Perhaps I was being too harsh but I don't consider behaviour you can sketch out on a 3x3 post-it note as AII think that's a little unfair.
Soup didn't say 'solved using A'I, but solved using principles ("techniques") learnt from studying AI. I guess he means getting a different set of mental models to view a problem from and being able to express it as Big Data or relationship models or a branchy tree structure or a car analogy...
is DSoup a spy?
Perhaps I was being too harsh but I don't consider behaviour you can sketch out on a 3x3 post-it note as AI
Yup. Lots of once promising things, like machine learning and decision trees, were evolved in the pursuit of believable AI. Many technqiues were discarged but have found uses in non-AI applications.
"We must scrupulously guard the civil rights and civil liberties of all citizens, whatever their background. We must remember that any oppression, any injustice, any hatred, is a wedge designed to attack our civilization." -- Franklin D. RooseveltI don't work for GCHQ. Our resources are used for entirely for the benefit of mankind and finding really funny cat videos on the internet. Actually, that does sound a lot like GCHQ.
Dont tell people that Machine Learning is not AI, then there are a few VCs around the world that will see their cash run the same way as their Crypto coins will
"So are you a spy or not?
We're waiting for a breakthrough in AI.Just silence. Followed by a broken neck.
What about Artificial Intelligence: A Modern Approach by Russel and Norvig ? I liked this one and thought it was a must read.Anything AI-related by Jack Copeland.
It's not very modern any more which isn't a problem and I think it is still a must read if you want an intro of how to get into the mindset needed for making machine learning systems.What about Artificial Intelligence: A Modern Approach by Russel and Norvig ? I liked this one and thought it was a must read.
It's not very modern any more which isn't a problem and I think it is still a must read if you want an intro of how to get into the mindset needed for making machine learning systems.
Fair enough. That chat was actually after Isgreen and Beeckman had taken me through a closer look at those features in the game, via a peek at the new Tamerlane campaign, the most striking of them being that new AI.
"If you look at the old AI, it cheated" Beeckman chuckled. "It totally cheated. It doesn't cheat anymore."
"If you put one AI, non-cheating, of the Definitive Edition version against seven of the original that do cheat, it just wipes the floor with them. No problem. Every time. So that's the kind of level we were aiming for here... the new AI is so good that you can watch replays of it to actually learn how to play the game, because it plays the game correctly now... it tries to follow the competitive meta - on the highest [difficulty] level - because we have 20 years of pro players doing tournaments and we tried to get that in the AI as well."
You can also talk to the new AI in a more sophisticated way. I didn't see this in action but the example I was given was how, in the past, interactions were limited to things like "do you have any spare Gold you could give me?" whereas now you could ask an AI ally "hey, can you attack, with Knights, that specific player at that time", and they'll go ahead and do it.