Saturday, December 7, 2019

What if Bandwidth were Free?

The "original insight" for Microsoft was the question: "What if computing computing were free?" It might have seemed a ludicrous question. Young Bill Gates reportedly asked himself what his business would look like if hardware were free, an astounding assumption at the time, when owning a computer was an impossibility, as they cost several millions of dollars. 

In 1970, a computer cost perhaps $4.6 million, as there only were mainframes.

“The mainframe we played tic-tac-toe on in 1968, like most computers of that time, was a tempermental monster that lived in a climate-controlled cocoon,” Gates wrote in his book The Road Ahead. “When i was in high school, it cost about $40 an hour to access a time-shared computer using a teletype.”

When Micro-Soft was founded, Gates concluded that the cost of computers would drop so much that the cost of the hardware was not a barrier to using them. In turn, that meant a huge business could be built supplying software for those computers. But nobody should minimize the near-crazy assumptions made at the time: that million-dollar computers would become so cheap that the cost of computing was nearly free. 

Prices dropped by orders of magnitude in the personal computer era, however, confirming the original insight by Gates. In 1972, an HP personal computer cost more than $500,000. In inflation-adjusted terms, an Apple II computer in 1977 cost $5,174, for example. 

In 2004, Gates argued that “10 years out, in terms of actual hardware costs you can almost think of hardware as being free. I’m not saying it will be absolutely free, but in terms of the power of the servers, the power of the network will not be a limiting factor.”

You might argue that is a position Gates adopted recently. Others would argue that has been foundational in his thinking since Micro-soft (the original spelling of what became Microsoft) was a tiny company based in Albuquerque, New Mexico in 1975. But prices did, in fact, tumble. 

Microsoft's newer "insight question" was: "What if digital communication were free?" It's the same scenario, only this time it applies to the capacity to move data--audio and video as well as text--from one point to another. 

Communications industry executives hate the idea, but facts tend to support the notion that the cost of using bandwidth keeps dropping to the point where there is almost no barrier to using it. 
As Intel CEO Andy Grove once famously said, "If you think PC prices have plummeted, wait till you see what happens to bandwidth. 

As much as telecom executives might rue the observation, bandwidth is approaching the point where its use does not impede creation and use of applications, no matter how bandwidth-intensive they might be. 

Among the biggest problems telecom service providers face is that connectivity prices in the digital era have shown a “disturbing” tendency to drop relentlessly lower, in some cases even trending towards zero. Consider the price of mobile service, which has dipped since 1997 by about half, while prices for other products have increased 100 percent to 200 percent. 


Also, some products sold by internet service providers, including entertainment video subscriptions, virtually require very-low bandwidth costs. According to Cisco, 80 percent of global traffic now consists of video. And that means virtually all networks must be designed to carry entertainment video at very low cost per bit. 

Bandwidth is not “free.” But it is affordable, and constantly getting more affordable. The key business implication is that internet bandwidth costs are, and will be, low enough so that use of internet apps is not impeded. 

Computing hardware, though not free, is no longer a barrier to widespread use. Neither is bandwidth. Though the trend has not yet reached ubitquity, that is the direction. A good analogy is electricity. It is affordable, not free, but costs are reasonable enough devices and applications based on its consumption are plentiful. 

So it will be with bandwidth.

No comments:

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...