By Elena Georgiou Strouthos
Co-Founder & CTO, Cocoon Creations
Mark Zuckerberg recently published a manifesto about the future of artificial intelligence. But as the mood around AI is shifting, with people in the US protesting data centers, everybody being worried about mass unemployment and doomsday scenarios floating around his manifesto is surprisingly optimistic. He doesn’t talk about humans becoming obsolete and machines taking over the world. Instead in his vision everyone has access to what he calls “personal superintelligence”: an AI that knows us, understands what we are trying to achieve and gives ordinary people capabilities that today would be unimaginable.
And he doesn’t only describe superintelligence solving scientific problems or running businesses. At one point, he talks about something much more relatable: using AI to find recipes to cook with his daughter.
That is telling of how he wants to imagine the future. Superintelligence isn’t for the elite and it isn’t supposed to live somewhere in a laboratory. In his vision AI will help us work, learn, create and live.
He also makes an interesting argument about safety and alignment. Instead of trying to build one safe AI, you distribute many of them widely and let competing interests check each other, the way democratic institutions do. His illustrations: one person with a superintelligent lawyer is unfair; everyone with one is justice. One actor with cybersecurity superintelligence is catastrophic; universal access hardens every system.
In his manifesto he talks about a future utopia. But there is one obvious problem with the messenger.
Mark Zuckerberg is the founder, CEO and controlling shareholder of Meta and his company is spending enormous amounts of money trying to shape the future of AI.
Mark Zuckerberg’s 6500 words manifesto conveniently contains ideas that align rather well with Meta’s strategy.
He argues, for example, for widely available open-weight AI models rather than a future in which the most powerful models are closed and controlled by a handful of companies. He asks the government to keep silicon export controls and to not slow down US model releases.
Are they legitimate arguments in favor of his points? Yes. But only a truly naive person wouldn’t realize that Meta has a very strong commercial interest in the world Zuckerberg is advocating for.
But the reactions against the manifesto are related to something deeper, namely trust.
Facebook and Instagram, 2 of the main products of Meta, carry 2 decades of baggage around privacy, data collection, political influence and the wider effects of social media. When the person behind Facebook tells us that the next technology revolution will empower individuals and distribute power more widely, people are skeptical.
The reactions are not so much a criticism of the vision but rather a reflection on who is presenting it and why.
Let’s forget Meta. Let’s put aside Mark Zuckerberg’s motivations. The central premise is rather intriguing.
Imagine everyone really does get a superintelligence. Imagine that increasingly capable AI eventually means that every one of us has access to intelligence that can help us research, write, analyze, design, program, learn, plan and create.
Then what? It’s logical that people are suddenly dramatically more productive. And then? A company that employs 100 people today can do the same thing with 20 people.
Does this sound like a future utopia? Or does it sound like mass unemployment?
The central problem here is that we assume that the amount of work worth doing is fixed. We assume that if producing today’s goods and services requires fewer people, then the remaining people are unnecessary.
But history teaches a valuable lesson. Technology doesn’t just make existing things cheaper to produce but it also makes entirely new things economically possible.
Twenty-five years ago, planning a complicated trip often meant going through a travel agent. The agent had something you didn’t: access to information, booking systems and expertise. They could find flights, compare hotels, understand connections and make reservations on your behalf.
The internet put much of that capability directly into the hands of the traveler. Travel agents became obsolete, but an enormous economy grew around that change: online booking systems, comparison sites, short-term rental platforms, travel bloggers and creators, experience marketplaces, review platforms and countless businesses serving travelers in ways that couldn’t exist before.
Technology automated the travel agent by redistributing the travel agent’s capabilities to everyone but it also created new businesses and jobs that couldn’t exist before.
Marketing provides another version of the same story. Not that long ago, reaching a large audience required access to resources most small businesses simply didn’t have.
You needed advertising agencies, designers, photographers, printers and the money to buy space in newspapers and magazines or airtime on radio and television.
Then the internet (Google and Meta being a large part of it) changed the economics of distribution. A small company could build a website. Search engines allowed customers to discover it. Social media provided access to audiences that once required significant advertising budgets. Digital advertising allowed a business to target thousands of potential customers with relatively little money and know exactly what works and what doesn’t.
The old barriers disappeared but marketing didn’t disappear.
In fact, we got SEO specialists, social-media managers, content creators, performance marketers, influencers, email-marketing platforms, analytics companies and entire industries that couldn’t have existed under the old economics of advertising.
Again, technology didn’t simply make existing marketing cheaper. It made marketing accessible to millions of organizations and individuals who previously couldn’t afford to do it at all while creating hundreds of new job roles.
AI could push this phenomenon dramatically further. What if the future is not mass unemployment? Instead, now individuals are capable of building products, offering services and solving problems that previously required a large organization behind them.
That appears to be at the heart of Zuckerberg’s optimistic economic argument. If AI dramatically reduces the resources required to create something valuable, then it doesn’t only make existing companies more efficient. It lowers the cost of creating new companies altogether.
For a small economy like Cyprus, that possibility should be particularly interesting. Scale matters. Our companies operate from a small domestic market and a limited talent pool. Building highly specialized teams is difficult and international expansion is expensive. If AI genuinely reduces the importance of organizational scale, some of those disadvantages become smaller. A company in Nicosia doesn’t need to become a 1,000-person organization before it can acquire some of the capabilities of one.
But access to AI isn’t enough
Most businesses already have access to remarkably capable AI. The technology is not particularly difficult to obtain. Yet the difference in how organizations are using it is enormous.
In one company, AI might mean employees occasionally asking ChatGPT to rewrite an email. In another, it might mean redesigning an entire process that previously took 5 people several days. Both organizations technically “use AI”. Their outcomes will be completely different.
The important divide over the next few years may therefore not be between companies with AI and without AI. It may be between companies that add AI to the way they already operate and companies that are prepared to rethink how they operate around what AI suddenly makes possible.
That requires a very different question from the one many organizations are currently asking.
Not:
“Where can we add AI?”
But:
“If we were designing this company today, with the technology that now exists, would we still do things this way?”
That question is critical. Sometimes the answer will involve AI. Sometimes it will involve automation, integration or simply removing a process that no longer makes sense. And sometimes the conclusion may be that technology isn’t the problem at all.
It’s easy to criticize Mark Zuckerberg. META’s history and his past actions gives us every reason to be cautious when he promises a technological revolution that will give “power to the people”. And, yes, he has a not-so-secret motive for advocating for what he is advocating.
There is a future in which companies’ employ fewer people and wealth becomes increasingly concentrated.
There is another future in which the cost of creating a company or a product is so low because individuals gain access to expertise and capabilities that were once available only to large institutions. In this future, things that weren’t worth building because they were too expensive suddenly become worth building. People create new businesses, products, services and perhaps entirely new categories of work.
And increased productivity gives us something technology has been promising humanity for a very long time:
Time to create something new, time to learn something new, time to solve a problem nobody could previously afford to solve.
Time to go out and have fun, time to travel the world, time to cook dinner with your daughter.
Zuckerberg isn’t just choosing to be hopelessly optimistic about where artificial intelligence is taking us, he has an agenda. But it doesn’t mean we can’t choose to be optimistic about the future.
---
Elena Georgiou Strouthos
Elena is the Co-Founder & CTO at Cocoon Creations. She writes about technology, product decisions and what it takes to build in a small market.
LinkedIn
Cocoon Creations helps organizations understand how emerging technologies can create practical business value, from identifying opportunities and redesigning processes to deciding what should be built, bought or integrated. If you’re wondering what your organization would look like if you designed it today around the technology that is now available, that’s a conversation we’d be happy to have.
