Before this blog, i was reviewing the game i played against a 1400 rated player and the most noticeable thing is the bar on the left of the chessboard on the weirdly green themed chess dot com. Stockfish, the engine evaluates the game and predicts probabilities of winning for both the players. It also gives you a complete analysis of the blunders (i made tons btw), the brilliancies and the average ass moves you made.
Stockfish is like the free tier GPT all of us can access and use.
It makes us better and some people learn quickly, some are slow and most don’t play at all. Some are stuck at 600, some are stuck at 1300, and some cheat to try to get ahead but also get caught pretty quickly.
The current state of chess on the highest level of gameplay
Openings are no longer an advantage in any game. Every opening is studied intensively and upto a depth of 30 - 40 moves. This means that a good chunk of the game is pretty much solved. All thanks to the engines, humans have to no longer think of all the possibilities, we all know what the standard openings are and what are the best lines to play.
Can a computer beat a grandmaster?
Chess engines before 1997 were taken much less seriously, however they were used by grandmasters to analyse positions and understand various lines. The legendary Garry Kasparov was crowned world champion in 1985. Bro reigned supreme but this noob, this rookie beat him. Garry had beaten him before, but this time he was back and in a 6 game rematch, he won. IBM’s Deep Blue, a computer had beaten the strongest human player, in a game we invented. Chess was the first arena in which humans fell, against the robots.
Fast forward, we‘re still playing chess right? we still have tournaments, we got Gukesh as the youngest world champion. Engines in chess have developed far beyond the understanding, at the highest depth, the moves these engines make are incomprehensible a lot of times. Chess engines are now extensively used in chess commentary. It is now so smart, that there are multiple instances where we have no understanding of why a specific move was made, such moves are impossible to be played by a human in a real game. There are moves it makes to optimise for a 0.2 or 0.3 advantage that the best grandmasters wouldn’t think of. These optimisations however have contributed so much to perfect opening theory.
If chess isn’t the best example of Humans and Super Intelligence coexisting together, i don’t know what is.
Now that we’ve established that humans and AGI existing together will neither create nirvana nor will it create 1984, let’s try to map this to our understanding of software.
The most common opinion since the release of o3, which is probably the first version of super intelligence is “Software jobs as we know it are dead”. Your CS Degree is gonna go to shit.
Credits to chessbase and this article, we have stats to understand the age and rating trends:
- The first FIDE rating list in 1971 featured only 82 grandmasters. By October 2020, this number had grown to 1,722
- In 2000, the lowest rating in the FIDE Top 100 was 2,594. By 2022, this threshold had risen to 2,650, indicating a 56-point increase.
- The average age of players in the FIDE Top 100 decreased from 33.19 years in 2000 to 31.5 years in 2020.
- In 2000, there were only three players aged 18 or younger in the Top 100. By 2022, this number had increased to five.
So what’s my prediction?
The number of software engineers being hired decreases because the barrier to entry will be so high. This will also create exceptional programmers, ones that will be 100x better than the older ones.
Experience will not be the factor that will make it better, but exposure to AI at an early age will.
This means the ones who are cozy at their job without putting the effort to stay updated with new methods, new ideas, new tools will be forgotten and pushed aside. Purist programmers will only survive if they’re the best, which is going to be much harder than using the AI tools and letting go of their ego. Will demand for software fall? nope, the hunger will increase, the suppliers however will be much better at handling it.
If you’re a software engineer who’s fearing automation and replacement, you were never a good software engineer. Imagine not having to go back and correcting the syntax error you made, while intensely typing the program of your dreams. Imagine not having to write the mundane tests. You get to dream up beautiful software while the worst parts of your job are done by someone else.