قالب وردپرس درنا توس
Home https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ Technology https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ NVIDIA Used Artificial Intelligence to Clone “Pac-Man”

NVIDIA Used Artificial Intelligence to Clone “Pac-Man”



Chip Design NVIDIA (NASDAQ: NVDA) will recreate the classic video game Pac-Man in an unusual way. The company has set up an artificial intelligence (AI) system known as NVIDIA GameGan, feeding it video clips and signal-control signals from 50,000 Pac-Man gaming session, and GameGan has given it four days to produce its own version.

It works. The NVIDIA research team has created a playful version of Pac-Man.

A banner displaying logos for the 40th anniversary of the corporate symbol of Pac-Man and NVIDIA.

Image source: NVIDIA.

Why is this idea unique?

You may remember how Alphabet (NASDAQ: GOOG) (NASDAQ: GOOGL) set up an AI system to play chess on its own, starting with nothing but basic rules and ending with the world’s most powerful chess computer less than 24 hours later. GameGan is similar to Alphabet’s AlphaZero, but this system creates new games instead of playing existing ones.

The formed Pac-Man the game is fully played but a bit rough around the edges. The graphics look much older-school than the original 1980s, and the unique movement patterns for each of the four ghosts are boiled in a single character.

Again, there is no human input into the mechanics of the game. The NVIDIA team has asked GameGan to study several thousand games and create its own rules to fit the input. That’s all.

The next step

Pac-Man is an early test of a more exciting ambition. Over time, GameGan should study more complex games, such as first-person shooters and real-time strategy games, and elements from a variety of games can be developed. combine to create completely new gaming experiences. Beyond that lofty goal, NVIDIA wants to use real-world data to create advanced world simulation systems.

For now, a new take Pac-Man was created by a neural network. That was a good start. A public release will follow next.




Source link