Steampunk rock  Smarter working is already possible. But too many think it science fiction

Factories were consequently built with reinforced ceilings and beams to support the immense weight, with the machines clustered around these rotating shafts, the sole source of power. Out of necessity, this whole gargantuan apparatus, a vast concatenation of shafts and belts, rotated continuously, regardless of how many machines were actually in use at any one time.

This was obviously inefficient. It was also almost unbelievably dangerous. In 1900 alone, just shy of half a million workers were maimed in industrial accidents in the United States. Around 35,000 people died.

When factory owners first electrified factories, they simply replaced the single steam engine with a single, equally enormous electric motor. This brought little advantage. Nor did safety markedly improve. Whether driven by steam or electricity, machinery remained tethered to a single power source, the whole arrangement of activity dictated by the constraints of the drive shaft.

It took a surprisingly long time – several decades in fact – before people awoke to the fact that this seemingly standard factory layout was no longer necessary. The reason was simple: whereas small steam engines were hopelessly inefficient, small electric motors worked perfectly well and could be scaled to the nature of the specific job in hand – or turned off independently when not needed. This difference opened up entirely new possibilities that couldn’t be realised within the constraints of the old framework.

Ultimately, manufacturers completely restructured their operations around distributed power. Power could be delivered exactly where and when it was needed through wires, rather than moving shafts. This enabled factories to be reorganised around production-line logic, rather than driveshaft logic, creating more spacious buildings with better workflow and
safety conditions.

Rethink and reimagine

This restructuring required a complete rethink: from the architecture and the production process, to entire management systems. What this lesson ultimately demonstrates is that revolutionary technologies often require equally revolutionary thinking to deliver on their promise. Benefits often don’t come from simply shoehorning some new technology into old systems, but from reimagining the entire process around whatever the novel technology newly makes possible.

I am sure many people will use this story as an analogy for the adoption of artificial intelligence within businesses. Don’t worry, I’m not going to do that. I suspect this has occurred to you already. And, for what it’s worth, yes, I think you are quite right to see a lot of parallels here – only this time it is the distribution of information, rather than motive power, which needs to be revolutionised.

But what I am going to suggest instead is that the story has much to teach us about how business institutions react to many other, less hyped technologies. And that the lesson we must learn from this story is as much about psychology as it is about engineering.

There seem to be two separate mental frames which govern our attitude to a new invention – naïve enthusiasm or excessive scepticism. If an invention immediately saves money in the short-term we often deploy it too hastily, simply justifying the investment on the basis of cost-saving without paying full attention to the longer-term opportunity costs. Supermarkets overinvested in self-checkout technology, without fully considering that this made it almost impossible for anyone to do a large weekly shop; open-plan offices cram staffers into a smaller footprint, but often stifle the very collaboration they promise to promote.

Alternatively, many of the most promising innovations require a great deal of upfront expense and experimentation. Hence, they are subject to much greater foot-dragging scrutiny and scepticism – but it is these more difficult innovations that present us with the greatest long-term opportunities. The biggest innovations generally require a great deal of behaviour change – and humans find behaviour change painful.

The reason for that 30-year delay in properly implementing the electrification of factories was not only a question of technical implementation, difficult though that was. It was also a problem of perception and behaviour. Looked at through the narrow lens of short-term gains, the first manifestation of a new technology is often slightly disappointing in some respects when compared to the incumbent technology. Its ultimate upside potential only emerges when people imagine it within an altogether new reality – and change their behaviour in response.

I have had repeated experience of this psychological phenomenon, in part because I’m 60 years old. Thus, I’m old enough to remember people saying things like: “Why would I want a mobile phone? I mean, what’s the big deal about being able to make a phone call while standing in the street?”

At a time when most people’s only conception of an out-of-home phone call was using a payphone in a telephone box, it was hard to envisage the ultimate possibilities offered by cellular communication. Moreover, if you were an early adopter of the cellphone, there was a protracted period when most other people did not themselves have a cellphone, so most of your calls were still to (or from) landlines. The freedom to call someone en route to meet you from a moving train  was still some years off and, until you experienced it, hard to understand. It was a value that had to be imagined before it was experienced first-hand.

The long view

Likewise, a century and a half earlier, when Rowland Hill introduced the penny post, the service lost money for the first few years. This was not because the service was not a good idea . It was a brilliant idea, in part inspired by Charles Babbage, one of the foremost mathematicians of the age. He was able to use his talents to calculate the remarkable economics made possible when the newly invented steam train could carry huge volumes of letters over trunk routes, with the resulting network effects making distance almost irrelevant to the cost of delivery.

Instead, the problem was psychological: at the time of its introduction, many people simply could not conceive of how their lives could change when they could routinely communicate overnight with people hundreds of miles away. Indeed, many people at the time simply confined their exchanges to people within a short walk of their home, because they could not envisage a world where they could do anything else.

The distinction you need to understand here is between short-term optimality and long-term optionality. For instance, the current debate around the adoption of electric cars often focuses on which is currently ‘better’. But comparing an electric car of 2025 with a petrol-engined car of 2025 is somewhat beside the point. The more important distinction is that, over the next 20 years, the potential for innovation around an electric powertrain – an exploration space which evolutionary biologists call ‘the adjacent possible’ – is inordinately larger than for the internal combustion engine. Miniaturised combustion engines (like miniature steam engines) don’t really work; miniature electric motors do.

The Lime bike, the e-cargo bike, the Heathrow Pod, the Waymo cab and indeed the drone are all early examples of new adjacencies. We must now try to imagine a world where there are hundreds more.

But there is one technology – and again I don’t mean AI – where we seem completely trapped in a steam-age mentality. I am referring to videoconferencing. Here, people seem completely straitjacketed in a mindset where the downsides are endlessly exaggerated while the upside possibilities are discounted.

Innovation ecosystems

Let me make my position clear. I do believe there is a huge benefit in many ways to face-to-face human collaboration, and to the serendipitous connections which emerge when people occupy the same space. But not everywhere, and not all the bloody time. Back in the 1960s, adman David Ogilvy confessed that he never wrote anything of significance in the office; he went home. The office “contained too many distractions”, he said. It is clearly to the advantage of knowledge workers everywhere to have some degree of autonomy over where they best perform different kinds of work.

But that is only one part of the possible upside. No one has asked the really big questions of high-quality videoconferencing or webinars which needed to be asked to transition from steam-power to electricity in factories.

For instance, are the people best qualified to solve this problem necessarily in the same building – or the same country? How good would a prospective chief executive be at addressing 10,000 employees simultaneously online, rather than impressing 20 stock-market analysts in a room? Is it reasonable to restrict our talent pool to people who already live in, say, London or those who can afford to move there (i.e. nobody)? Is it even ethical to do so? Should we encourage staff to take early retirement, or retain them remotely for a few days a month? Are there a significant number of people who we could never afford – or even want – to employ full time, but who could justify 50% of a salary by being on call?

Is the best way to reward staff not to pay them more, but to allow them to live somewhere with affordable housing? What are the effects on staff retention? Should we abandon traditional hierarchical reporting structures in favour of project-based, fluid team formations that are assembled virtually for specific outcomes? Can we create ‘innovation ecosystems’ that generate breakthrough ideas without physical proximity and serendipitous encounters?

If history is anything to go by, we should expect to see some really inventive answers to these questions. In around 2055.