Riding the AI train
Is it time to board?
In November 1995 I found myself at Finsbury Park station in North London. It was 8am in the morning and I was on my way to a conference in Central London. I’d never travelled at rush hour before, so this was a new experience. A train drew into the platform. It was crammed full of people, but somehow, when the doors opened, more people managed to squeeze onboard. It didn’t look pleasant - and there was another train due in two minutes - so I decided to wait. But when that train arrived it was just as full. This time the next train was only one minute away so I decided to wait again. And, yes, this train was even fuller than the previous two…
The current AI race is like those trains. At some point you’ve got to get on board. You can choose to wait for an empty train, but doing so means you risk missing out. I’m reminded of the saying “there are only two ways of dealing with exponential change – either too early or too late”. Start too early and you will waste some effort. But starting too late is worse – there’s a real risk of losing a competitive advantage. Of not being able to react sufficiently quickly as AI transforms our world of software development.
Moving fast requires allocating resources to investigating AI. To experimenting. To failing. There are no established patterns yet. It’s a leap into the unknown. Some of the experimenting won’t work. Some of it will get replaced over the next few years. But some of it will stick. At worst it means you don’t fall behind. At best, it might give you a competitive advantage.
It’s a commitment. Investigating AI isn’t a one-shot deal. It’s an ongoing habit. AI keeps improving. We all need to keep learning.
You might view this as an externally imposed race. And you’d be right. The pace of change is not something any of us controls. Even if we choose not to participate in the race it will affect all of us. Companies that don’t understand will get left behind. When new startups can create software for a fraction of the cost, the current incumbents will become irrelevant.
Remember Data General, Digital (DEC), Sun, SGI, Compaq, WordPerfect, WordStar? They were once leaders in their fields. But change came. They failed to keep pace. And now they are no more.
AI is exciting. It’s scary. It’s exhausting. But there is no choice but to keep pace. We need to stay relevant – for our companies, for ourselves and for society. It may be uncomfortable squeezing into that train but, if you’re not already onboard, now is the time to do so.
Thought you’d appreciate this Price Pritchett quote I came across today:
“Organizations can’t stop the world from changing. The best they can do is adapt. The smart ones change before they have to. The lucky ones manage to scramble and adjust, when push comes to shove. The rest are losers, and they become history.”
As someone who has spent a lot of time on that train line, I love the analogy. However, I’m less on board with the point you make - I think it is less black and white that you suggest. I agree that one might as well adopt (experiment with, learn from, etc) AI as soon as possible, as it may provide some advantages, and it’s not going away, But I’m not sure I agree that failing to adopt AI immediately will mean irrelevance and death.
It will vary by industry - I’m sure, for example, that there’s significant advantages to most software developers in adopting AI today, and failing to do so will create more and more competitive pressure. However, I see enough poor uses of AI to be sure that it’s going to take a while before there’s a killer app. An example that it relevant to me today is ebay - when listing an item ebay offers to have AI to create the description for an item. But as a serial buyer, they are worse than useless. All of the descriptions sound identical and tell me nothing I don’t already know. I want to know whether the item came from a smoking/non-smoking home, whether it was stored in a dry loft or damp shed, etc - none of which AI can or will tell me. And worse, I used AI to list a product I wanted to sell, despite hating it as a buyer. </rant>
And, just because there are examples of companies dying because they failed to adopt new trends - there are plenty of converse examples of them coming to them late, and still succeeding. Microsoft, internet …