The digital revolution, and how to spot a turning point
Follow the fall in the cost of information from the transistor to the phone in your hand, then build a five-question checklist for judging whether something happening now is a real turning point, and try it on a case.
BEFORE
When a fact had to sit somewhere
Until about seventy years ago, every piece of information occupied physical space and cost real money to move. A record was paper in a cabinet. A calculation was work: until the 1950s, computer was a job title for a person, usually a woman, who did arithmetic all day. Sending a short message across an ocean by telegraph cost a serious part of a week's wages.
The first general-purpose electronic computers of the 1940s filled rooms, used thousands of glass vacuum tubes that burned out constantly, and drew as much power as a small street.
One trick, repeated for seventy years
- 1947 · The transistor
A switch with no moving parts and no glass to burn out, made from a slice of semiconductor. Small, cheap, reliable.
- Late 1950s · The integrated circuit
Instead of wiring transistors together by hand, print many of them onto a single piece of silicon. Now the number of switches can grow without the work growing.
- 1971 · The microprocessor
A whole processor on one chip. A computer stops being a building and becomes a component you can design into anything.
- 1970s–80s · The personal computer
Cheap enough for an office, then a home. Software becomes an industry of its own.
- 1990s–2007 · The internet, then the phone
A public network joins the machines; the web makes it usable by anyone; and from 2007 the smartphone puts the whole thing in a pocket. More than five billion people are now online.
Check yourself
Sahar asks what the transistor, the chip, the internet and the smartphone really have in common underneath. Which answer gets closest?
- Each one made storing a fact, making a copy and sending a message cost far less
- They were all invented by a small number of companies in a single country
- Each was smaller than the thing before it, and smaller is always better
- They all connect people to each other, which is what people want most
Show the answer
Each one made storing a fact, making a copy and sending a message cost far less
Yes. Strip away the devices and one quantity keeps falling: what it costs to store, switch, copy and move information. That is the same quantity paper and the printing press cut in lesson 5, cut again by a far larger factor.
Start at 1. Double every two years.
2 years 2
10 years 32
20 years 1,024
30 years about 33,000
40 years about 1,000,000Forty years of quiet, steady doubling is a factor of a million.In 1965 Gordon Moore noticed that the number of transistors on a chip was roughly doubling on a regular schedule, and it went on doing so for about fifty years. Nothing else in this course improved like that: a steam engine in 1900 was perhaps a few times better than one in 1800. It is also worth saying plainly that this rate has slowed in the last decade. Nothing doubles for ever.
Check yourself
If something doubles every two years, then after forty years it is more than a million times bigger.
Show the answer
True
True, and most people guess far lower, which is one of the most common mistakes made about technology. Doubling compounds: forty years is twenty doublings, and twenty doublings is about a million times, not forty. It is why computing was underestimated for fifty years, and why the honest answer to 'what will this be like in ten years' is rarely 'slightly better than now'. The same reasoning warns you in the other direction too: the moment the doubling stops, all the forecasts built on it stop with it.
The printing press, and this
The same story
Far more voices. Truth and nonsense both spreading faster. Old gatekeepers losing control of what can be read. Authorities answering with bans and licences. And a period of confusion while people learn new habits of checking, exactly as readers had to after 1500.
Genuinely different
A copy now costs nothing, so there is no printer deciding what is worth the paper. A message crosses the world instantly rather than in months. And the system records who read what, which no printing press ever did. That last one would have been recognised at once by the temple scribes of lesson 2: records give power to whoever keeps them.
What it gave, and the bill
- Gained. Knowledge that used to need a library. A message home that costs nothing. Work, study and markets reachable from far outside the old centres. Machines that read scans, translate languages and forecast weather.
- Lost work. Typists, telephone operators, travel agents, film developers and many factory jobs are gone, exactly as the copyists of lesson 5 and the handloom weavers of lesson 8 went.
- Records. Almost everything you do now leaves a trace held by somebody else. That is the oldest theme in this course, and the newest version of it.
- Attention and rumour. A rumour has always moved faster than a correction. It has never before been able to circle the planet in an hour, aimed at the people most likely to believe it.
Check yourself
Which of these repeat the printing press, and which are new?
