A Computer Scientist Is Training an AI Robot To Be the Most Impressive Pokémon Competitor
“I just started this for fun and I want to take it as far as I still find joy out of making it.”
In the present decade, the tech world has seen great advancements when building AIs. When it comes to AI robots, society has seen them beat the best of the best gamers whether it be in chess or esports. Now, a computer scientist by the name of Albert III is hoping to create an AI that has the ability to be the most competitive Pokémon player.
Albert recently posted a video explaining how it works to create a bot that takes all the information available about players of the game while extrapolating the data and all the multiple possibilities for how the game will play out. The AI then has the ability to sift through the data and choose the option that will lead to the greatest number of bets outcomes all within 15 seconds with each turn. Currently, the AI is set to beat all the players but the very best human competitors.
Prior to Albert’s AI, other projects have also tried their hand at beating games such as Pokémon. One of the earlier examples includes Technical Machine which only supported Pokémon up to Generation 4. However, in comparison to Albert’s it was unable to actually tell if it was successful. A comment on Reddit stated, “Technical Machine at its smartest was still leagues worse than a normal player.”
The Verge reported that currently, Albert’s goal with his AI is to ensure that it is “playable in actual Pokémon games.” He hinted that “he wanted to make something that could tie in with the releases of Brilliant Diamond and Shining Pearl.” One of Albert’s most recent successes is using the AI to beat the Sword and Shield final boss.
It remains to be seen how far Albert will take this project.
In other gaming news, Bethesda has released a new Todd Howard featurette for Starfield.
