A book, written by Gary Kasparov, has really made me question my opinions on Artificial Intelligence (AI). The book is called Deep Thinking.
It is an interesting reflection of the two decades he has spent battling against chess computers.
Kasparov was, for many years, the greatest chess champion ever seen, until Magnus Carlson, the youngster from the Nordics, took that crown (there is an interesting Netflix doc on him, too, which I can recommend) and the last great figure of a special era for professional chess.
Kasparov based his astonishing game plans on his own mental strength and memory. He won battles because he was better prepared than his opponent. He knew the moves they would be making because he had endlessly studied them.
In fact, he already knew the next 10 moves as he was able to remember the numerous possible scenarios that might possibly play out. He was a machine, as they say. But then he was faced with a real machine.
The first leg of the now infamous matchup of man vs. machine had been played a year earlier, in 1996. Kasparov, the grandmaster, was the first to play a computer, an artificially intelligent chess brain called Deep Blue.
Kasparov won that first leg with relative ease. The second time around, in 1997, it went very differently.
The six-game match was competitive, but Kasparov was defeated. The AI performed moves that were unworldly. Moves that would never have been done by a human, according to Kasparov.
This assertion has been made time and time again. AI bucks the trend of the human player. This was repeated by Ke Jei, the Go World Champion, who, like Kasparov, was defeated by AI.
Go, being both the oldest board game in the world and one of the most played to this day was the next step up for AI, challenged to learn given how many more potential scenarios that can played out during a game when compared to chess.
The general theory around these AI champion-beaters is that they have developed an ability to digest data, and therefore analyse probability, at breathtaking speed. What are the next possible moves and which has the best chance of leading to a victory?
I mentioned how Kasparov could think a certain amount of moves ahead and thus play out certain scenarios in his mind. Deep Blue wasn’t able to do this in the early days but quickly caught up and easily overtook humans. Victory, therefore, became easy and inevitable.
Magnus Carlson, the next and remaining chess champion, is Kasparov +1. The +1 is what led Deep Blue to its famous victory. Because Carlson was of the generation who trained alongside AI, his brain has essentially been augmented by machine learning. His preparation is better simply because he has a machine’s help to guide, and influence his chess playing abilities.
There were those in Kasparov’s age who conspired to make the machines beat humans. Those machines now form the basis of all chess playing. They make humans better chess players.
Why am I saying this? Because I believe this is the only way we move forward, a belief which has been influenced by recent events for me.
Firstly, I was in a LinkedIn discussion with an Australian acquaintance of mine, Kylie Davis. She has been doing some reviews on the use of AI in the real estate space.
To quote her, “I ended up down a rabbit hole”. I replied stating it was no surprise when we at Unissu are analysing over 400 PropTech firms looking at using AI as part of their offering!
Secondly, I am 41. I was (we were) fortunate enough to be on the property ladder when you were able to make good returns on capital quickly. My wife and I reinvested in a few rental properties which I have (perhaps stupidly) held on to.
Our rentals have always been plain sailing till just recently. In two of them, we have particularly difficult situations. Tenants have caused us some issues. Both outgoing tenants have caused damage in excess of rental deposits held. In one instance, incoming tenants have then refused to move in (for various reasons which I won’t bore you with).
What is the relevance, I hear you ask?
With the rental situation, I had used technological inventory services to help me with deposit situations. They were a life saver in both situations. They were cheap and easy to set up. I was able to self-serve and be in control of what I did and what I didn’t. Technology was a huge help.
However, this situation is only really being resolved by human intervention. Humans are backing up the technological work done. In isolation, the situation wouldn’t be resolved.
Equally, these situations have not been easy. It isn’t easy on an emotional level and neither is it easy on a skills level - while I used to run lettings teams, I am way out of step with legislation. While I can read about it everywhere, the interpretation just isn’t there for me. I need a human level of input.
I have needed a service offered by humans to make sense of it all.
This brings me on to the point that Kylie was making in her post. There are so many applications for AI in real estate. She referenced an excellent piece of research from CapGemini Consulting that looked at use cases of AI in certain industries.
This should form the basis of all the areas AI can, will, and must be used in the real estate sector. However, a human element will again be needed.
And herein leads the reference from Gary Kasparov once again, to bring this article full circle. He essentially summarises his feelings but looking at how chess has evolved. And I’m paraphrasing:
In chess, computers will now always beat humans. No-one has beaten a computer at chess for a number of years, not even Magnus.
However, if you pit a human working with a computer against just the computer, the combined entity wins.
That is what you have to realise today. Humans can’t win on their own. Technology can’t either. Combine humans (even us lowly estate agents) with technology and you have a winning formula.