Meta's AI can now improve itself, says Mark Zuckerberg
Mark Zuckerberg just shared that Meta's AI is starting to "improve itself"—a big step toward smarter, self-optimizing tech.
He admits it's early days and progress is slow but real.
At the same time, he's urging caution about making advanced AI open-source, saying safety needs to come first.
Self-improving AI and its potential
Leading Meta's Superintelligence Labs, Zuckerberg imagines a future where AI-powered assistants and smart glasses help you hit personal goals—think reminders or even real-time info right in your ear.
He points out that this kind of self-improving AI is still new but being explored by others like NVIDIA and Google DeepMind.
Concerns over privacy and practicality
Zuckerberg has made waves before.
His push for "personal superintelligence" and always-on smart glasses also has people worried about privacy and whether these ideas are actually practical.