The next breakthrough will likely come from multimodal AI models, which are armed with multiple senses, such as the ability to use computer vision and audio to interpret things, Eck told me. The next big step will be figuring out how to build language models into other AI models as they perceive the world. This could, for example, help robots understand their surroundings through visual and language cues and voice commands.
Yann LeCun, Meta AI Chief Scientist
Generative AI will get better and better, LeCun said: “We’re going to have better ways of specifying what we want from them.” Currently, models react to prompts, but “at this point, it’s very difficult to control what the text generation system will do,” he added. In the future, he hopes, “there will be ways to change the architecture a little bit so that there is a certain level of planning that is more deliberate.”
Raia Hadsell, Research Director at DeepMind
Hadsell was also excited about multimodal generative AI systems, which combine audio, language, and vision. By adding reinforcement learning, which allows AI models to be trained by trial and error, we could see AI models with “the ability to explore, have autonomy, and interact in environments,” Hadsell told me.
What mass layoffs on Twitter mean for your AI workers
As we reported last week, Twitter may have lost more than a million users since Elon Musk took over. The firm Bot Sentinel, which tracks inauthentic behavior on Twitter by analyzing more than 3.1 million accounts and their activity daily, believes that around 877,000 accounts were disabled and another 497,000 were suspended between October 27 and November 1. . That’s more than double the usual number. .
To me, it’s clear why that happens. Users are betting that the platform will become a less fun place to hang out. That’s in part because they’ve seen Musk fire teams of people who work to ensure the platform is safe, including Twitter’s entire AI ethics team. Musk is likely to have some regrets. The company is already hiring engineers and product managers to 13 positions related to machine learning, including roles involved in privacy, platform manipulation, governance, and defending online users against terrorism, violent extremism, and coordinated harm. But we can only wonder what damage has already been done, especially with the US mid-term elections looming.