A Manifesto Meant to Calm an AI Backlash
A recent Verge essay sharply criticizes Mark Zuckerberg’s AI manifesto, arguing that Meta’s vision of the future treats relationships, hobbies, creativity, and even family life as problems to be optimized.
The article centers on a roughly 6,500-word essay from Zuckerberg that attempts to describe a positive future for AI. The timing matters: AI companies are facing growing public unease over data centers, electricity costs, job disruption, and broader social anger. The Verge piece notes that Anthropic CEO Dario Amodei has warned of “unusually painful” job losses across multiple industries, and that OpenAI CEO Sam Altman’s home has been targeted with a Molotov cocktail and later gunfire.
In that context, Zuckerberg’s essay is framed less as a neutral forecast and more as an attempt to reassure the public that AI will improve everyday life rather than hollow it out.
Personal Agents and the Question of Attention
One of Zuckerberg’s central claims is that everyone will have an exceptionally capable personal agent that understands their goals and works around the clock to improve relationships, health, career, finances, home management, hobbies, and more.
The Verge critique focuses on the inclusion of relationships and hobbies. A personal AI agent may be able to recommend a birthday gift, schedule a call, or predict someone’s preferences. But the author argues that it cannot replace the act of personally paying attention. In human relationships, the time spent thinking about another person is not just a means to an outcome; it is part of the relationship itself.
That distinction is central to the essay’s broader point: AI may optimize decisions, but it cannot automatically preserve the meaning created by effort, care, and presence.
When Hobbies Become Productivity Systems
The article highlights one example from Zuckerberg’s vision: using AI to choose a personalized recipe to bake with his daughter. The criticism is not that AI cannot recommend a useful recipe. It is that the process of choosing may itself matter. A parent might think about what a child enjoys, what skills they can learn, or whether a family recipe could open a conversation about childhood and memory.
Even a failed recipe could become a shared story. By contrast, an AI-optimized choice risks making the activity feel more like successful execution than time spent together.
The same logic applies to hobbies. Reading a book is not the same as receiving a summary. Knitting is not merely a way to obtain a scarf. Climbing, running, yoga, or other physical activities are meaningful because the person participates in them directly. The Verge author argues that if AI turns hobbies into ways to “accomplish more,” it may weaken the very relaxation, immersion, and self-formation that make hobbies valuable.
Creation Tools Are Not a Substitute for Practice
The essay also extends this critique to AI-generated creation. It opens with an anecdote about an AI-made motivational poster, using it to ask what people are supposed to respond to when an output contains little visible effort, skill, or personal expression. Even imperfect human-made art carries the maker’s time, attention, and motor skill.
That does not mean AI cannot be a creative tool. The point is to distinguish between obtaining an output and going through the process of making something. People already have many tools for creative expression; what remains difficult is practice, taste, failure, and improvement over time.
For example, the pleasure of making a video is not only the final output. It can include filming with friends, experimenting with technique, learning from mistakes, and later seeing how one’s eye has improved. Faster production does not necessarily create deeper creative satisfaction.
The same caution applies to productivity and business use cases. More tools may help more ideas become products, but they do not guarantee success. More competition could make success harder, and reliability, governance, and deployment risks remain open questions for powerful AI systems.
Industry Takeaway: AI Needs a Better Answer Than Efficiency
The deeper issue raised by the critique is not whether AI can be useful. It clearly can be. The question is what kind of life AI companies are trying to promote.
If the future of AI is described mainly as smoother execution, higher output, and constant assistance, it may fail to address why people value slow, personal, imperfect activities in the first place. Relationships, learning, creativity, and leisure are not always bottlenecks to be removed.
For AI companies, the next challenge is not only building stronger models or more capable agents. It is explaining how these systems can support human attention, care, and practice without replacing the very experiences that give those activities meaning.