Hacker News Reader

If AI is the general purpose pattern recognition algorithm and real world has patterns then as long as data i.e. real world has juice, AI will do better. Recursive self improvement is included in the domain. The only constraint will be how long it takes to gather feedback from the real world when AI does experiments.

View on HN

I’m sorry, but I didn’t understand what you meant on and after the word “juice”

I’m not sure any of that was relevant to what I was saying.

And you still didn’t answer my question.

I am only doubtful about the exponential part. I am hopeful that yes all fields will ultimately subsumed by AI.

I believe exponential (as opposed to logarithmic) advancement would be needed for that to happen.

If the improvement curve is linear, it would be prohibitively expensive as and if logarithmic, it’d be impossible.

Another point in your favour is the problem space itself will become exponentially large. Imagine getting a requirement to synthesize a material with 40 properties rather than 1 or 2.

Reply on news.ycombinator.com