Skip to main content
Posts by LB

Exponential — first slow, then fast

Exponential is not our Normal View of the World

One of the take home mantras I got out of The Singularity is Near , Superintelligence: Paths, Dangers, Strategies and the Wait But Why article on AI was that exponential technology acceleration is hard to fathom but underpins the world in a very real way. One book that presents this idea well is a short book by Marshall Brain called Manna , it offers both sides of the potential exponential future (of course with the advent of strong AI), on one hand there is a near future where we are all essentially perpetually supported by the State while a handful of business owners use AI to fuel a set of monopolies that have a version of robots providing all the physical and creative labour needed. The other side is a kind of utopia where the idea of ownership is antiquated and technology takes us to the solar system and also to completely engaging virtual worlds — people can take their pick. Another book, part of a trilogy, The Dark Forest  — presents a far future where technology is astonishingly advanced, wireless power is the norm, we have screens on every surface of every wall and massive cities underground. The connection between these two books is the clear framing of how exponential technology growth makes a difference. The first book presents this fanciful future as ‘near’ (< 50 years) but the second book presents a similarly incredible future as far (400 years). The second book however — the world is in a predicament where technology cannot advance exponentially, only linear and this is the critical difference.

If things keep going as they are, as we think they go, new social networks will come and go, batteries will get better and screens will become larger/smaller and generally ubiquitous. The thing is that things going as they are is probably the wrong way to look at how things will go. In our day to day, week to week life we have a sense that things are changing but it is hard to comprehend this change on a bigger scale. I remember watching a documentary about a tribe in Africa (excuse the lack of specificity) that roamed the land to live off and as they did not really have any written language they communicated through stories across generations, the tribe members being interviewed conveyed their sense of the effects of climate change being felt by the family through their generations. This slow change was real to them, they felt closely the impacts of what took possibly three to four generations, this perspective over a time period spanning more than one generation is unique.

It is hard for us to perceive how things change over a long period of time without reading about it or being told stories, but these do not convey the reality in a way that we actually experience it. Experiencing technological change is the same, we know it changes but the rate at which it changes is hard to process as in a recent lifetime the major shifts have not seemed that great (Landline phones -> mobiles -> web chat -> telepresence/VR chat coming soon), but go back just a generation and a bit ago it was (letters -> landline phones) and before that there was nothing.

What is Exponential

The term exponential is a mathematical one, and as with most mathematical things they get boring in the eyes of most people very fast (exponentially even). So to help us think of things exponentially, let’s just think of ‘ first slow, then fast ’. Occasionally in conversation people will ask me about something to do with technology, either what is the next big thing or what cool stuff is coming out, probably because I usually ramble too much about these things and my friends are trying to appease me. I will then go into a part-rant about AI, self-driving cars, battery technology or energy storage in general and after the blank looks I then mention something that Facebook are doing that is cool. In general the response is ‘so what, how does that affect me’, and once we start thinking about exponential technological acceleration we can see that, it will affect me soon or just a bit after soon.

Let’s look at a few technologies and their access percentage of the population and some of my personal experiences around these.

  • Most successful kickstarter a few product iterations in — thirteen million people have access (0.19%)
  • Solid sales for a proof of concept — twenty-seven million people have access (0.38%)
  • Successful product that no one you know owns — fifty-three million bought it (0.75%)
  • That technology you may have heard about in a few buzzfeed articles — one hundred seven million (1.50%)
  • The new thing that you should get (if you had the coin to buy it) — two hundred fourteen million (3%)
  • That time you owned an iPad and everyone asked you about it in cafes, yes this happened to me, a friend of a friend owns it — four hundred twenty-eight million (6%)
  • Remember when your friend got their first smartphone — eight hundred fifty-five million (12%)
  • At least one person in your family owns a fitbit (or similar) — one billion seven hundred million (24%)
  • If you don’t regularly use voice to control a device of some kind, your spouse will — three billion four hundred million (48%)
  • Everyone has this thing — six billion eight hundred million (98%)

Depending on where you live, the distribution of technology may be different but you can hopefully recognise some of those indicators like something being a seemingly too expensive go to everyone you know seems to own one.

The scale above is exponential, every step represents a doubling of access for the entire globe and many things we use today have jumped each step in an equal timespan.

Telephone lines available (assuming this means access per person) jumped each step every thirteen years.

Internet users has been jumping each step about every four years.

Mobile phone subscribers have jumped this step about every two years.

Whatever is next may even jump this step every twelve months, or less.

All of these started out slow, then suddenly they were essentially everywhere and this trend is most likely going to increase. The first seven steps seem almost inconsequential, not even one billion people have access to the technology, then three steps later — everyone has access.

Why Does It Matter

In summary, the real question is why does it matter to me and I think that the key understanding is that this rate of change itself is not just happening at a consistent pace, it too was happening slow and is about to start happening really fast.

This rate of increased access, let’s call it inverse half life — the rate at which reasonable access to a technology doubles — is itself increasing.

The inverse half life of Internet access was 208 weeks, mobile phone access was 104 weeks, the next big communications shift will be 52 weeks and after that only 26 weeks. This rate of change is hard to fathom and might not even be possible to cope with but it is something to be aware of, prepared for and potentially investigated if you want to continue to utilize technology in the future.

For further reading I strongly recommend you subscribe to the Azeem Azhar: The Exponential View email list for some epic reads.

Research Spreadsheet for reference.