Culture Will Decide When We Adopt AI

Once Upon a Time

AI won’t replace humans, but those who use AI will replace those who don’t. 

We see this opinion everywhere, but is it true? Technology takes time to reach universal adoption. When the last companies get around to it, there is no friction or risk.

When I first started working in the research department at Schroder Investment Management in 1991 there was a typing pool. Analysts hand wrote notes, walked them over to the pool and waited to edit or approve what came back. Even those who typed reports had them reformatted in the house style on a typewriter.

We had computers but not everyone used them. I remember the top ranked bank analyst in the City coming to meet me. He rolled out three sheets of graph paper sellotaped together to illustrate his estimates of HSBC’s future earnings. While we talked he rubbed a number out and redid some calculations.

Spreadsheets at the time were Lotus 1-2-3, an IBM product introduced in 1983 and phased out almost 20 years later. By then, Microsoft owned the corporate desktop operating systems market.

Schroders was a traditional institution. I had an in and out tray and there was an internal post service. It was efficient but slow. Firms that adopted email could send more messages and get them delivered faster.

A Brief History of Email

Email started out in academic, government  and tech circles, with only a handful of corporate staff having access, usually in research or IT roles. It was first introduced in corporate environments in the 1970s, but widespread adoption didn’t happen until the mid-to-late 1990s.

In the early 1990s email spread to more industries, but was the preserve of managers, executives and those in specialist roles. Many companies still relied on fax and internal memos, like the ones generated in the typing pool.

It was the introduction of Lotus Notes, Microsoft Exchange and cheaper networking that made email a standard corporate tool. By 1998-2000, surveys indicate that over 50% of office-based corporate employees in developed economies used email daily.

Then, in the early 2000s, email became near universal in white-collar environments, as questions of security and privacy were answered once and for all.

Digital Resistance

There are famous examples of companies failing because of an inability to adopt new technology. Kodak missed digital imaging because its culture was resistant to rapid IT adoption. Sears was slow to adopt e-commerce and digital customer engagement, including email marketing and CRM integration. Yet I cannot find an example of a company that went under because it refused to use email.

The late email adopters, such as small legal firms and medical practices, caved in either to customer demand or regulatory requirements. Manufacturing in parts of Japan and Germany stuck with fax until the 2010s and survived on niche customer franchises. When these businesses change hands it is simplicity itself to adopt email. It’s a zero friction tool.

Email adoption happened over a period when many other technological and market changes were happening. These include the internet boom, globalisation and Enterprise Resource Planning rollout. Companies that were behind on email were slow to adopt new technologies in general. While refusal to use email did not cause company failures, broader digital resistance did.

It is probable that AI will introduce a new way of communicating. Agents will talk to other agents, both within and between companies. They may transact with each other using digital dollars, stablecoins, or cryptocurrencies.

Companies will adapt at the pace of their customers and competitors. Adoption will be dictated by services that spring up to make things easier, as Microsoft Exchange did for email. In many cases AI will come as standard in software and run in the background.

Turning Full Circle

Investment analysts today use AI to research companies and draft reports. The smarter ones use it as a sounding board to head off objections to a thesis before they arise. Some companies claim to have replaced analysts all together, but this is unlikely to be for research that is distributed outside of the firm.

In many ways we’ve come full circle. When I started out, analysts’ reports were written for them. Their skill was interpreting the numbers and the conversations they had with company executives. This was of greater value than writing, especially in a regulated house style.

Now AI can write reports for them. It can interpret numbers but it cannot yet interpret a nod and a wink from a human. Maybe one day AI agents will tip each other off. If and when that time comes, every company employee will have adopted AI at a pace that suits their industry and culture.

Questions to Ask and Answer

1. When did I last ask clients how they like to communicate?

2. How do my customers feel about me using AI?

3. How do my staff feel about using AI?

Find out more. Hit reply to this newsletter and ask about Digital Transformation.