Forums
Most Recent
- My account has been tagged as suspicous regarding fair play policy !
- Is Pirc/Modern Defence too risky for a club player ?
- Square Off - Chess Set
- Report Spam HERE!
- Why Is the Rook so Short?
- 4/25/2018 - K.Begley, 6th Prize, Mat Plus 2008
- Fix The Connection Issues
- Bulldog Chess with Witch- WareDude versus JosephYossi
- JosephYossi and HorribleTomato V. Waredude, Junebug666, and LittleBug
- If You Were to be a GM for a Day, Who Would You Be?
Forum Legend
- Following
- New Comments
- Locked Topic
- Pinned Topic
I don't understand why people are so amazed by this. It had to play a bunch of games before it got stronger. Just like a human. If you gave AlphaZero a Facebook account and told it to pick up the kids from school, buy dinner, gave it access to Youtube videos, etc... It wouldn't be as strong.
The answer is quite clear, humans are not playing chess as strong because of lifestyle choices. AlphaZero had no distractions, no other thoughts creeping in. Humans could be just as powerful if they do the same.
Please don't procreate, ever.
When I heard about it I was able to guess how it worked in about five seconds and he said I was right.
I am sure you said something that was right. However, I bet you did not say everything that is in the paper on the subject.>>>
That would be to assume I could write that paper in five seconds instead of the tens of thousands of hours it undoubtedly took, without knowing anything about it.
The person is called Shaubo Hue. He works for the company but didn't work on this particular project. He did send me the paper on it. I've known him since he was completing his PhD round about 2007. We were on opposite sides of the very ferocious PRC China vs. Tibet debate. Since then he lived in Manchester for a while. He was annoyed at me for my scepticism concerning this. I think my son also gets quietly annoyed at me too but I'm trying to continue to teach him to be creative. Since he got his theoretical physics PhD he worked in defence for three years and is now working at a far less demanding job, programming computers, because he needed to learn that as a professional would need to understand it. He's interested in A.I. and he was talking about taking a Masters in Electronic Engineering. He thinks it might take two years part time with the O.U. because he thinks he might not have to do any of the maths. Is that a reasonable time scale? But I'm sure he wants to work in A.I. and he thinks it's best to understand the hardware properly.
This will cause your head to explode. Click Here
Thanks, Optimissed: very interesting.
Regarding time needed, I am sure there will be specific standard advice from the OU. Took me a while to get what you meant about the maths - you mean he has done a lot of this already. I am sure the OU will explain to him what he can skip, and work out a schedule for what is left.
Hardware in AI is surely important in several ways. One is specialised computing hardware such as google's TPUs, but also GPUs, which serve a similar purpose for most deep learning specialists (because they can be faster than CPUs for the big matrix calculations. Working with parallel hardware of other types is also useful. He may be interested in robotics, which has a lot more varied hardware requirements than most AI work: interfaces in both directions are important. I would see the role of a hardware specialist as complementary to those working on the software and algorithms.
As an amusing coincidence, long, long ago, after I had deserted academia for microcomputers I got interested enough in electronics to do a few modules in it where I could, and found use for that experience later when building automated experiments in applied physics research. An MSc would have been a choice with more potential, but I wasn't so committed to that area. Mostly, I have stuck to working with software, which is more my metier.