Why AI-driven enterprises are the future of entrepreneurship

Artificial intelligence is helping to level the playing field between startups and established companies, a situation fueled by a new class of organization — the AI-driven enterprise — that harnesses AI across every functional business area from day one.

“AI-driven enterprises achieve phenomenal impact with far fewer resources in far less time,” said MIT Sloan School of Management senior lecturer Paul Cheek in a presentation at the most recent MIT Research and Development Conference. “Companies are becoming more efficient, and they are emerging into the marketplace and stealing market share from incumbent organizations faster than ever before.”

Organizations that deploy AI across every functional area from the start are able to develop products and acquire customers at lightning speed — with few employees and little capital outlay. It took Netflix years to get to 1 million users, but it took ChatGPT just five days, Cheek said.

“We have seen the product development life cycle shrink significantly,” said Cheek, a founder of the AI-Driven Enterprise Institute and a senior advisor at the Martin Trust Center for MIT Entrepreneurship. “What used to take years to develop a product, if we look back 50 years ago, has decreased to days, hours, minutes, and, in some cases, actually milliseconds.”

It’s no longer about just the product development process but the process of creating entirely new ventures that has accelerated, he said.

Nvidia, Amazon, Meta, and energy producer SLB rank highest in terms of the most AI-mature enterprises in the S&P 500, according to the 2026 AIDE Index, which is based on a framework developed by Cheek and AIDE Institute co-founders Felipe Scherer and Kate Reed. It tracks objective signals of AI adoption using publicly observable data such as LinkedIn profiles and posts, earnings call transcripts, job postings, patent filings, regulatory filings, and corporate communications.

Cheek explores the AI-first phenomenon in his forthcoming book, “No One Works Here: How AI-Driven Enterprises Are Dramatically Redefining Business, Leadership, and Competition.”

Here’s what Cheek believes competitors need to understand as AIDEs begin to remake entrepreneurship and business more broadly.

ARR to FTE is the new business metric to chase

Historically, startup success has often been measured by the number of employees hired and the amount of investor dollars secured. That has changed, Cheek said. Many entrepreneurs and investors measure the success of AI-driven enterprises and AI-native startups by looking at annual recurring revenue per full-time equivalent employee, which provides insight on a company’s workforce productivity.

A high ARR per FTE suggests that a company is efficient and can generate more revenue per person. Established firms or those in industries with business models that aren’t necessarily reliant on recurring revenue may substitute that metric with revenue per employee.

ARR-per-employee figures show a stark divide between fading unicorns and fast-growing AI-native companies, Cheek said. He pointed to:

• Bench, the bookkeeping and tax-advisory fintech, which generated as little as $23,000 in ARR per FTE before filing for bankruptcy in January 2025.
• Bolt.new, the browser-based AI app builder, which had $1.3 million ARR per FTE in March 2025, five months after it was founded.

“That’s an efficient business,” Cheek said of Bolt. “That’s a business that is growing at a rapid clip and bringing in significant revenue per FTE.”

In this scenario, founders whose businesses initially looked strong based on traditional metrics are increasingly getting caught off guard, Cheek said. Even those with “phenomenal metrics” in a recent seed or Series A round are finding that the bar has already moved when they want to raise additional capital.

“The goalpost has moved since they brought the last round of funding into the company. Investors are seeing far more efficient companies out in the marketplace that they would rather put their capital into,” Cheek said.

Leading the AI-Driven Organization

In person at MIT Sloan

Register Now

Your teammate, direct report, or manager might be an AI agent

Most people still think of “companies” as a group of people working together, but AIDEs are better described as a network of nodes, some human and some AI agents.

“I’m not saying we should get rid of people, but we should be redefining the word ‘organization’ to reflect what organizations really are today,” Cheek said, noting that there are already AI agents running companies that have sold real products to real people in the real world.

“We’ve now had the first unit-profitable sale of a product from an AI agent to a human. This is really important as we look at what happens next as enterprises continue to scale and how AI companies emerge into the marketplace,” he said.

People are often the most inefficient part of an organization — a deficiency that technology can help offset, he said. For example, an organization that includes AI agents alongside human head count can easily operate around the clock, said Cheek, who sketched out two scenarios for his audience:

• A human CEO who wakes up every morning, takes in data reports from the previous day, reprioritizes what they believe will advance the organization, and then delegates to their team.
• An AI CEO that continuously takes in data, reprioritizes, and delegates to other AI agents “all night, every hour, every minute, every 30 milliseconds.”

“Who would you place a bet on?” Cheek asked.

AIDEs show economic promise but also peril for incumbent competitors

AIDEs offer real advantages for founders. Smaller teams mean lower overhead and a smaller payroll, and because such companies need far less investment capital to launch and scale, entrepreneurs get to retain more control over what they’ve built.

“We don’t see these entrepreneurs giving up quite as much equity in their companies. They’re retaining more ownership, which is really exciting and a good thing for entrepreneurs,” Cheek said.

As an example, Cheek cited an investor update for a startup that had restructured around AI-driven efficiency. In under three years, the company had been able to increase revenue while decreasing its burn rate by over 85%.

In aggregate, such improvements could impact competitors and the economy more broadly.

A distributed economy with thousands of AIDEs creates less systemic risk than dependence on few large employers or sectors does, he said. If one large innovation-driven enterprise loses market share, jobs vanish. But thousands of smaller, efficient AIDEs can provide stability.

“We always think of innovation-driven enterprises as being beneficial to the economy because they create a lot of jobs, but the reality is that [situation] also creates risk in the economy,” he said. “If you had thousands of these AI-driven enterprises, you would have a much more distributed economy that has less risk, something that’s good for all of us.”

A new skill set for managers

Given that individual managers have the capability to manage far more agents than humans, Cheek believes that the workforce will transition over the next 20 years to one composed of many more agents than humans.

While this shift won’t necessarily result in long-term widespread job loss, leaders will require a different skill set to manage an agent-majority workforce, Cheek said.

Watch: AI-Driven Enterprises — The New Arithmetic of Exponential Growth

Paul Cheek is a senior lecturer at the MIT Sloan School of Management, senior advisor for entrepreneurship and artificial intelligence at the Martin Trust Center for MIT Entrepreneurship, co-founder and executive director of the AI-Driven Enterprise Institute, and founder of Entonomy. He is the author of “No One Works Here” and “Disciplined Entrepreneurship: Startup Tactics” and the recipient of MIT’s Monosson Prize for his impact on entrepreneurship education.