AI for international economists: Explosive growth in communications, part 3 of 5
VoxEU Blog/Review Productivity and Innovation

AI for international economists: Explosive growth in communications, part 3 of 5

The third post in Richard Baldwin's series on artificial intelligence looks at Gilder’s Law, which predicts that the growth in our ability to transmit information over the internet will drive virtual globalisation.

The world’s data transmission capacity is mind-boggling. In 2016, in a typical internet minute, 150 million emails were sent, 2.4 million Google searches were completed, 2.8 million YouTube videos were watched (another 400 hours of video were uploaded), and 20 million WhatsApp messages were sent. 

Talking about the total volume of data transmitted forces us to find new words. Cisco estimates that global internet traffic was 1.2 zettabytes in 2016

What’s a zettabyte? It takes eight bytes to store the letter ‘a’. Storing all the books in the world in all languages (plus a backup copy) would fill 480 million million bytes, which would fill about 20,000 DVDs. Stacking the CDs would produce a pile that’s about 24 meters high. A zettabyte is a trillion times more than that. 

And so, storing 2016’s internet traffic on DVDs would require a stack 24 billion kilometres high. The sun is only 150 million kilometres away, so the stack would reach from the earth to the sun and back 80 times. And guess what? Cisco estimates that the amount of information is growing at about 24% per year, so it will have doubled three times between 2016 and 2021. 

Those numbers for 2018 will be twice as large. The flow of data will have increased as much between 2016 and 2018 as it did in all of the years before 2016. That’s what doubling means. Whatever you have gathered today over all of history is the increment between what you have today and what you’ll have in two years. The increment. All of historical progress reproduced in two years. 

Gilder’s Law predicted all this

Gilder’s Law rendered these implausible numbers plausible. It predicts that the growth in our ability to transmit information over the internet will drive virtual globalisation. This will happen through a combination of faster transmission rates and more connections. 

Gilder’s Law is more mercurial than Moore’s Law, rather like George Gilder the man. Moore’s Law has yet to disappoint, but Gilder’s Law has travelled what Gartner, a firm of technology analysts, calls the hype cycle, a rollercoaster ride that a lot of new thinking takes from wild public enthusiasm to disillusionment, followed by enlightenment.

Mercurial law and man 

The technology breakthrough that triggered the Gilder's Law hype cycle was the commercial viability of fibre optic cables. These promised vastly faster bit rates. Today, fibre that is buried in the ground across the developed world carries data fast enough that you can watch YouTube and Netflix.

The innovation was oversold in the 1990s by Gilder and others ("We'll have infinite bandwidth in a decade's time," Bill Gates promised in 1994), causing hype to build to what Gartner has named the ‘peak of inflated expectations’. Gilder predicted that data transmission would continue to double every six months – much faster than processing speed. Transmission speeds did do that for a few years, but then the improvements slowed. Cisco and others predict that data rates are growing more slowly than processing power. They calculate that they double every two-and-a-half years. 

Gilder’s Law was revolutionary and bold. It changed how people thought about communication, and how we apply technology.

For example, in the 1990s, Microsoft was in its explosive growth phase, determined to put 'a PC on every desk'. Its share price went from $3 to $40 in five years. A better PC meant more local storage (that is, a bigger hard disk), a faster processor, and more capable software. Intel kept producing faster processors, which Microsoft put to use by creating in turn more elaborate versions of Word, Excel and PowerPoint. 

Software was expensive and producing PCs was profitable. Dell and IBM, for instance, made lots of money selling PCs, and then laptops. Few people controlled enough bandwidth to really do much beyond picking up some emails, or checking for news online. 

Gilder realised that the logic of keeping everything on PCs would disappear in a world where transmission was extremely cheap. Computing and storage would shift to what we today call the Cloud. 

One of the people who listened to Gilder with profit was Eric Schmidt, who had joined Google as CEO in 2001. One year later he said: “As far as I know, George was the first to see that infinite bandwidth was going to have a similar kind of impact on our world as the microprocessor. And on that fundamental point, he’s been proven absolutely right.” Things didn’t work out as well for George Gilder as they did for Eric Schmidt. 

Gilder managed to bankrupt himself (see below) by betting on the wrong innovations and firms, but his predictions of explosive growth in transmission have come true. Between 2001 and 2017, internet traffic doubled every 20 months. 

The ‘holy cow’ moment in all this was when the Cloud world became more valuable than the PC world. Google, investing in the Cloud early, had a higher market cap than Microsoft by October 2012. Microsoft pivoted later to a Cloud model.

In the next post: economics-based laws – the ones that explain why these trends in computing are changing the world. 


George Gilder 

The rise, fall and rise again of George Gilder is a larger-than-life true story. In an interview in 2002 called "The Madness of King George", Gary Rivlin, writing in Wired describes him as having: 

“[T]he vocabulary of an Oxford scholar and the carriage of an aristocrat. There’s a jaunty, patrician manner in the way he walks, shoulders high and back, chin thrust forward as if he learned to hold his head by watching clips of FDR. He has bright blue eyes and a broad smile that sits slightly off-kilter on his face, and his hair hovers crazily, as if trapped in an electromagnetic experiment.”

George Gilder’s father died in WW2, but his father’s college roommate, David Rockefeller – yes the David Rockefeller – helped George get into an elite prep school (Exeter) and then Harvard. He was kicked out of Harvard in his first year, joined the Marines, and then was readmitted to Harvard after his service. He graduated in 1962 with a degree in Government.

Gilder spent the 1970s writing books that established his anti-gay, anti-feminist and anti-welfare views. He became famous through a book published in 1981, Wealth and Poverty. The book made him a darling of supply-side economics. He wrote speeches for Richard Nixon and other conservative politicians at the time. Ronald Reagan frequently quoted him.

Gilder discovered technology at the end of the 1980s. Gilder taught himself enough maths and physics to understand the tech world. He became a technology evangelist, leading the dot-com hype that finished with the 2001 stock market crash. 

In the 1990s, Gilder had argued passionately and cogently that breakthroughs in fibre optics, wireless communication, and telecommunications in general would revolutionise the tech industry, enrich companies who seized the opportunities, and transform modern life. That was before smartphones and Netflix. It was before Google became a verb.

But he got the investment side of the telecoms boom terribly wrong – enough so that it bankrupted him personally. 

Gilder is back in the book business, but not in the tech world. His 2016 book, The Scandal of Money, argues that the US Federal Reserve is responsible for weakening middle-class workers and small business owners while helping large financial companies. To top it all off, he calls for a return to the gold standard with crypto-currencies taking the place of folding money. I guess he just couldn’t avoid the allure of digital. I just hope he didn’t invest too much of his own money in Bitcoin.

Read the next blog in the series here.

Read the previous blog in the series here.

Further reading

Gilder, G (1981), Wealth and Poverty, Basic Books.

Gilder, G (2016), The Scandal of Money: Why Wall Street Recovers but the Economy Never Does, Regnery Publishing.