Still not SWOS! (3 Viewers)

I fully agree, pruning is the important thing. It saves time which is essential when you are playing against the clock.
This is the fantastic thing about Houdini, the engine I mentioned earlier on in the thread. It is like an anti engine, engine. It will go to where positions are unclear in its evaluation and give it some wellie in the evaluation department where other engines might of just pruned it off. Doesn't seem much but this romantic way of playing has been killing other engines. This is where subjectivity beats objectivity. Concentrating on other engines weaknesses. Kasparov made a killing at it.

There are chess engines though that are self learning, These are the new breed, which are far and few between. Giraffe is perhaps the latest well known one but it is not one which I would determine as true self-learning. Those beasts are a few generations off at least.

The Chess Engine that Died So AlphaGo Could Live

Today's news two weeks ago!

'Superhuman' Google AI claims chess crown

I'll have to go through the games and see what's going on. I'm a bit suspect on what's being claimed here especially with no opening book in its system.

This is the claim.

Entire human chess knowledge learned and surpassed by DeepMind's AlphaZero in four hours

Seems the hype machine has gone into overdrive on this.

Been doing the rounds on the chess sites seeing what people had to say on this, and they're not impressed.

They have limited the capabilities of Stockfish to work out its moves. Also, the learning capabilities of AlphaZero seem to tail off after a few hours.

They still has a long way to go.