This week in my ongoing years of transformation series, I looked at the start of the Web–whether you want to call that 1989, 1990 or early 1991 (all have a claim to make as I discussed).
But 1989 was among the more momentous years in the 20th century due to an event which was far more significant at the time, the fall of the Berlin Wall, that signalled the beginning of the end of the Soviet Union. For those of us who lived through those months–the protests in Tiananmen Square in Beijing, the sudden and dramatic opening of the borders between Soviet block countries like East Germany and ‘The West’, and the collapse of brutal regimes like that of Nicolae Ceaușescu in Romania these events seem to redefine the world. Now they are perhaps just footnotes in history for those who didn’t experience them.
But if you had asked historians or journalists in 1989 to name the most significant event that year, I imagine you’d have got one of two answers.
The Tiananmen Square, or the fall of the Berlin Wall (at the time most would have considered the latter to be the more significant).
I do know precisely zero would have said it was an obscure computer network idea from some random guy working on contract at a high energy physics lab in Switzerland.
What’s the lesson in that? Perhaps that events which seem momentous at the time are the culmination of many years–they represent an end more than a beginning.
While the beginning of things is only detectable in hindsight, when their compounding consequences have played out. As Steve Jobs is often quoted as saying
“You can’t connect the dots looking forward; you can only connect them looking backwards. So you have to trust that the dots will somehow connect in your future.”Steve Jobs
So what lessons does that have for today?
I’d focus less on the big picture and grand visions. Whatever we imagine will happen is likely to be far further off than we might think, if it arrives at all–and like Tim O’Reilly (among others) has said ‘being early is the same as being wrong’.
Things take much longer than we would think to happen. They happen slowly, then all at once (as Hemingway put it about bankruptcy–’gradually, then suddenly’). Because change is compounding and so exponential.
Roy Amara, a Stanford computer scientist in the 1960s said “we overestimate the impact of technology in the short-term and underestimate the effect in the long run” (Bill Gates is most often attributed with a related quote, but that was decades later, and Gates has more than enough of, well, everything, so let’s give Amara the credit here).
Right now I see quite a lot of people trying to solve enormous problems with generative AI–within their organisations, or in the wider world. They are overestimating what can be done in the short term. The Wall Street Journal reported in the last day or so that
Consulting firm EY … recently completed a $1.4 billion investment into artificial intelligence … and that it would train its 400,000 employee workforce on AI.
Like what does that even mean?
Instead, I’d begin with one little thing and explore how these technologies help there.
Here’s an example from my work. Mundane, but hopefully instructive.
I rely on a very jury-rigged systems of forms, spreadsheets, and hand rolled scripts to produce the schedules for our conferences and publish them our websites.
Because we typically only used them a few times a year, I’ve never invested the time and energy to automate some of the time consuming, error prone manual aspects of the process. I’d knew it could be done, but would involve, for example, some arcane spreadsheet functions I knew nothing about. I didn’t really have a sense of how long it might take to work out what I needed to do. And I knew it would take maybe 20-30 minutes to do it the way I already did–so stayed with that.
But not only did this consume time that could be more valuably used, it meant I’d avoid doing the process until I really had to, so we updated the conference website less frequently.
And because humans are inherently more error prone, I’d introduce mistakes into the output.
Earlier this year, as I started to work more with chatGPT, I began automating those processes. At first just very simple things–filtering and sorting columns in spreadsheets with values imported from other spreadsheets. As I said nothing complex at all, but just enough effort to not really warrant working out how to do what I’d long done with copy and paste.
But once chatGPT took away the concern I’d do a bunch of work for no result, and the worry about whether it would take hours to work out, all these little steps quickly got ironed out.
I use chatGPT for a lot more than wrangling spreadsheets–from generating tweets to writing bash scripts and Python (a language I had never really used before and frankly don’t even really know), but each new use starts with a simple use case, that then compounds.
And our publication process is almost entirely automated now–with nearly 100 talks across 7 tracks at Web Directions Summit, thankfully so!
A little less imagination, a bit more action
Now I think is less the time for imagining big picture transformative uses of these technologies and more a time for just using them every day, on small simple mundane tasks. Developing an intuition for how they work. Exploring what they are good for.
There’s no need to do courses on prompt engineering, just roll up your sleeves and get started.
Start that interest compounding today–and who knows what a decade from now you’ll be achieving.