gowan wrote:I doubt that there would be any super strong AIs except that for playing games like chess and go are simple, easy test projects. The success of AlphaGo and other strong AIs is commendable but go and chess are still just games with little significance in the wider world. AI should get on with the real job of improving life on the planet in general.
What specific "real jobs" do you have in mind? There have already been huge strides in a number of fields due to:
* Advancements in hardware
* Lots of data
* Machine learning techniques that can exploit the two items above
Computer vision, speech recognition, etc., have improved immensely. Customized experiences on the web (e.g. google news, Amazon product recommendations, etc.) have made the user experience online much more personal. Machine learning and AI are affecting many areas of life right now. And with the data and hardware that's currently available, things that used to be less feasible are now becoming reality.
Even "AlphaGo" was used to save 40% on cooling costs for data centers:
https://deepmind.com/blog/deepmind-ai-r ... g-bill-40/, which is a very "real" problem.
I think in our lifetimes, many aspects of living will change at a rapid pace...