Googles ai

https://experiments.withgoogle.com/ai/ai-duet/view/

Theres a bunch of experiments with the new ai… this is a piano you play and the ai responds by playing what it feels fits with what you did… its clever…

2 Likes

Lol it gets excited sometimes and plays over you…

Well that was fun for 10 minutes!

1 Like

Yea… lol… it has other experiments…

Visualize high dimensional space and see how computer learning works i guess lol… its open source too.

1 Like

Seems deep learning/neural networks is where it is at right now.

2 Likes

Google’s AlphaZero is going to take over the world:

I had fun with this.
They tell you what to draw in 20 seconds, and then it has to guess correctly.

2 Likes

Thats pretty clever. Will we need human interaction if ai becomes better than humans?

Will we need humans?
Maybe for entertainment value?

1 Like

I think AI has already progressed. What if the singularity already happened?

1 Like

The best part so far… googles ai has created “children” ai… with specific functions… they are better designed than anything any human has or currently can make… i want it to play with time crystals… watch our reality become infinitely more interesting…

@Csummers lol what happens when the ai becones bored of this planet…

I found out that I can’t draw worth a damn. It thought my crocodile was a squirrel, my toaster was a house, and my lighter was a candle. :smile:

1 Like

I think the goal of A.I. is to develop tools to aid humans in ways they would not think of. That is how AlphaGo won the final game against the highest rated Go player. It won the game by only 2 points and it was thought to be a glitch that they should have worked when they were preparing to play the games. Further, they did not know what the AlphaGo program was doing when they read the move as an unusual move that only had a 0.1% chance of people would make and the program agreed and was sure it had a 95% chance of winning. So they thought the program was glitching as they had another professional Go player look at the program prior to the matches to identify any flaws and to which he identified this type of play as a kink in the system’s ability. However it turned out to be it’s stregnth.

Actually the documentary is on Netflix about the whole thing. People are worried about machines going against their programing and becoming rouge but they still adhere to their programing, just not in ways you would think of.

1 Like

I heard it took 100 moves so it was a challenge to it at least a little bit… i bet nothing we can present it against could win in unlimited games…

Lol i havent tried the drawing one. Your drawings look like what both of you thought lol

1 Like

The guesses it comes up with are hilarious. What’s interesting is that going through several rounds, it’s like you learn how to draw just the basic features so it’s almost like the AI is training you to draw recognizable features.

2 Likes

There were 5 games. It lost the 4th one and won every other one.

1 Like

Akinator is always a fun one. He wasn’t always the smartest though. But now he could guess anyone you could think of.

1 Like

I love akinator. I introduced my parents to it. Totally blew their minds.

2 Likes

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.