After attracting a host of leading AI researchers from companies such as OpenAI, Google, and Apple, and offering them compensation packages worth hundreds of millions (with one case reportedly reaching a billion dollars), Meta co-founder and CEO Mark Zuckerberg is elaborating on his vision for “superintelligence,” a move that has stirred the tech world.

In a recently published plain text note online (full text below), Zuck explains: “In recent months, we’ve started to see our AI systems gradually enhance themselves. Although the progress is currently slow, it is undeniable. The development of superintelligence is now within reach.”

He further elaborates on his and Meta’s concept of superintelligence, emphasizing its personalization — a hallmark of Meta’s entire suite of products like Facebook, Instagram, WhatsApp, Threads, as well as its AR glasses and VR headsets, all of which offer users a degree of customization and personalized content.

What stands out to me, however, is the contrast Zuck makes with “others in the industry who advocate for superintelligence to be centrally focused on automating all valuable work.” This appears to be a subtle critique aimed at his competitor and recruitment target OpenAI, which defines artificial general intelligence (AGI) — a precursor to superintelligence — as “highly autonomous systems that surpass humans in most economically significant work.”

AI Scaling Hits Its Limits

Power limitations, increasing token costs, and inference delays are transforming enterprise AI. Join our exclusive salon to learn how top teams are:

• Turning energy into a strategic advantage

• Designing efficient inference for real throughput improvements

• Unlocking competitive ROI with sustainable AI systems

Secure your spot to stay ahead: https://bit.ly/4mwGngO

OpenAI co-founder and CEO Sam Altman recently mentioned at a Washington, D.C. conference that AI could lead to entire job categories being “completely, completely eliminated,” citing customer service as an area where automated AI systems might dominate.

In contrast, Zuck presents an alternative vision:

“At Meta, we hold the belief that individuals pursuing their personal goals have always been the driving force behind progress in prosperity, science, health, and culture. This will continue to be crucial in the future as well…This decade appears to be pivotal in determining the trajectory of this technology, and whether superintelligence will serve as a tool for personal empowerment or become a force aimed at replacing significant segments of society. Meta is committed to developing personal superintelligence that empowers everyone.”

In a video shared with the note on X and other social platforms, Zuck states: “At Meta, we are committed to placing the power of superintelligence in the hands of individuals, allowing them to direct it toward what holds value in their lives. While some of this will involve enhancing productivity, much of it may have a more personal focus.”

Interestingly, this vision closely aligns with the one Altman shared on his personal website almost a year ago: “The transition won’t be immediate, but soon we’ll collaborate with AI that enables us to achieve far more than we could without it; eventually, we’ll each have a personal AI team, comprised of virtual experts in various fields, working together to create almost anything we can envision.”

Just last month, Altman also wrote on his blog:

“The entire industry (not just OpenAI) is crafting a brain for the world. It will be highly personalized and accessible for everyone; we will be limited only by innovative ideas. For years, technical professionals in the startup sector have mocked “the idea guys,” individuals with ideas seeking a team to realize them. It now seems their moment is arriving.

OpenAI has evolved into many things, but fundamentally, we remain a superintelligence research company. We have substantial work ahead, yet much of our path is now illuminated, with the shadows swiftly diminishing.“

Thus, these competing visions of superintelligence may indeed be more aligned than opposed. Explore Zuck’s complete note below:

Personal Superintelligence

In recent months, we’ve begun to witness our AI systems enhancing themselves. The progress is gradual but undeniable. The prospect of developing superintelligence is now within view.

It seems evident that in the coming years, AI will enhance all our existing systems and enable the creation and discovery of new, currently unimaginable innovations. The direction we take with superintelligence remains an open question.

In some respects, this heralds a new era for humanity, but in others, it’s a continuation of historical trends. As recently as 200 years ago, 90% of people were farmers growing food for survival. Technological advancements have progressively liberated much of humanity to focus less on subsistence and more on chosen pursuits. At each stage, people have leveraged newfound productivity to achieve previously unattainable goals, advancing science and health, while dedicating more time to creativity, culture, relationships, and enjoying life.

I am highly optimistic that superintelligence will help humanity accelerate our progress. However, perhaps even more crucial is that superintelligence has the potential to initiate a new era of personal empowerment, granting individuals greater agency to improve the world in their chosen directions.

While the abundance AI might one day produce is profound, a more significant impact on our lives will likely stem from each person having a personal superintelligence to help achieve their goals, create desired changes in the world, embark on any adventure, be a better friend to those they cherish, and grow into the person they aspire to be.

Meta’s vision is to make personal superintelligence accessible to everyone. We believe in empowering individuals to direct this power toward what they value in their lives.

This differs from others in the industry who advocate for superintelligence to be centrally focused on automating all valuable work, leaving humanity to live off its output. At Meta, we believe that individuals pursuing their personal aspirations have always driven progress in prosperity, science, health, and culture. This will continue to be vital in the future as well.

The intersection of technology and human life is Meta’s focus, and this will only grow in importance in the future.

If current trends persist, you’d anticipate people spending less time in productivity software and more on creation and connection. Personal superintelligence that knows us deeply, understands our goals, and aids in achieving them will be the most valuable. Personal devices like glasses that comprehend our context by seeing what we see, hearing what we hear, and interacting with us throughout the day will become our primary computing devices.

We believe the benefits of superintelligence should be shared globally as broadly as possible. Nonetheless, superintelligence will introduce unique safety challenges. We must rigorously mitigate these risks and carefully decide what to open source. Still, we believe building a free society requires aiming to empower people as much as possible.

This decade seems poised to be decisive in determining the path this technology will take, and whether superintelligence will serve as a tool for personal empowerment or focus on replacing large segments of society.

Meta is deeply committed to developing personal superintelligence that empowers everyone. We possess the resources and expertise to construct the vast infrastructure needed and the capability and determination to deliver new technology to billions through our products. I am eager to direct Meta’s efforts toward building this future.

– Mark