There's extensive discussion about how AI will reshape our economy and world. With all the buzz, there's more noise than signal in online discourse. Here is my perspective on what I believe will happen.
The General Direction
We are approaching a time when AI will surpass the intelligence of even your smartest friend. While models might not be there today, we're likely 1-2 years away from reaching this point. This means that virtually any knowledge work task will be performed faster, cheaper, and better by AI. Given that firms are incentivized to maximize profit, companies will naturally shift the majority of knowledge work from humans to AI. This will reduce both product creation costs and operational expenses by orders of magnitude, eventually resulting in cheaper services for consumers as profit margins get competed away in this hyper-competitive landscape.
Resource Allocation in an AI-Dominated Economy
The remaining challenge will be resource allocation. In a distant future, creating an app might cost just $1. Still, someone must decide where to invest each dollar to create value. While this decision-making will likely also be handled by AI, there will still be a human at the origin of that value creation chain directing the AI's efforts.
Market Fragmentation and Decentralization
An obvious implication is a dramatic increase in company creation, leading to market fragmentation. Niches and market definitions will become increasingly narrow because it will be economically viable to build highly specific products for small markets while remaining profitable. I expect profit distribution to shift from being concentrated among a few major players to being decentralized across a long tail of specialized market segments. The fundamental business principle remains unchanged: identify what people want and serve their needs better than anyone else. However, there will be orders of magnitude more companies serving ever-smaller, more focused segments. This benefits consumers, who will receive better products tailored to their specific use cases at lower costs.
Two Phases of AI Adoption
Regarding the replacement of knowledge work with AI, I foresee two phases. The first is the era of vertical AI applications—think of specialized AI for the German accounting market. Some companies will excel at serving specific verticals better than general-purpose AI and will capture value accordingly. These markets will have their own competitive dynamics. However, I believe this phase operates on borrowed time. Eventually, general-purpose AI will become so proficient across all verticals that one AI can "do it all," leading to the rapid obsolescence of vertical AI companies. Consumers will prefer having one comprehensive AI rather than switching between multiple specialized agents. I'm confident this outcome will materialize for knowledge work, though timescales may vary.
Beyond Knowledge Work
Will this mean that one general-purpose super AI captures the entire economy? Perhaps not. Knowledge work (services, software, products) represents approximately 33% of global GDP. Product creation won't disappear—consumers and businesses will still need and want to purchase physical products. Personally, about 80% of my spending goes toward products of some sort, with only 20% on knowledge work services. This gives me reason for optimism. However, there's an important caveat to consider: monopolization tendencies. While I predict increased market fragmentation and competition in product markets, the development of general-purpose AI itself remains concentrated among a few tech giants. If only one or two companies achieve the breakthrough to truly superior general-purpose AI, they could capture disproportionate value across the entire economy, regardless of how diverse the applications become. The company controlling the best general AI might effectively become the infrastructure layer for all economic activity, extracting rent from every transaction and decision made through their system. This concentration of power could offset much of the democratizing effect I described earlier.
Opportunities and Challenges
Founders and companies will be able to create better products at lower price points and serve niche markets more effectively than ever before. However, competition will intensify. I believe the more specialized your niche, the greater the potential rewards. The question I'm still contemplating is how to align organizations and work to maximize exposure to these opportunities, especially considering long-term dynamics. This reflects the nature of our economic world—disruption occurs frequently, and humanity must adapt accordingly. New fields of opportunity will emerge that aren't obvious today.
Conclusion
It's important to remain grounded in present realities. As the saying goes, "All models of the future are wrong, but some are useful." What's unquestionable is that we're experiencing a massive shift in how our economy is organized. Opportunities to solve problems for the world—and be rewarded for doing so—will always exist. While the methods for solving these problems may change, the fundamental mechanics of markets and innovation remain constant.