How the End of Moore's Law Will Change the Economy

If reductions in the cost of computing power are no longer driving the global economy, the question becomes what will take it place - if anything is even capable of matching its impact. JL

Matthew Lynn reports in The Telegraph:

In the next two decades, the economy may not be driven by computing power in the way it was in the past five – and that will change the way it operates far more than most people yet realise. We might even need a law to describe it.

Forget the laws of supply and demand. Stop worrying about the law of diminishing returns, or any of the other iron rules you might find in the economic textbooks. In many ways, by far the most important law for the economy over the last 50 years has been a technological one - Moore’s Law, the famous prediction by one of the founders of Intel that computing power would double roughly every two years.

It has held true for a remarkably long time; the incredible power of cheap micro-chips has underpinned the internet revolution and just about every other form of technology.

There is a problem, however. According to the people running Intel now, it may have come to an end. Chips are already so powerful it just isn’t going to be possible to keep doubling their speed, at least not nearly as quickly as in the past.

Of course, there is plenty of debate about whether that is true among technologists and computing engineers. But if so – and it surely must be true one day – it is also going to have a huge impact on the economy. We will see much less innovation, and less disruption of old industries. But we should also see more inflation and, paradoxically, higher productivity as well.

It was a half century ago in 1965 that Gordon Moore put forward the proposition that the number of transistors you could squeeze on a circuit would double every two years. In fact, when he first described his “law”, he argued it would double every year, but in 1970 he revised that to two years, and forecast the process would last about a decade.

As it turned out, it has held true for a lot longer than that. With only minor blips along the way, chips have continued to get far more powerful, and a lot cheaper at an incredibly fast rate. That explains why the phone in your pocket packs a lot more punch than a mainframe from the 1980s, and why most people’s idea of hell would be to stuck on a desert island with a machine running Windows 95. We have become used to the idea of rapid advances in technology, and expect it to carry on as it has for the last 50 years. And that has generated vast wealth. Last Friday alone, Google added $67bn to it market capitalisation – the largest one day gain in history.

^{Windows 95: pure torture in the 21st century}
But nothing lasts forever. Last week, Brian Krzanich, Intel’s current chief executive, argued that the era of Moore’s Law might be coming to an end. In a discussion with analysts, he suggested that although his firm had “disproved the death” of the law many times, the next generation of micro-processors would take longer to produce. They could no longer double in power every 24 months.
Technologists will argue both about whether that is actually true – and whether it matters. A new generation of quantum or possibly bio-engineered processors might take over. Or it might well turn out that smarter software, close to artificial intelligence, becomes more important than the chips themselves.
We will see what comes out of the labs. Even so, hyper-fast advances in technology have become built into the economy. In reality, the demise of Moore’s Law would have a big impact outside computing. But like what? Here are four places to start looking:
One, there will be less innovation. The regular doubling of computing power was what lay behind the creation of a vast number of new products, from smart-phones, to internet TVs, to streaming services. Without that, your mobile would still be the size of a small house, and you’d still be choosing between four channels on your TV. But if chips develop less slowly, we will see less of that.

^{The floppy disk}
It is not just obviously tech products such as phones or tablets that will be impacted. Ever more powerful computing lay behind the vast expansion of global trade – you can’t run ‘just-in-time’ manufacturing systems using dirt cheap factories in Vietnam, or build ever more efficient robot-dominated factories, or cars that use less fuel, without processors that are continually quicker and cheaper. There won’t be as many ‘new new things’ as we have grown used to expecting.

^{A classic: Atari game Space Invaders}
Next, get ready for higher prices. The great inflation of the 1970s has gradually disappeared – indeed, these days we worry far more about deflation. Better monetary policy was one reason for that. But the power of computer chips was also part of the story. It drove the relentlessly falling prices of every kind of electronics, so that a new TV ended up costing less than a family meal out. And it lay behind the globalisation of manufacturing, which drove prices down even more. Without that, we may well move back into an economy which is prone to inflation once again – and all it will take is some cheap money to spark those flames.

^{Would your children know what this is?}
Thirdly, expect less disruption. The technology industry was very good at creating new companies very quickly – and disrupting established powers. You could see that in retailing, the media, telecoms, and right now in finance. Old established players suddenly found themselves under siege by brash new companies, and were very quickly fighting for their life.
Even if they survived, they were often much diminished. No one cares much about ITV, once a colossus of the British media industry, anymore. British Telecom has had to change shape several times. And yet, if technology slows down, so will the rate at which old industries get shaken up. The result? The existing global multi-nationals should be able to increase their lock on the market, with more stable returns, and higher profits. And that will make them a better investment.

^{A Nintendo Game Boy from the 1990s}
Finally, we might well see higher productivity. One of the oddities of the IT revolution was that, contrary to what you might expect, it didn’t do much to increase output per person. As it has advanced, we have become very bad at producing more. That might be because it generated lots of fairly pointless activity.
Or it might be because, while some new industries were created, just as many were destroyed. Whatever the explanation, there is no doubt that it happened. So if technology starts to slow down again, companies may well be able to start figuring out ways to get productivity rising again. If they can, that will improve living standards.
Of course, innovation will always continue. The one thing that entrepreneurs in a free market are very good at is working out new ways of doing things, and figuring out new products for which there will be a demand. But in the next two decades, the economy may not be driven by computing power in the way it was in the past five – and that will change the way it operates far more than most people yet realise. We might even need a law to describe it.