AI from the Archives
[AI seems to be on everyone’s lips these days. Media talking heads say that there will be massive job loss. Companies are using the excuse of AI for basic cutbacks in the cost of labor for increased profits. Parents worry that their recently graduated from college children won’t be able to get entry level jobs. Everywhere there is both hand-wringing and fear of losing control. People have started to use the term “AI slop” when talking about LinkedIn and other social media platforms”. Entire industries are in a panic.
Given all the new subscribers to this newsletter this year and the general subscriber churn, I decided to republish the columns I have written on AI as they still are what my thinking is and perhaps they will help all of you to have a deeper understanding of the potential and possible problems with it. These columns have been published here over the last three years and still represent my developed thinking.
Since ‘move 37’ back in the spring of 2016 [see the first column published below] I have been speaking and writing about AI. It is important to learn the history of the technology, what it could do, and the evolutionary power it has within it. It will rival every major transformational technology in history: the invention of eye glasses, the moveable type press, the invention of the internal combustion engine, radio and television and of course the Internet. It could well be far more impactful than any of these.
Currently we are in the early-stage ‘corporate product competition’ of implementation. This is not what it will be, just a toe in the water of a great ocean of possibilities. So please start with today’s column. They will be published in chronological order.
The column below was published almost exactly two years ago]
“AI” is the Next Big Thing – Move 37
FEB 14, 2024
Over the past few decades, there have been a lot of “next big things” expected to happen technologically. These include VR [ Virtual Reality ] AR [Augmented Reality] BMI [ Brain Machine Interface – now Brain-Computer Interface, BCI] and, drumroll please ….Artificial Intelligence[AI].
VR, AR, BCI are all finding their way to a slow scale-up through time. “AI” is here.
[This column is the first of several about “AI.”]
It was back in 2016 when I realized that AI had begun to arrive and would, indeed, be the next big technological thing. I receive an inordinate amount of inbound emails in order to cast a wide net across multiple categories to spot trends, developing patterns and simultaneities. Over the course of two days, I saw several inbound emails with “move 37” in the subject line.
Huh? Move 37?
So I did the necessary deep dive to investigate. The result was that I saw that AI was finally going to be the next big thing. More on the specifics of move 37 further down.
Quick history of “AI”
The quick history of “AI” is that in 1996, the reigning global Grandmaster in Chess, Garry Kasparov beat IBM’s Big Blue 4-2 in a six-game match. Kasparov was regarded as one of the greatest grandmasters of all time. Big Blue was the most powerful supercomputer in the world at the time. One year later, in May, 1997 Deep Blue defeated Kasparov 3.5-2.5 over six games. This was the first time the reigning grandmaster had ever lost to the world’s fastest computer. Never again would the reigning grandmaster beat the world’s fastest computer.
The reason for computer dominance over the best human was simple: massive computational power to play thousands of games against itself daily. It has been said that grandmasters can think 30+ moves in advance. The supercomputer used its massive power to figure out all the possible moves by simply playing a human lifetimes’ worth of chess games in a single day. Twenty years later, in 2016, Kasparov spoke about this match:
“While writing the book [Kasparov’s book about the match]I did a lot of research – analyzing the games with modern computers, also soul-searching- and I changed my conclusions. I am not writing any love letters to IBM, but my respect for the Deep Blue team went up, and my opinion of my own play, and Deep Blue’s play, went down. Today you can buy a chess engine for your laptop that will beat Deep Blue quite easily”
Wikipedia
So, a supercomputer beats the reigning grandmaster of chess in 1997 and by 2016, any consumer could purchase a computer program that would easily beat the 1997 supercomputer. That is rapid development of “AI”.
The next big event occurred in 2011 when the two greatest Jeopardy players competed against Watson from IBM [Big Blue had been renamed to honor the founder of IBM]. Think about Jeopardy. The goal is to provide the question to the shown answer. Massive amount of human contextual knowledge. This goes well beyond the high-speed massive computing power of the 1997 Big Blue and into knowledge across a wide variety of human endeavors
Watson beat both players!
Now we come to 2016, the event and the year when “AI” truly became the next big thing in technology. The Google “AI” computer – AlphaGo- is programmed to compete in the Go game against the reigning Go champion of the world.
The board game Go has been thought of as the most complex board game ever invented. Evidently, a good bit of intuition is needed as the game has far more potential moves than chess or any other game. As Google stated before the competition, there are as many potential moves in Go as there are atoms in the Universe. Go is largely a game of encirclement and capture.
[After the U.S. lost the war in Vietnam, one of the explanations for the loss was that while America played checkers, North Vietnam played Go.]
Google set up a match between its’ DeepMind AlphaGo and the reigning Go master Lee Sedol. It was a five-game match known as the DeepMind Challenge Match. DeepMind won 4-1. Wikipedia has a really good summary of the match and it’s significance
A historical footnote is that, in the year prior to this competition, it was thought that it would take “AI” a good decade, if not several decades, to get to the level of being able to beat Go Grandmasters. So the 2016 victory showed how dynamically “AI” had been, and continues to be, the fastest-developing technology in human history.
Back to “Move 37” coming into my inbox.
The competition was held in a room with judges and two players. One player was making the moves that AlphaGo directed him to make and the other was Lee Sedol. When, I believe in the second game, AlphaGo made the 37th move, Lee Sedol pushed back his chair and asked the judges for his right to take a break and leave the room.
Now, outside this room, there were a number of South Koreans describing the match to people looking on from all over the planet. Go aficionados from around the world were suddenly mystified by Sedol’s action. This video shows this moment, recorded live.
When Sedol was asked after his loss of this game why he asked for a break, his answer was that it was not a move that a human would make, that it was surprisingly intuitive and was a completely new, creative, and different move than he had ever seen. Google had evidently introduced some Go-specific random algorithm into AlphaGo.
This told the world that “AI” had arrived. Here is a video from someone who created a course around Move 37, which also explains why it was so significant and its ramifications of it. A great video to watch if you want to learn about both Go and “AI”. It explains how “AI” functions better than I could.
“AI” finally became “the next big thing” in technology with this match in 2016, after decades of anticipation.
The next column will explain the insights I learned right after this match
Please share with anyone who you think might like this column. Your word of mouth is so important to the growth of this newsletter. Thank you! Don’t forget to also hit the like button below as it helps with the Substack algorithm.