A year ago, it was reported that a Go player beat AI several times. It is said t

A year ago, it was reported that a Go player beat AI several times. It is said that this Go player attacked the loophole of AI's algorithm...
====================================
According to the Financial Times on the 18th (local time), an amateur Go player named Kellin Pelline recorded 14 wins out of 15 against the Go AI, Katago. Lee Se-dol declared, "I retire because I can't beat AI," which means that an amateur has crossed the barrier of AI. Lee is no longer the last human to beat AI in Go.

What's interesting is that the key to Pelin's ability to beat Go AI is AI. Through AI, we found weaknesses in Go AI, and Pelin was able to attack and win. AI startup FAR AI has identified weaknesses in Go AI through more than a million matches. FAR AI CEO Adam Gleave said, "We have identified 'blind spots' available to humans."

Pellin placed the stones in the other corner of the board to distract the AI while slowly weaving the stones into a large 'ring' to surround one of his opponent's groups. Pellin said that the AI did not notice the vulnerability until the siege was almost complete. He added, "It would be easy for a person to notice."

=====================================
It's about an article...
Because of these concerns, they argue that autonomous cars can never come out.
No matter how well you make it, there's a problem with the algorithm...

What was especially interesting about the Go game was that after Katago lost in Go, I modified the algorithm again, so I played the game again, but I changed the algorithm again, and I lost again... It went like this.

In the end, even if you modify an algorithm, there's something that you can't fundamentally overcome
Now that you're talking about it...

It seems that AI needs more time to catch up with humans that have evolved over a long period of time.