Are CPU and memory speed limits preventing us from building AI systems?

Question

Many technology optimists say that after 15 years, the speed of computers will be comparable to the speed of the human brain. That is why they believe that computers will reach the same level of intelligence as humans.

If Moore’s law holds, then every 18 months we should expect a doubling of the CPU speed. 15 years - 180 months. Thus, we will have a doubling of 10 times. This means that in 15 years the computer will be 1024 times faster than now.

But is the cause of the problem? If so, we will be able to create an artificial intelligence system NOW, it will be only 1024 times slower than in 15 years. This means that it takes 1024 seconds (17 minutes) to answer the question instead of an acceptable 1 second. But do we now have a strong (but slow) AI system? I think no. Even if now (2015) we give the system 1 hour instead of 17 minutes, or 1 day, or 1 month, or even 1 year, it still cannot answer complex questions formulated in a natural language. So the problem is not speed.

This means that in 15 years our intelligence will not be faster than 1024 (because we do not have intelligence). Instead, our "stupidity" will be 1024 times faster than now.

Answer 1

We need both faster hardware and better algorithms. Of course, speed alone is not enough, as you indicated. We need self-modifying meta-learning algorithms that can create hypotheses and conduct experiments to test them (for example, people). Systems that learn to learn and improve themselves. Algorithms that can prove that this self-modification is optimal in a certain sense and will lead to even better changes in the future. Systems that can reflect and test their own software (can you call it consciousness?). Such studies are being carried out and may create superhuman intelligence in the future or even technological singularity, as some believe.

. , , , . . , , ( ) . ( ). pleasure 5 . , , . . : " ?". , - , . . ?

Answer 2

. (. Big O). , , . , . , . . : . . , . , , , . , , . , . , (: GPU), , - .

Answer 3

, . , . , , . - , , ? , , . . , :

. , , . , ALOT , 70 . . , , .

. A, . , , , .. , , :

, , I .

, .

, . , , , . , , , , . . , . .

, : " ?

:

, ovens, dangerous, question

and indicate the definition of dangerous as: painful, bad

and then provide more evidence that this is the case, for example, personal experience:

Yes , the stoves are dangerous because I was wounded in my grandmothers house alone.

Thus, intelligence seems to be a combination of events, correlation, and data retrieval to solve a solution. I am sure that there is much more than that, but this is only my understanding of intelligence.