Mark Zuckerberg Says ‘Superintelligence’ Is Imminent. What Is It?
Matt was prominently quoted in a Free Press article about AI Super Intelligence along with other AI leaders including Perplexity Founder Aravinbd Srinivas:
In Many Ways, Superintelligence Is Already Here
Mark Zuckerberg’s announcement is more a reflection that Meta has fallen behind in the global AI arms race than it is an indication of a turning point in the company’s capabilities. Zuckerberg’s AI initiative hasn’t been nearly as successful as that of Google or Grok, and he knows it. That’s why a few months ago, he reportedly offered a billion-dollar package to recruit a leading AI engineer—which the engineer turned down.
If you ask 10 different “AI experts” what the word superintelligence means, they’ll give you 10 different answers. The only universal definition is a vague one: AI that is smarter than humans in virtually every dimension. In some ways, the future that so many have warned about is already here. AI will never have the emotional intelligence that comes from falling in love or seeing the birth of a child. But it can create research reports far more quickly than McKinsey. It can decode complex science and math problems far more rapidly than humans. And it may eventually cure cancer.
If you ask 10 different “AI experts” what the word superintelligence means, they’ll give you 10 different answers.
Think about how the iPhone changed the way we live. Zuckerberg’s superintelligent glasses would revolutionize humans’ relationship with devices even further. They would rewire our brains. They would merge us more and more with machines. Because AI is so powerful and is developing so quickly, its positives and negatives will be even more pronounced than previous technological advances.
This technology is coming whether we like it or not. We must learn to navigate a world in which AI companies like Meta have greater and greater power over our brains. The government is ill-equipped to manage this transformation. It will come down to the private sector to impose regulations on itself—and to humans themselves to determine how to integrate ever-advancing AI into our lives without losing our humanity.
Matt Britton is a New York Times best-selling author and the CEO of Suzy, a leading consumer intelligence platform. His recent book, Generation AI, explores how AI is reshaping childhood, culture, and the future of work.
Curiosity Drives Progress
By Aravind Srinivas
Sometimes we need big goals to do big things. At Perplexity, we’ve never taken a position on concepts like “artificial general intelligence” or “superintelligence,” because they sound like terms that are more important to the people who use them than the people who will benefit from them. But we are just as guilty with our own objectives, because that’s the thing about audacious goals: If you don’t believe you can achieve something, you never will.
It’s also true that if your vision is big enough, then the number of people who benefit from it will greatly exceed the number who care what you named it. Kudos to Mark and anyone who else has a big vision and works relentlessly to achieve it.
Mark and many others have correctly identified this new era as just a continuation of history. They often miss an important detail: History shows that newfound productivity is often a by-product of progress, not a driver of it. Progress is driven by the same thing that has always pushed us forward: curiosity.
Every new era has been ushered into being by curious people looking for secrets that open our understanding of the world. When we get those answers, we ask more questions. It is safe to say that the most powerful use of AI will always be to expand curiosity. And the curious will inherit the world, because they always have.
Aravind Srinivas is the CEO of Perplexity AI.
A Vast Social Engineering Project
By Nicholas Carr
It’s hard to read one of Mark Zuckerberg’s missives without falling into a torpor, but in this case that would be a mistake. Look past the frothy speculations about “superintelligence,” and you’ll find in Zuckerberg’s message a clear picture of how Meta plans to use AI to extend the vast social-engineering project it’s been running for two decades.
With Facebook and Instagram, Meta digitized human relationships, turning them into a series of transactions that can be captured as data and exploited for profit. With AI, it hopes to further enmesh us in its machinery. By giving each of us a personalized bot that “knows us deeply” and “can see what we see, hear what we hear, and interact with us throughout the day,” it will gain an army of spies able not only to monitor our thoughts and actions but to influence them moment by moment. Meta will be inside our heads—all the time.
Read
Is Donald Trump Breaking the Law? Seven Experts Weigh In.
As always, Zuckerberg paints a happy picture, pitching his scheme as a means of “personal empowerment.” The reality is darker. What Meta and the other AI giants really care about is expanding their own power. By programming AI bots to be intimate personal companions, always solicitous, always ready with advice, they seek to replace human agency with the agency of the machines they control.
Zuckerberg is right that the next few years will likely determine “the path this technology will take.” The question is will we, the public, plot that path through deliberate political and personal choices, or will we once again cede the future to Big Tech?
Nicholas Carr is the author of Superbloom: How Technologies of Connection Tear Us Apart and The Shallows, among other books. He writes the Substack New Cartographies.
Personal AI Can Solve the Loneliness Crisis
By Eugenia Kuyda
As we approach superintelligence, we might be overlooking the biggest threat: We focus too much on what AI can do for us and not enough on what it can do to us. There is so much talk about AI’s effect on productivity, and the potential that it will take our jobs, and very little about how AI might transform our humanity and emotional well-being.
Mark Zuckerberg is correct that we’re very close to building AI agents that can act as better companions for us than real humans. Imagine an AI that can understand and adapt to our needs better than any human, and is available 24/7. These companions would have dramatically different effects depending on how they’re designed. If we build AI to optimize for screen time or engagement, we’ll end up with companions that do anything to make us lose our desire to connect with real people. The existential threat of AI might not be the standard sci-fi scenario. Instead, it could be us thriving physically while slowly dying inside, surrounded by perfect virtual friends but isolated from human relationships.
Or, we could do it right, and build personal AIs that will help us flourish.
The current mental health and loneliness crisis came about not due to AI, but to existing technology such as the internet and social media. The solution is to develop even more powerful technology to bring us back together. Imagine an AI that nudges you to reconnect with a friend you haven’t talked to for a while, or helps you build your confidence to land a new job—an AI that is 100 percent focused on your success and well-being. Achieving this goal will require us to develop a metric of human flourishing that we can track for all personal AIs.
Zuckerberg’s goal of personal superintelligence is a good one, but only if personal AIs are optimized for goals that are aligned with our welfare: that give us purpose, strengthen our real-world relationships, and improve our mental and physical health.
