No. All decisions about inputs are made by a player. Bots will be used occasionally for repetitive tasks or to make manipulation of RNGs easier, but ultimately the decision is left to a human as to what inputs are selected. A fully automated TAS hasn't been made yet as far as I know, but it could happen sometime.
A piano playing a piano roll is not one which is running off AI. It's just a predetermined series of instructions, which is not being calculated/generated in real time via an artificial thinking machine.
This is exactly the same as what i discussed with him, so he started to search and found that there is a concept called TAS-bot and TAS is achieved by bot, human can't do such gameplay even with slow motion and S/L. Such a fool.
Joined: 4/17/2010
Posts: 11495
Location: Lake Chargoggagoggmanchauggagoggchaubunagungamaugg
Just wow.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
It baffles me that there are really people coming to the conclusion that the runs at AGDQ are done by Tasbot itself....
If we really managed to create such a bot we'd probably be in all kind of computer science magazines.
Well, "artificial" just means something made by people instead of occurring naturally, right? And people are made by people, so by that logic, people are artificial (aside from whatever you decide the "first" people were, who of course occurred naturally). Assuming then that you buy that people are intelligent, people are artificial intelligences. ;)
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Let's just quote wikipedia on Artificial intelligence:
Well that was somewhat to be expected. Alright, then let's quickly switch to the Wikipedia page of Intelligence:
The only information TASBot can perceive is the latch from the console asking for input, to which it replies with the input data stored in its brain. (This information in the brain, however, is being given by a human and it's not being modified)
This means that whether or not you want to call TASBot an AI, it's never going to be an AI playing a game. At most, it's going to be an AI being really good at playing back a movie.
Though, I don't know if we can just simply apply definitons to this question.
Yes they can. That's how it's being done.
Flutter A.I is superior anyways!
Warning: Might glitch to creditsI will finish this ACE soon as possible
(or will I?)
Philosophically, I think that artificial intelligence requires the ability to think in all of the ways that humans can think. In a Turing test, a bot can fairly easily mimic a human in knowledge and problem solving. (A human and bot would answer 'what is the atomic number of hydrogen' or 'what is 2+2' the same way.)
Creativity is where AI falls short, at least for now. If you ask a human to 'list some creative uses for a paper clip,' you will get answers like a fish hook, lock pick, earring, etc. A Turing test bot won't respond that way. At best, it will say something like, 'how about you list some creative uses for a paper clip.' More likely it will say something irrelevant.
Specific to TASing, even a run generated completely by a bot isn't really AI. We can program a bot to find faster solutions, but what about creativity? A bot playing the SMB3 airships wouldn't bounce off of the cannonballs to obtain 99+ lives magnificently.
Brain Age is a good test of AI. Could it create (on its own) a beautiful drawing? Could it draw something funny? Could it draw something sad? Until it can create a Brain Age TAS the way humans can (original ideas, including art, humor, emotions), it's not really AI to me.
Joined: 4/17/2010
Posts: 11495
Location: Lake Chargoggagoggmanchauggagoggchaubunagungamaugg
So how fruitful was that move in the end, regardless of how odd and beautiful it looked?
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.