Bagholder has been an active tournament chess player for decades. When you consider all the people who take it serious enough to get rated, and play in USCF or FIDE tournaments, Bagholder’s rating is in the top 1% of all tournament players. Bagholder was blessed in his chess career to have made friends with Grandmaster Igor Ivanov. In the 1970’s-1990’s he was an absolute terror over the board. Igor, in his only career opportunity, beat former world champion Anatoly Karpov (linked here). Igor was, without doubt, the top 1% of the top 1%. Knowing Igor as well as Bagholder did, was a very humbling experience. The difference in skill and talent between Igor and myself, was absurd. He was just on a different level. Make no mistake, Bagholder can play - but not like this guy.
In the late 1990’s computers were reaching the point where they could beat the best chess grandmasters on the planet. Perhaps some of you remember a pair of highly publicized, 6 game matches between IBM’s Deep blue and world champion at the time, Garry Kasparov. Garry won the first match in 1996, and deep blue won the second in 1997. Fast forward to today, and the top Grandmasters cannot compete with the top computers, anymore than Bagholder could compete with Igor.
This begs the question, are computers more intelligent than Human beings? Is artificial intelligence (AI) really a thing? Webster’s defines intelligence as the “ability to acquire and apply knowledge”. Bagholder would argue computers are not good at either acquiring knowledge, or applying knowledge. What computers are good at, is doing what they are programmed to do. Even a dollar store calculator can multiply 6 digit numbers better than any human alive, but ask that same calculator to spell the word “dog”, and your average 5 year old is better. The reason is, the calculator has NO ability to acquire knowledge on its own.
Bagholder, always looking to acquire knowledge and improve his chess game, once asked Igor how many moves ahead do you usually look, when playing chess against world class opponents? His response was “Only one, but it is usually the right one”. When Deep Blue beat Kasparov in a match, it was because the computer could calculate every possible move, and every response to that move, and every response to that response, and on, and on…. In other words, the computer was looking at millions of possible variations, to find the best move. So even though computers can beat the top Grandmasters, are they really more intelligent if they have to look at millions of variations to do it? The Igor’s of the world are able to intuitively hone in on what is important in a given position. They are not wasting time or energy considering millions of variations. They are wisely considering just a few, and choosing the best - without all the calculation. Bagholder would call that, intelligence.
If we were to Instruct Deep Blue to only consider a few variations, and choose the best, like humans do - Bagholder could play blindfolded and easily beat Deep Blue. The reason is, computers are indeed good at raw calculation, they are NOT good at applying knowledge in any way other than blunt force. It is not just chess either, you can load a computer with the contents of every book in your average University library, something no human could ever do. But ask that same computer to write like Hemingway, Shakespeare, or Twain, and it doesn’t have a chance. Again, it has NO ability to apply the knowledge it has been given.
Many in the industry believe AI is only a matter of when, not if. Bagholder is not buying it. Multiplying long numbers, or beating a grandmaster at chess is remarkably easy for machines specifically programmed to do those tasks. They are good at it, because that is what they were told to do. Ask yourself, who is more intelligent, the guy who always does what he is told, or the guy who thinks for himself? Whether it is people or machines, doing only what you are told is not a sign of intelligence, it is a sign of a lack of intelligence.
If and when they build a computer which exhibits human traits like ambition, creativity, intuition, wisdom, curiosity, or understanding nuance, Bagholder might be a believer. Fact is, none of those traits exist, even in the most powerful computers. Until such time as they do, Bagholder would take human intellect over a computer, every time. After all, there is a reason it is called “Artificial Intelligence”. By definition, Artificial means “fake, or fraudulent”. That sounds about right.
This is a spot-on analysis. I appreciate it.