What’s next for AI | MIT Technology Review

In the United States, the Federal Trade Commission also closely monitors how companies collect data and use AI algorithms. Earlier this year, the FTC forced weight loss company Weight Watchers to destroy data and algorithms because it illegally collected data on children. In late December, Epic, which makes games like Fortnite, dodged the same fate by agreeing to a $520 million settlement. The regulator has spent this year gathering feedback on potential rules for how companies handle data and create algorithms, and Chair Lina Khan said the agency intends to protect Americans from the practices. illegal commercial surveillance and data security with “urgency and rigor”.

In China, authorities recently banned the creation of deepfakes without the subject’s consent. Through the AI ​​Act, the Europeans want to add warning signs to indicate that people are interacting with deepfakes or AI-generated images, audio or video.

All of these regulations could shape how tech companies build, use, and sell AI technologies. However, regulators must strike a delicate balance between protecting consumers and not hindering innovation, something tech lobbyists are not afraid to remind them of.

AI is a field that is developing at lightning speed, and the challenge will be to keep rules precise enough to be effective, but not so specific that they quickly become obsolete. As with EU efforts to regulate data protection, if the new laws are properly implemented, next year could usher in a long-awaited era of AI development with more respect for privacy. and fairness.

—Melissa Heikkila

Big tech could lose its grip on basic AI research

AI startups flex their muscles

Big Tech companies aren’t the only players at the forefront of AI; an open source revolution has begun to match, and sometimes surpass, what the wealthiest labs are doing.

In 2022, we saw the first major community-built multilingual language model, BLOOM, released by Hugging Face. We also saw an explosion of innovation around the Stable Diffusion text-to-image open-source AI model, which rivaled OpenAI’s DALL-E 2.