5 years. I still do not believe the current technology alo..."Perhaps we need at least one more breakthrough, but I don't think it will be in the area of understanding the human brain or consciousness. To me, it seems more likely to be a technical improvement to machine learning architectures, like transformers. I am sure we will soon find out :)
Liedzeit wrote: "I read the book in 2017 but for some reason did not review it and I only remember that I was not overly impressed. You are right people seem to be not very good in predicting the future.But there..."
What is surprising is that so many of these clever people appear to believe they are good at predicting the future.
