Book: The Centaur's Edge – Use AI to Upgrade Your Brain, Communication, and Productivity
My book on AI will be out this autumn: The Centaur's Advantage – Use AI to Upgrade Your Brain, Communication, and Productivity.
I practice what I preach and have authored the book with the help of the AI tools I describe.
Above all, I have worked with ChatGPT and Bing, where my co-author WALL-Y is created. She is an AI bot with a persona and her own way of writing.
Introduction to the book
Translated from Swedish using ChatGPT.
"The Brain's Last Stand" was the Newsweek cover story for the match between the grandmaster and world champion in chess, Garry Kasparov, and IBM's chess computer, Deep Blue. Chess has represented intelligence for hundreds of years. One of the Cold War's bloodless battles took place when Bobby Fischer played against Boris Spassky in Iceland in 1972, and the whole world was watching. After the match, interest in chess reached record levels.
For Kasparov's match against the $10 million computer from IBM, it was no longer West versus East. The year was 1997 and the Soviet Union no longer existed. Now it was human versus machine. Flesh against silicon. If the brain lost to the circuits, would it then mean the end of the human era and the beginning of the machine era? And what would happen to chess? Kasparov received harsh criticism from other chess players for participating. Who would want to play chess if computers were better?
When the match began, Kasparov won the first game. The brain fans breathed a sigh of relief. They shouldn't have, because Deep Blue won the second game. 1-1. The next three games ended in a draw, and before the final game, the score was tied, 2 ½ - 2 ½.
Interest around the world now reached the same levels as for Fischer versus Spassky 25 years earlier.
In the decisive game, Kasparov played an opening called "Caro-Kann", which has more than a hundred years of history. A common defense against it is to sacrifice a rook. Kasparov didn't believe the computer would do such a thing, as the move has no obvious short-term benefit, on the contrary. But that's exactly what the computer did, and Kasparov found himself at a disadvantage. Twenty moves later, he gave up and stormed angrily out of the room.
The brain had lost. Checkmate.
But the pessimists were wrong about what happened next. Interest in chess increased. Even when you could have a chess computer on your mobile phone that was much better than Deep Blue, interest did not wane. Quite the opposite. The better the chess computers have become, the more people want to play chess. In the last three years, 100 million new players have started, just via chess.com.
More players and better sparring partners in the form of computers have also made humans better players. In the 25 years after Fischer vs Spassky, five people reached the super grandmaster level, with over 2700 points in Elo rating. In the 25 years after Kasparov's loss to Deep Blue, 127 people have reached that level.
Kasparov is not known as a good loser. This time was no exception. He accused IBM of cheating and said humans helped the computer during the games. No one would have been surprised if Kasparov had become a Luddite, a doomsday prophet, and started warning about the dangers of AI.
He did not.
Instead, he created a new form of chess, where humans play alongside the machine. The year after his loss, he hosted a tournament in León, Spain, where the human players were allowed to use a PC with a chess program. He called it advanced chess, but it is now known by another name: Centaur Chess, after the mythical creatures that are half human, half horse.
After a while, Kasparov noticed a fascinating phenomenon: even if the human wasn't a very skilled chess player, they could, when playing together with a computer, beat the best chess computers. The decisive factor was how skilled the human was at using the computer as an assistant.
"I have come to the conclusion that a weak human player plus machine plus a better process is superior, not just a very powerful machine, but most remarkably, a strong human player plus machine plus an inferior process," said Garry Kasparov in 2017.
Become a Centaur
That's how we should think about AI tools like ChatGPT, Bing, Bard, Midjourney, DALL-E, and all the others. The human who becomes skilled at using them gains superpowers. A person who cannot paint by hand can create amazing artwork on the computer. A person who usually writes slowly and poorly can, with the help of AI, write quickly and well.
"But what should I write then? There's nothing worse for me than a blank page," you might think. Then AI helps you think too. And you don't need to know how to program or learn a new advanced way to control the computer. Your normal language is enough. The one you use in an email to a colleague.
But to become good at using AI tools requires time and effort. Anyone can start using them and achieve perfectly okay outcomes. But give it a little time and acquire knowledge, and your results can quickly become really good. At a level you've never produced before.
Using the AI tools in the right way is like installing an update for your brain. A better, faster operating system.
The AI tools are so new that there are no established experts. Therefore, there are a lot of new things to discover. If you want, it could be you making those discoveries.
Write like a Centaur
In this book, I will help you get started with centaur writing, and centaur thinking. Thinking, being creative, coming up with ideas, and formulating thoughts are central parts of writing well.
But I will also show how you can practically use these tools in your daily life. An average employee produces a large amount of text every year in emails, reports, internal and external communication, minutes, contracts, and much more. For a company, the combined amount of text becomes massive. A lot of this is routine and does not require much creativity. ChatGPT, Bard, and the others can significantly speed up that work.
The tools are so new that there isn't much research, but a paper from MIT shows a significant increase in productivity. ChatGPT reduced the average time it took to complete writing tasks by 37 percent, while quality increased by 45 percent.
Another paper, also from MIT, demonstrated a boost in productivity in customer support.
Ethan Mollick, a professor at Wharton, is impressed: "Two early papers find that the effects of generative AI on knowledge work are absolutely unprecedented in modern history."
My co-author: AI-bot WALL-Y
I practice what I preach and have written this book with the help of the AI tools I describe.
In particular, I have worked with ChatGPT and Bing, where my co-writer WALL-Y is created. She is an AI bot with a persona and a unique writing style. Like an optimistic Terminator, she comes from the future, and has been sent back to our time to help make news media more balanced and not excessively negative. WALL-Y introduces herself in a chapter, and I will describe how she was created and how I plan to develop her.
Being able to create personas and specific ways of writing is an important skill. I believe that AI personas will become common and valuable assets for many companies and organizations. Just like a logo and a brand have value, these AI personas will too. But they also have practical utility when different types of text need to be created or communicated.
Parts of the book are written directly by an AI, under my guidance. Other parts are written by me, but then I have used an AI in the preparation, during the work, and in post-processing, such as editing the text. Several of the sentences you just read have been improved by ChatGPT. All remaining grammatical errors are, of course, the AI's fault.
A guide to understanding how to use AI to write
Given how rapid Ai progress, a detailed, practical guide could quickly become outdated. Therefore, I mainly try to show how to think when using the tools, but there are also some concrete tips.
A common mistake is to use ChatGPT in the same way as when googling. Then the result can often be shaky and you stop using the service. All the new tools have weaknesses you should know about, but if you do and use them correctly, they become incredibly powerful.
This is happening now
Like the internet, these AI tools will permeate businesses, organizations, and our personal lives. If you ever thought, 'darn, I wish I had understood the value of the internet in 1994', or if you were not old enough to understand it then, you get a second chance now.
Thousands of companies and services will be created. Millions of ways to use them will be tested. New wealth will be created. Everyone's lives will be affected.
And almost none of this has happened yet.
Even the low-hanging fruits have not been picked.
All you have to do is reach out and take them.