FACEIT joins Google Cloud and Google Labs (now Jigsaw) to end expressive utterances in gaming. These expressions are labeled as ‘toxic,’ which means they must be banned across the board.
Up until recently, censorship was considered an expensive and risky endeavor. On larger platforms, companies had to hire a team of properly-trained censors who can discern context to pick out forbidden expressions and attitudes. AI algorithms and machine learning make such economic concerns completely obsolete. Intangible AI bots are being trained to scour the internet and eliminate any expressions that are deemed inappropriate by the powers that be. This kind of news will become common and, eventually, not even newsworthy.
— FACEIT (@FACEIT) October 25, 2019
It always starts as celebratory and obvious. Who doesn’t want to remove infantile and tiresome insults from game chats?
This particular AI bot, called Minerva, supposedly does what you expect from human intelligence — sussing out context. As we move forward with the wide-scale implementation of this technology, the problem is two-fold:
- What exactly is “toxic”? Who decides that and it based on what criteria?
- It costs almost nothing to have AI speech bots running everywhere and expanding their purview constantly. It busts the door wide open to a totalitarian system you read about in cyberpunk dystopian novels.
Can you imagine if AI speech bots are scouring the internet 24/7 and are set to label and remove any speech that counters the establishment agenda by abeling content as hate speech, conspiracy theories, and/or toxic — all of which can extend to the entirety of human knowledge and factual expression?
Google has been caught multiple times tampering search results and engaging in mass YouTube censorship — deleting channels that break no rules but are politically inconvenient. Now, are we to celebrate Google deploying sophisticated censorship bots that delete silly insults in gaming chats?
Images courtesy of Twitter, Pixabay.