"Power comes not from knowledge kept but from knowledge shared," Bill Gates said in 1999 — the same year OpenAI CEO Sam Altman turned 14. "A company's values and reward system should reflect that idea." Yet across the business world, the opposite tendency persists. Academics Catherine Connelly and David Zweig even gave it a name: "knowledge hiding." Inside many organizations, knowing what others do not still confers advantage, and information hoarding can be treated as a competitive skill.

The question is whether such behavior can survive the AI era. Sir Martin Sorrell, founder of digital media agency S4 Capital, believes it cannot.

Before launching S4 Capital — built around a digital-only, tech-first model in contrast to the legacy agency holding companies he once led — Sorrell founded the media giant WPP and served as the longest-tenured CEO of any FTSE 100 company, holding the role for 33 years. At 81, he brings a long-term perspective to technological disruption, remembering a time when fax machines were cutting-edge.

Democratizing Knowledge Inside Organizations

"I think the most fascinating area [in the development of AI] is what I call democratization of knowledge," Sorrell said. "AI flattens the organization. In a multi-branded, multi-vertical, multi-siloed company, people develop power bases by controlling information. If information is available to everybody, if you give them access, it changes the dynamics. Good people who share will be the real kings and queens."

Even kings and queens need jobs. In 2024, Altman suggested that 95% of marketing roles — including content and campaign creation — could eventually be replaced by AI. Sorrell does not expect the revolution to be that sweeping, but he sees significant efficiency gains on the horizon.

"I don't think 95% will go, but some will go," he said of an industry that remains surprisingly manual in its execution. "That's pretty tough, because your interest as an agency was to prolong [projects] for as long as possible, and to make the campaign in the most expensive location possible. You don't have to do that anymore. Essentially making ads is cheaper, and we have to change the model to be cheaper and more efficient." His comments come as the major advertising holding companies — including his former employer WPP, along with Publicis, Omnicom, and IPG — are all racing to embed generative AI into creative production, media planning, and back-office operations.

A Comparison to Asset Management

Sorrell draws a parallel between advertising and asset management, where computerized workflows have long been the standard rather than the exception.

"Do you think that Larry Fink does, or his people do, manual spreadsheets, or semi-manual or semi-automatic?" Sorrell asked. "No, it's all automated. If you think about machine trading on Wall Street, it's a very high proportion. In our business we have 250,000 people, mainly 25-year-old media planners and buyers, and they're doing their work semi-manually, often wholly manually. That can't work — in a hyperscaler-driven world it is going to be done algorithmically."

By hyperscalers, Sorrell refers to the giant cloud and AI infrastructure providers — Amazon, Google, and Microsoft — whose computing platforms increasingly underpin how digital advertising is targeted, measured, and delivered.

He noted that AI-generated content is now widely accepted even in the automotive sector, an industry he described as "as picky as it gets." "You can now do extraordinary things with technology," he said. "And we've got to get used to doing it."

Public Anxiety and the Case for Optimism

Growing evidence suggests public sentiment is turning against AI amid concerns over global security, personal data protection, and job security. In June, a Reuters/Ipsos poll found that half of American adults fear the rise of AI could cost them or someone in their household a job. College graduates expressed more concern than other groups.

"I think we have to look at what's happening in a positive way," Sorrell said, recalling a conversation from the early 2000s with the principal of one of Mexico's leading technology universities.

"It was a university of about 30,000 kids and the principal said to me: 'It's amazing what the internet has done to Mexican villages,'" Sorrell recounted. "They didn't have libraries, but today they have access to the internet, and the marginal cost of information is zero. And those kids now have access to information sources."

"If you think about AI, which is like the internet on steroids, the marginal cost of information — not just in a company, but inside government, inside universities, to kids — is zero and we can do things with these tools that we never could do before."

Distributed Work and Generational Shifts

Sorrell argued that the Covid-19 pandemic ushered in an era of "distributed working," in which anyone with a laptop can work from virtually anywhere. The youngest generations have entered the workforce at a time when working remotely is not just possible but expected.

"A reporter interviewed somebody who was running a firm with a lot of Gen Z employees, and she said: 'It's effing difficult,'" Sorrell recounted. "And I think she was very brave to say it, because it's quite a difficult thing to say, particularly in this environment, but I think it's dead true."

"I think it's to do with a different attitude developed through Covid, developed through distributed working," he continued. "It is much more difficult to manage that than it is a workforce of 200,000 who come into offices on a regular basis. And part of the reason for the reaction that we've seen from JPMorgan Chase, WPP — 'You have to come back' — is because the people that run the business are uncomfortable."

Those mandates were part of a broader 2023–2024 wave in which large employers across finance, tech, and media tightened in-person attendance requirements after years of flexible arrangements.

"In a business like ours, it does depend to a large degree on interface and interaction," he added. "The point about the young people you recruit is how can they understand the culture if they're 55 miles away?"

Keynes Was Right — Just a Century Early

In his 1930 essay Economic Possibilities for our Grandchildren, the economist John Maynard Keynes predicted that the great challenge of the future would be "how to occupy our leisure." Elon Musk has echoed similar themes, describing a coming "age of abundance" in which super-productive AI eliminates the need for widespread labor.

Sorrell, who studied economics at Cambridge, believes Keynes' essay deserves renewed attention. "My basic thesis now is that Keynes was right, but he was 100 years too early."

The AI revolution, in his view, could finally deliver on Keynes' vision of a 15-hour workweek.