![The Museum of the Future, Dubai. Pic:](https://www.redsharknews.com/hubfs/museum of the future.jpg)
Everything’s a bit different now. It’s not that we don’t have a future any more: we just don’t have one we can predict.
There are several ways to analyse progress. And there are many ways to define progress. We won’t agree on all of them. Conventional mainstream politics would say that growth is the engine of the economy. But that kind of growth also tends to be harmful to the planet. Maybe what we should be aiming for is a zero-growth economy. I’m not qualified to say. But I have been thinking about progress and the rate of change for a long time.
Until last year, it was fairly predictable, as long as you understand exponential growth. Most people don’t: it’s not intuitive.
Consider the lilies
Here’s an example of why most people don’t get it. Imagine a lily pond, where the lily leaves double in area every day. After forty-two days, they cover exactly half the pond. How long until they cover all of the pond? It’s forty-three. Why? Because they’re doubling every day. So if there’s half the pond already covered by the leaves, it only takes one more day of doubling to cover the whole pond.
To me, that’s not surprising because I understand the math(s). But in another way, I still find it to be different to what my gut feeling says it should be. That’s why it’s hard to predict an exponential future.
But now, we have AI, and things are very different. How different? We’ll come back to that in a minute.
The consequences of getting exponential growth wrong can be severe and even deadly. The effects of getting it right can be stupendously beneficial. Most people missed the early stages of Covid-19 infection growth. But in hindsight, it’s clear that the increase was already exponential in the early days, around the beginning of 2020. Very few people noticed it because the numbers were small at the time.
You still get pretty small numbers if you keep doubling small numbers (2, 4, 8, 16…). But after another ten doublings, you get gigantic figures. And after another twenty doublings, you have figures bigger than the number of grains of sand on the planet. After another ten, you have numbers greater than the number of molecules in the known universe.
Betting on convergence
When Apple was developing the iPhone, it knew technology was growing exponentially. Other companies did, too, I’m sure. But the difference is that Apple acted on it. It knew it could build a phone when specific lines of development converged at a point in the future, and by following an exponential path, it predicted when that would be. That put it ahead of the competition, which had to play catch-up for several years.
That growth started centuries ago - but it didn’t seem like it because it was so slow. We didn’t reach the really big numbers until we had computers. Over the last twenty years, you can see exponential growth very clearly in the rise in sensor resolution. We’ve gone from SD to HD, 4K and now 8k. 4K is 4x HD, 8K is 4x 4K and 16 times HD. 8K is also around 85 times SD. It would have seemed ridiculous if you’d told your colleagues in 2000 that video would be 85 times better within 20 years. (I’m maybe stretching it a bit - I’m talking about pixel count, not the linear resolution, which is what we are more likely to notice).
And now, with AI, the growth rate is off the scale if there is a meaningful scale for it to be off.
The end of the typing pool
In the late seventies, offices started to look different. In hallowed corners of the building, you could find the very first examples of a phenomenon called a word processor. These weren’t PCs: they were dedicated, inflexible computers. At the time, hard drives cost as much as refuelling a space shuttle, and 8” floppy discs were state of the art. Employees watched in awe as perfectly-formatted, fully-justified pages of text clattered out from a daisy-wheel printer. Rather quaintly, the operator’s job title was also 'Word Processor'.
Imagine those workers’ reactions if you told them you could ask a black slab of metal and glass in your hand to take a dictation and write an article for you. Or to summarise a recent meeting. Or do optimise your resume. That was forty-five years ago. (You could say 45 years BC - before ChatGPT). Now, imagine what the future will be like in only ten years. Five? You can’t. The rate of change is so fast that it almost appears to be vertical to us. We can’t intuitively predict anything any more.
Preparing for uncertainty
But there are steps we can take to prepare for an uncertain future.
Look for trends - even if they’re vague and ill-defined. Get a feel for where technology is taking us. Understand the range and scope of AI’s potential. Make sure that even if you don’t know where the sun will rise in the future, you’re not standing with your back to it.
And then choose technology solutions that offer as wide a scope as possible. Look for scalability but, above all else, flexibility. Build open, cooperative, multi-layered platforms that you can easily reconfigure. Ensure you update each layer with the latest and even more flexible technology. Sometimes flexibility will matter more than speed.
Not every industry is as vulnerable to AI. We will always need infrastructure. If AI runs in the cloud, we will require networks to support it and servers to run it on. We will therefore still need buildings. Microphone manufacturers can feel reasonably safe (assuming that AI doesn’t become the default way to make a decent recording from a crappy one). The same goes for loudspeakers and probably lens manufacturers - although it would be surprising if those vendors didn’t start building AI-based correction and enhancement into their lenses.
Don’t fight against technology. Fight for consensual, ethical regulation. See it as an opportunity to improve our lives while accepting and mitigating the existential threat that it will get into the wrong hands and ultimately leave us disadvantaged or excluded from a new economic reality.
Above all, even more essential than understanding what a given technology does: try to figure out what it means.
Comments