In early May, IBM CEO Arvind Krishna said he wanted to replace roughly a third of back-office functions with artificial intelligence. He wasn’t the first CEO to announce this kind of change. At the moment, though, these messages are primarily directed at the stock market. The ranks below top management now have to think about how to execute these statements. They’re the ones who have to turn lofty announcements into actionable plans. Welcome to the future of work and AI.

But make no mistake: AI is coming for the jobs of white-collar workers. This is a group that has never had to fret much about technological change. They are highly educated, well-paid and the most creative cohort. Why do they need to worry now? The short answer is: productivity and efficiency. The back offices of big corporations are primarily doing clearly structured and formalised tasks with predictable outcomes. This makes them ripe for automation. It’s what we’ve done to blue-collar work for ages. Let’s remember:

We automate repetitive and structured tasks with predictable outcomes as soon as (a) we can and (b) the cost falls below that of human labour.

For much white-collar work, both (a) and (b) recently have come true. What we are going to see is a boring, slow-grinding efficiency play: automating well-defined back-office processes. Consultancies will produce heaps of slideware before implementing software, mostly in the cloud, and this will take years or maybe decades to unfold. Here we are, again, in the game of efficiency.

The big question

It’s what established companies always do first when a new technology arrives: efficiency innovation. It’s about doing more with less, freeing up resources for other uses. In theory, those uses could be market-creating innovation. But before we’ll arrive there, we’ll see the AI-ification of existing business models. This will go beyond implementing AI to replace work among today’s incumbents in the near future. Start-ups will recreate existing products and services using AI, at a fraction of the costs.

Bill Gates has identified health and education as two fields where AI can have a great impact. Ironically, these two fields were strained by the pandemic which exposed their massive weaknesses and shortcomings. In healthcare, the obvious use case for AI is dealing with the ever-increasing amount of paperwork that keeps doctors and nurses away from the patient. In education, it’s more of a hope that AI can deliver the promises that computers and the internet couldn’t.

Of course, AI isn’t limited to these fields. Many established companies will get into trouble when AI-powered start-ups eat their lunch. Just take a look at the digital revolution of the past three decades to get an idea of how that will shape up. It won’t happen to all industries at once, though. There will be low-hanging fruits for start-ups to harvest first. Over time, most companies will eventually feel the impact.

The big question for companies and workers is roughly the same: How much of their work and their value creation can AI take over in the future, and how easily and how fast can it do that?

How AI can replace work in the future

The answer will vary by industry, and by the degree of digital maturity. The higher the level of digital maturity, the easier it is to deploy AI technology. A company that already lives in the cloud can easily add AI to the mix. An online shop can integrate AI with less effort than a brick-and-mortar store. Work that’s already been digitised can be shifted smoothly to current and future AI systems.

But the biggest impact might well apply to industries with low degrees of digital maturity. If the reason for lagging behind is the complexity of the business and the huge amount of domain-specific knowledge, then AI is a possible key technology. It could be the missing piece that unlocks the digitisation of laggards, allowing them to rebuild products, companies, and industries on the basis of AI.

AI is going to digitise processes and types of work that we couldn’t digitise in the past. This will have a huge impact on white-collar workers. The first waves of digitisation were about replacing paperwork with digital forms and databases. The next waves will be about replacing entire workflows and knowledge bearers, or at least augmenting them. It will be about learning faster than the competition, to borrow a phrase from Ash Fontana’s 2021 book The AI-First Company.

After digital-first, mobile-first, and cloud-first start-ups, we can expect a wave of AI-first start-ups. They will not only recreate existing products, but also invent new ones, creating entire new markets. This AI-first world has been in the making for quite some time. But it’s still early days when we think about interfaces. Chat as an interface for generative AI is nice, but clearly isn’t the future.

New interfaces?

It reminds me of the early days of the internet before the arrival of the web. Text-driven interfaces were all the rage until the fancy world wide web took over. Or, to extend the analogy even further, think of MS-DOS before the advent of Windows. Command-line interfaces are great, but only for the right purposes. How often do you fire up your Lynx text-mode browser to access the web?

So, how will AI-powered interfaces look and feel? A few years ago, we expected voice interfaces – the likes of Alexa and Siri – to take off. So far, they haven’t. Most of our computing today is still based on touch interfaces that were introduced into the mainstream by the iPhone in 2007. Is AI going to change that? To answer this question, we need to look at the specific new combinations of hardware and software that AI could enable. The touchscreen had been there before 2007, but the iPhone hadn’t.

Beyond the efficiency play that will change how we work, AI-first start-ups will rethink and redesign much of our world, including work. And since work is what remains to be done by us mere humans, the interface matters. It matters a lot:

Automating tasks is going to be amazing for rote, straightforward work that requires no human input. But if those tasks can only be partially automated, the interface is going to be crucial.

At work, computing today is predominantly defined by the PC paradigm. Only the form factor has changed, from desktop to notebook. We have added the Blackberry and then replaced it with smartphones. What is AI going to change here?

I’ll leave the last words to Amelia Wattenberger:

I want to see more tools and fewer operated machines – we should be embracing our humanity instead of blindly improving efficiency. And that involves using our new AI technology in more deft ways than generating more content for humans to evaluate. I believe the real game changers are going to have very little to do with plain content generation. Let’s build tools that offer suggestions to help us gain clarity in our thinking, let us sculpt prose like clay by manipulating geometry in the latent space, and chain models under the hood to let us move objects (instead of pixels) in a video.

Photo by Lukas on Unsplash