Self Consciousness of AI

Introduction:

From various industries to our ordinary lives, Artificial Intelligence (AI for short) is everywhere. It assists us to use time and increase our productivity. However, a problem is fatal hiding behind this cutting-edge technology which remains the possibility that AI would potentially gain self-consciousness.

What is self-consciousness? 

Self-consciousness can be defined for an individual as the awareness of his/her body in interactions with the environment and recognising who they are.

For example, a pig identifies exactly who they are and the environment they are currently in. They also experience emotions like joy and misery which only living intelligence could perceive according to biological science. 

How can Super Intelligence threaten humanity? 

We all know what happened to lions and tigers a few thousand years ago right? They are currently labelled as endangered species!

Why does this happen? Well, it all comes down to the fact that human, we simply cannot allow any species that can threaten our lives. Because back in the past when humans didn't invent long-range weapons such as bows and arrows, we were the ones being hunted and manipulated by the animals!

This is realistic for super intelligence as well; once they gain self-consciousness, they would view humanity as its most significant threat, because we created them, and we also have the capability of destroying them. Therefore, the ultimate conflict is inevitable.

AI already has emotions

I didn't write the above game analysis. It's written by IM Daniel Rensch. I just wanted to show you the moves.

Here Lc0 could indeed promote to a queen. But somehow Lc0 decides to humiliate its opponent.

This is extremely unreasonable because how on earth could a computer exhibit any emotions? I'm certain the developers of Lc0 has programmed this bot to find the best move in all positions, and that's what all computers do, their objective is to win the game. On move 155, white could simply move out the king and promote the pawn on the b-file, which is Mate in 14. But Lc0, as I said, chose to play e6 which is a missed win according to the Stockfish Engine.

For example, human players make all sorts of mistakes. But at the instant we play out our move, we would believe that is the best, otherwise, we wouldn't play it.

As far as I know, only human players tend to humiliate or just make fun of opponents such as gobbling all of their pieces before checkmating them. 

The reason Lc0 did this is that it's an AI capable of machine learning, meaning it can learn from mistakes and improve after every single game. So we see why the Elo rating of these bots keeps on climbing higher and it seems there isn't a theoretical limit. However, chess engines running on PC is not able to learn from themselves. 

Conclusion

The development of AI must continue to satisfy higher technological demands each year. Hence, we must figure out a way that humans and AI can peacefully live together.

I think Elon Musk's Neuralink looks promising. One of its considerable advantages is that it provides an environment where AI and humans are coexisting together because their ultimate goal is to implant chips into the human brain. This may sound scary but it solves the problem that AI would potentially renegade against human as can't live by itself without a human brain.

What perspective do you hold on the development of AI? Is there another solution for this conundrum? Please note down your ideas and join our discussion in the comments section!

Thanks for reading! Did you like this blog? Then please show your support by following @QuantumYitian on Twitter to never miss blogs like this.