- Many more voices, and gatekeepers losing control
- Rumour travelling faster than the correction
- Authorities answering with bans and licences
- One more copy costing effectively nothing
- A message reaching the far side of the world at once
- The system recording who read what
Show the answer
The same as 1500: Many more voices, and gatekeepers losing control, Rumour travelling faster than the correction, Authorities answering with bans and licences
New this time: One more copy costing effectively nothing, A message reaching the far side of the world at once, The system recording who read what
Five questions for anything called revolutionary
- 1. One product, or a new input?
A hit product changes a market. A turning point is something other industries rebuild themselves around. Electricity, not the electric fan.
- 2. Has the cost of something basic collapsed, and is it still falling?
Paper, then printing, then electricity, then computing. When the price of a basic thing falls a hundredfold, people start doing things nobody would previously have considered.
- 3. Has anyone rearranged around it yet?
Expect a dull first decade. The gain arrived when factories were redesigned, not when the motor was bought. If nothing has been redesigned, it is early, not dead.
- 4. What can it do that had no substitute?
Writing solved memory among strangers. Roads and couriers solved distance. Printing solved the cost of a copy. If you cannot name the unsolved problem, be suspicious.
- 5. Who loses, and what rides along uninvited?
Every change in this course sent a bill: the copyists, the weavers, the plague, the smoke. Name the losers and the side effects before you call it progress.
The checklist on one case: cheap solar power and batteries. This is a judgement, not a fact. The checklist does not hand you the answer; it shows you where the argument is.
| The question | The case |
|---|---|
| One product, or a new input? | Electricity is an input to everything, so a cheaper way to make it is too |
| Has a cost collapsed? | Panel and battery prices have fallen by roughly ninety per cent since 2010 |
| Has anything been rearranged? | Only beginning. Grids, vehicles and buildings are being redesigned now |
| A problem with no substitute? | Power with no fuel delivered, in places no pipeline reaches |
| Who loses, what rides along? | Coal and oil regions; and mining, land use and waste shift elsewhere |
Note the third row. A silicon solar cell was built in 1954 and did almost nothing for fifty years. Lesson 5 and lesson 10 both warned you about exactly that gap.
Check yourself
A company launches a video app that gains fifty million users in a month, and everyone calls it a turning point. Which checklist question exposes the weakness of that claim fastest?
- How many users did it gain in its first month?
- Is the company profitable yet, and how fast is its revenue growing?
- Do people say it has changed their daily life for the better?
- Is it a new input other industries rebuild around, or one product?
Show the answer
Is it a new input other industries rebuild around, or one product?
Yes. Fast growth is the weakest signal on the list. Plenty of wildly popular products are only products. The test that separates a turning point from a hit is whether farming, medicine, banking and teaching end up rebuilt around it, the way they were around cheap paper, cheap power and cheap computing.
The habits this course was really teaching
- The four questions. What was life like before, why did it change when it did, what followed including who paid, and how do we know?
- Causes. Big changes have several. A condition shared by places that never met explains more than a lone genius, and a trigger is not a cause.
- Consequences. Look hard for the ones nobody planned. The web that carried silk carried plague; the road built for a king's couriers carried an invader.
- Sources. Ask who made this record, for whom, and who left none. A dull office ledger usually beats a grand story.
- Time. Name the window you are judging by. Industry reads as a disaster in 1830 and a triumph in 1900, from the same evidence.
Check yourself
Final review: match each situation to the habit that handles it
Show the answer
- A friend says a new tool will change everything within a year → Adoption lags invention; expect a slow, dull start
- A report covers only what one office chose to write down → Ask who made the source, why, and who is missing
- The same idea appears in two places that never met → Look for a shared condition, not borrowing or luck
- A change is presented to you as progress → Ask progress for whom, and at what cost
- One dramatic event is blamed for a huge outcome → Separate the trigger from the loaded conditions
Course recap
- The digital revolution is the cost of storing, copying and moving information falling further than any earlier technology brought it.
- Like printing it gave more voices, faster nonsense and a fight over control. Unlike printing, a copy costs nothing and the system remembers who read it.
- Doubling compounds: forty years of it is a million times, which is why this change kept outrunning the forecasts, and why it will stop outrunning them when the doubling stops.
- To judge a change now: one product or a new input, has a cost collapsed, has anything been rearranged, what problem had no substitute, and who pays?
- And keep the four questions from lesson 1, because they work on the present as well as the past: what came before, why now, who pays, and how do we know?