AlphaGo followed by ChatGPT, the future of Go and humans

Ask ChatGPT. What is the meaning of the confrontation between AlphaGo and Lee Sedol? In less than a second, long answers come out one after another. “AlphaGo, a Go AI program created by Google Deep Mind, defeated Lee Se-dol, the world’s best player, in 2016. A monumental moment in the AI ​​realm. It proves that AI has the potential to accomplish great things.”

Q: Then, what are the similarities and differences between AlphaGo and ChatGPT?

A : “Both have the ability to decide and carry out tasks that were previously the purely human domain. AlphaGo was designed for the Go game, and ChatGPT is a language model that composes sentences like humans.”

Q: Do you know AlphaGo Zero?

A : “The uniqueness of AlphaGo Zero is that it plays Go by its own reinforcement learning without any prior knowledge or human input. This program surpasses previous versions of AlphaGo and all human capabilities, demonstrating that machine learning algorithms can perform at extremely high levels in complex domains.”
AlphaGo Zero played Go among themselves, not knowing anything about the Go rules and the goal of winning. He learned Go on his own by playing 30 million boards in just 40 hours by self-taught without notation, and later faced AlphaGo 100 times and won them all. AlphaGo Zero shows how far AI can go with just reinforcement learning without any data. So he asks

Q: Can ChatGPT perform reinforcement learning like AlphaGo Zero without external input?

A : “I don’t practice reinforcement learning in the traditional sense the way AlphaGo Zero did. I have extensive information and knowledge because I have been trained on the vast amount of data I have acquired from the Internet. We can use that information to generate responses in a way similar to reinforcement learning.”
The answer, ‘It’s not reinforcement learning, but in a similar way’, draws attention. It foreshadows that ChatGPT and similar AIs will evolve at a frightening speed. How far can their abilities go?

Q: Numerous Go AIs have appeared since AlphaGo, and they are considered to be ‘gods of Go’ because they are much stronger than humans. Can ChatGPT one day become a god who finds the right answer for humans?

A : “As a language model developed by OpenAI, I do not have human-like emotions or desires, and I do not control my abilities or usage. Whether or not I become ‘godlike’ to humans depends on how I am perceived and used by the people I interact with, and how AI technologies develop and evolve in the future.” 카지노
In the end, I mean, it depends on the person. The infinite competition has already begun. ChatGPTs will certainly evolve at an astounding speed as the enormous capital of global conglomerates has embarked on a new war called ‘Finding the Right Answer’. It must be said that an enormous amount of reinforcement learning has already begun. The AI ​​never sleeps and has infinite stamina. In order for humans to play 30 million Go games, it takes about 10,000 years, 10 games a day. AI takes 40 hours.

Go is already an AI world. It is said that the checkerboard is the epitome of life, but can we guarantee that what happened on the checkerboard does not happen in the world?

There is no moral good or evil in Go. No ideology, no religion. So, the decision of Go AI is accepted without any objection. The world is different. AIs such as ChatGPT continue to evolve and finally only give the right answer, and people think of a world where they believe in it, and suddenly chills run down their spine. The dystopia of future novels, I can feel the dark energy. It’s a problem even if there is no right answer in the world, but it’s even scarier that there is only one right answer.

Many words are appearing with ChatGPT. Neutrality and fact, fallacy and bias, manipulation and derivation, and so on. However, while talking about institutional preparations while talking about this, it seems that AI will go to a very far place that is out of our reach and smile.

Ask ChatGPT. What is the meaning of the confrontation between AlphaGo and Lee Sedol? In less than a second, long answers come out one after another. “AlphaGo, a Go AI program created by Google Deep Mind, defeated Lee Se-dol, the world’s best player, in 2016. A monumental moment in the AI ​​realm. It proves that AI has the potential to…