Anthropic recently released a compelling economic report on AI-driven job displacement. Their framework introduces a new measure they call Observed Exposure, which quantifies the tasks that AI systems like LLMs are already automating, and those they may soon target.
Key takeaways from Anthropic’s report:
While AI job displacement is a real concern, Anthropic cautions against overestimating its speed. The report points to past predictions, like offshoring, where a large percentage of U.S. jobs were expected to disappear, but most remained resilient.
Following Anthropic’s report, Morgan Stanley raised alarms about an AI breakthrough coming in 2026. With "unprecedented" computing power, AI will undergo a transformative leap. OpenAI’s GPT-5.4, for example, is already outpacing human experts in certain domains.
Sam Altman predicts we’ll see entire companies run by just 1-5 people leveraging AI to compete with larger corporations. TurboAI is a perfect example of this playbook. Two college students launched a business with just $300 that converts lecture notes to flashcards and quizzes. They’re now earning about $1 million per month with 8 million and just 13 employees.
For me, this feels like capitalism on steroids — but I mean that in the most optimistic sense. We may be entering a golden era for AI-driven entrepreneurship, where small, nimble teams can punch far above their weight class. Further, if Jevons Paradox holds true — the idea that greater efficiency doesn't just replace demand, it expands it — then AI won't just reshape work. It will create categories of work we can't yet imagine.
The real tension sits between two parallel forces: large enterprises substituting AI tokens for headcount, and tech-savvy founders launching lean startups that scale to millions with a skeleton crew. Since the U.S. workforce is roughly split between small businesses and large corporations, the net employment outcome hinges on which force moves faster. History suggests entrepreneurship wins in the long run — but the transition will be uneven, and not everyone will have equal access to the tools or training needed to adapt.
That's the part that keeps me up at night — and gets me out of bed in the morning. The AI arms race will reward those who develop the new tradecraft: knowing not just how to use AI, but how to deploy it strategically, ethically, and at scale.
At Trellist, that's exactly what we help clients do — turn AI from a vague threat into a concrete competitive advantage. Curious where you stand? Let's talk.