[AI Story] Artificial Intelligence Challenging Humans (1) Deep Blue (Deep Blue)

2024-07-04

AlphaGo's shock was just around the corner, but artificial intelligence technology is already evolving at a very rapid pace. In the future, artificial intelligence will have more impact on our lives in more diverse fields and everyday life.

‍

However, as we live in such a rapidly changing era, it seems that vague fears have grown along with rosy expectations about the future with artificial geogong. It seems that artificial intelligence will quickly replace humans soon, and many people are uneasy about the uncertain situation where it is difficult to predict the future.

‍

At times like this, how about predicting the future through the past of artificial intelligence, as there is a saying, “If you know the past, you can see the future.” I would like to look back on the history of artificial intelligence, which has developed by challenging humans until now, and find insights for a future where people must live coexisting with artificial intelligence in the future.

‍

Let's start with a story about Deep Blue (Deep Blue), which first broke the stereotype that computers can't beat humans.

Humans vs. artificial intelligence, the beginning of a fierce confrontation

‍

MaChack, Image source

‍

The history of artificial intelligence's challenge to humans is older than expected.

‍

MacHack (MacHack), created by MIT hacker Richard D. Greenblatt (Richard D. Greenblatt) in 1967, won a chess match between humans. However, the opponents at the time were amateurs rather than professionals. McNack was no match for a trained professional chess player.

Since then, the challenge of artificial intelligence has continued; in the early 1990s, the artificial intelligence Chinook (Chinook) tried to become a chess champion but lost. There have also been several other chess matches with humans, but until then, no computer could overcome human barriers.

‍

The artificial intelligence computer that beat the chess champion

Gary Kasparov VS Deep Blue, Image source

‍

Deep Thought (Deep Thought), developed by IBM in 1989 (prior to Deep Blue), has a fateful first confrontation with world chess champion GarryKimovich Kasparov (GarryKimovich Kasparov). However, they lost on the first challenge. At the time, it seemed like it was still a long way off for machines to surpass humans in advanced brain games.

‍

Then, in 1996, IBM tried again with the more advanced Deep Blue (Deep Blue) and achieved a historic victory. Almost 30 years after the artificial intelligence challenge began, they finally defeated humans. However, the final stats failed to achieve victory with 1 win, 2 draws, and 3 losses.

However, in 1997, Deeper Blue (Deeper Blue), which eventually improved Deep Blue, succeeded in defeating the human chess champion with 2 wins, 3 draws, and 1 loss. The impact of being the first at the time and the shock of losing the legendary Grand Master of Chess were added, and it attracted tremendous attention around the world.

‍

As you know, the rate at which artificial intelligence overwhelms humans will gradually increase in the future.

‍

Why Deep Blue was able to win

‍

Chess is a complicated game. There can be a large number of cases (not as many as in Go). So how did Deep Blue beat humans in such an advanced brain game?

At the time, the Deep Blue development team said, “They believed that pure intelligence alone could not beat a world chess champion, and instead took the approach of using tremendous computational power to make a difference.” * “To this end, we created artificial intelligence software at the level of a world chess champion by combining artificial intelligence-type algorithms and supercomputer-level computing power,” * he stated.

At the time, Deep Blue was a supercomputer with a huge level of parallel processing** capabilities. With strong hardware performance, we investigated the number of all possible cases and determined the next number. “In general, professional chess players expect about 10 numbers, but the Dipper Blue system that beat Kasparov had the ability to predict 12 numbers.” It's ***.

However, it is said that this kind of Deep Blue would not have been able to beat a human master in Go, which is much more complicated. After all, it's an old technology that appeared 20 years before AlphaGo. On the other hand, AlphaGo, which applied the latest artificial intelligence technology, was able to build more powerful performance thanks to machine learning and neural network learning.

‍

While finishing

‍

Kasparov's TED talk, Image source

‍

Paradoxically, by learning about Deep Blue this time, I was able to get a glimpse of greater possibilities for the future of humans. That was also thanks to Kasparov, who was the first human representative to experience a bitter defeat for artificial intelligence.

‍

From a TED talk a few years ago, this great chess legend “Don't be afraid of artificial intelligence machines; cooperate with them. (Don't FEARINTELLIGHTY MACHINES. Work with them.)”It says It conveyed insights from experience for future generations who will have to work and live with artificial intelligence in the future.

‍

Rather than worrying about competition with unknown machines, I was able to envision a new future of human intelligence through his words to see greater potential to further enhance human abilities through artificial intelligence and collaboration.

‍