Australian Open 2022 Tennis Men’s single final that happened on 30-Jan-2022 between Rafael Nadal and Daniil Medvedev was a testament of endurance, skill, sportsmanship and above all a supreme battle of Tennis. While rest of the world was enjoying the outcome of the match, award ceremony, post match speeches and highlights of extreme rallies, the world of technology was pondering on the million dollar question – Is AI ready to make right decisions for future of humans?
At one point in the game, the score board showed as 6-2, 7-6, 1-0 in favor of Medvedev. It is at this point that a Win Predictor was shown to the world which is primarily an AI model and it predicted only 4% chance of Nadal win. It was this prediction that was heavily discussed and criticized after the match when world saw Nadal fighting a steep uphill during the match and coming out as Australian Open 2022 Tennis Men’s Singles champion.
There were several posts on social media criticizing that AI failed to predict this outcome, which is true. Few others celebrated the creativity and unpredictability of human intelligence that its Artificial counterpart lacked, which is also true. A section of commenters, went beyond this to judge that AI cannot be entrusted to run our lives.
Lets look at it factually, at the point of the match when AI predicted only 4% chance of victory for Nadal, below were the factors considered:
- Nadal was seeded 6 Vs Medvedev who was seeded 2.
- Nadal was playing his first Grand Slam since Aug 2021 Vs Medvedev had a great 2021 season including his first Grand Slam win
- Nadal was back from COVID and a major surgery.
- In the specific match itself, Nadal was 2 sets and a game down
- In Australian Open 110+ years history, only 5 times a player won the championship after losing first 2 sets
With these facts, even a human brain would attribute very low percentage of victory to Nadal. An AI algorithm that has been trained with millions of data sets predicted 4% chance of victory for Nadal. This is pretty close to how human Intelligence and logic would have performed in this situation too. To put things in right perspective, Win Predictor predicted right winner on several hundreds of occasions but got it wrong in several other situations as well. Which is same as Human guess or Intelligence. Even an educated human guess or intelligence can go right or wrong in a given situation. What is different though, as humans, Nadal’s fans knew few additional qualities about Nadal that constantly gave them a glimmer of hope which turned out to be true on this occasion. Clearly, Win Predictor’s AI model was unaware of these qualities of Nadal. This is the gap between Artificial Intelligence and Human Intelligence that can never be bridged and perhaps this is what several commenters on social media platform were trying to point out. They are neither wrong nor right about AI in this specific scenario.
While AI is not ready to run our lives completely and never will be as many concluded, but AI has come a long way and definitely ready for primetime in many applications/sectors. It cannot replace a human brain by any means but AI can aid or complement a human brain to provide a help, confirmation or validation of the high impact decisions being made on a daily basis. While no amount of training data sets can train an AI model to become HI, at the same time, no human brain can access or remember the amount of data that an AI model has been trained with. This is the why we can safely say, AI can complement HI and not replace HI. As long as we are able to keep AI at this sweet spot, HI and AI can co-exist to accelerate human race’s development into next gen.