The Question Every CEO Needs to Answer
Peter Diamandis and Salim Ismail , who wrote Exponential Organizations , are now asking every executive the same question:
Could a two or three person team with Claude disrupt a major line of your business in 60 to 90 days?
If the honest answer is yes, the follow-on question is: what are you doing about it?
When they wrote Exponential Organizations, they did not anticipate how quickly their framework would prove prescient. Now, watching agentic AI arrive, they believe every organisation faces a structural choice: retool around intelligence, or get disrupted by someone who has.
The Architecture That Has to Change
Every major organisational structure built over the last century was designed around hierarchy. Information flows up. Decisions flow down. Management layers exist to filter, interpret, and route.
That architecture was built for a world where human judgment was the bottleneck. In a world where AI agents can handle the routine judgment calls, the architecture becomes friction rather than function.
"If you don't retool your organisation, you will be disrupted. Because someone doing it is going to eat your lunch."
The model they are proposing: build around intelligence, not hierarchy. Agentic workflows that execute without requiring the management filter at each step. Decision-making pushed down to the edge, where the agents and the humans working with them can act without waiting for the chain.
The Four Moats That Survive
If two people with Claude can replicate your business in 60 to 90 days, what stops them? Diamandis and Ismail identify four durable moats:
Proprietary data. Information you have that cannot be replicated elsewhere. Historical transaction data, specific customer relationships, operational knowledge built over years. This is the clearest moat , the more specific and irreplaceable the data, the harder it is to replicate from scratch.
Regulatory moats. Licences, certifications, compliance infrastructure. These can be eroded over time, but they hold in the near term. Healthcare, finance, and defence all have regulatory layers that slow disruption even from well-funded AI-native competitors.
Learning loop advantage. "If you can learn faster than everybody else, nobody is going to catch you." The companies with the most advanced AI deployment are building feedback loops that improve their systems faster than competitors can match. Claude's training loop and GPT's training loop are already ahead of most alternatives , and the gap compounds.
Brand and purpose. The emotional connection with end users that goes beyond functional utility. If your customers trust you and identify with you, that trust is not easily replicated by a two-person team with better AI tools.
The Timeline Is Shorter Than Most Plans
"It's going to restructure how every company, every industry is being run. Not in five or ten years. In the next one year, in the next two years at most."
That claim is contested. But the underlying dynamic , that agentic AI changes the unit economics of starting and scaling a business faster than most organisations are moving to respond , is not particularly controversial among people who have deployed these systems in production.
The question is not whether this happens. It is whether your organisation figures it out before the two-person team running Claude figures out your most profitable line of business.