Mark Zuckerberg’s AI manifesto is exactly why people don’t like AI

On Monday, Meta CEO Mark Zuckerberg unveiled a sprawling, 6,500-word manifesto detailing his vision for the future of "personal superintelligence." While the document serves as a comprehensive roadmap for the systems currently under development at Meta AI, it also highlights a growing disconnect between Silicon Valley’s optimism and the public’s deepening skepticism toward artificial intelligence.

While elements of this essay have surfaced previously—appearing in a Wall Street Journal op-ed and during various Meta earnings calls—this latest iteration stands as the most exhaustive articulation of Zuckerberg’s philosophy. Some observers have labeled it an "anti-doomer" manifesto, but that characterization misses the mark. Instead, it is a personal declaration of why Zuckerberg believes AI will fundamentally reshape society for the better. Yet, in his attempt to paint a utopian future, he inadvertently underscores the very reasons why the public remains wary of the technology.

The Trust Deficit

To understand the friction surrounding this manifesto, one must acknowledge the current climate of the tech industry. Facebook—both as a product and as a corporate entity—remains deeply unpopular with the American public. Recent data paints a stark picture:

• 64% of Americans believe social media has negatively impacted democracy.
• A similar majority supports stricter government regulation of the industry, a sentiment that transcends traditional partisan divides.
• The company recently faced a $567 million court fine related to its impact on children.

The "vibes" are undeniably poor. While it is not suggested that Zuckerberg should retreat from the public eye, the fallout from the social media era is a primary driver of the current anxiety surrounding AI. When the public looks at tech executives, they do not see benevolent architects of the future; they see leaders who have historically struggled to mitigate the societal harms of their products. Rather than addressing this trust deficit, Zuckerberg’s manifesto seems to exacerbate it by ignoring the messy realities of how these tools are actually deployed.

Hazy Generalities and Philosophical Oversimplification

A significant portion of the essay relies on vague, high-level abstractions about intelligence—a rhetorical style reminiscent of the "free speech" arguments once championed by tech leaders like Jack Dorsey and Zuckerberg himself.

"As everyone gains more powerful tools, each person will become more capable of shaping the future, not less. … People and institutions with competing interests naturally check and balance each other to lead towards positive outcomes. The best and most realistic path to building a positive AI future is by delivering superintelligence to everyone."

While these statements sound philosophically profound, they function more as a marketing pitch for Meta’s "personal intelligence" products than as a rigorous analysis of societal dynamics. By framing these products as inevitable engines of progress, Zuckerberg sidesteps the possibility that conflicting interests often lead to negative, rather than positive, outcomes.

Out of Touch with Reality

The most jarring moments in the manifesto occur when Zuckerberg applies his theory to real-world scenarios. His vision for education, for instance, feels disconnected from the current classroom experience:

• The Vision: Every student will have a personalized, infinitely patient tutor with a PhD in every subject, leveling the playing field for those who cannot afford private instruction.
• The Reality: Consumer chatbots like ChatGPT, Claude, and Gemini already exist. While they offer learning potential, they are currently being used by students to bypass the learning process entirely—writing essays and completing homework assignments.

Because there is no industry-standard, robust watermarking system, educators are left in the dark. By failing to acknowledge that technology designed for "good" often produces unintended consequences that exacerbate existing problems, Zuckerberg risks appearing out of touch with the very people he claims to be empowering.

The Legal and Economic Paradox

Zuckerberg’s thought experiment regarding the legal system further illustrates this disconnect. He posits that if only one person had a "superintelligent lawyer," it would create an unfair advantage. However, he argues that if everyone had one, justice would be served more efficiently.

Critics might argue that this ignores the reality of bureaucratic complexity. Rather than streamlining justice, such tools could just as easily clog the courts with a new wave of "legal spam" or vexatious litigation.

Similarly, his description of a "freemium" model for compute power raises eyebrows. He suggests a dynamic auction mechanism to ensure the lowest price for users. While the theory of spot-compute markets is sound, the user experience of "surge pricing" on essential tools is notoriously poor. It is difficult to believe that a company would subject its end users to the volatility of real-time compute auctions, yet the fact that Zuckerberg presents this as a viable path forward suggests a lack of focus on the practical, human-centric side of product design.

A Lesson in Communication

There is a stark contrast between Zuckerberg’s approach and that of other industry leaders. Figures like Sam Altman (OpenAI) and Dario Amodei (Anthropic) have adopted a different strategy:

1. Direct Acknowledgment: They openly discuss the inherent dangers of AI. 2. Emphasis on Precautions: They highlight the safety measures and guardrails they are building into their systems. 3. Trust-Building: They attempt to convince the public that they are taking the risks seriously.

While communication alone cannot solve the existential risks of AI, it is a necessary component of the industry's survival. When safeguards fail—as they inevitably do—the lack of a trust-building foundation makes the industry look reckless. By refusing to engage with the risks and instead doubling down on a purely optimistic, abstract vision, Zuckerberg isn't just failing to convince his critics; he is providing them with more evidence that the people building the future of AI are not sufficiently concerned about the consequences.

Ultimately, if the goal of this 6,500-word manifesto was to win over the public, it has likely achieved the opposite. It serves as a reminder that for many, the "future" being sold by Silicon Valley feels less like a promise and more like a threat.