Investment giant BlackRock recently dropped its midyear investment outlook and, unsurprisingly, AI was one of its chief areas of focus. Buried in the report is a fascinating chart that shows just how quickly new technologies can take hold and spread. On first look, the snap conclusion here is that it’s no wonder everyone’s abuzz about AI. Take a pen out and trace what you think the AI adoption curve will look like. My bet is that it’ll look a bit like the smartphone curve (red line). Smartphones went from zero to almost 75% penetration (that’s the percentage of US households using the technology) in around 15 years – a time span that should be within every long-term investor’s horizon.

Unfortunately the internet’s adoption curve isn’t plotted here. But according to the World Bank, it took just 17 years (1990 to 2007) for penetration to reach 75%. That’d be another reason to jump up and down about AI, then.

But this clever chart also hints at looking at AI’s adoption from a different angle. Just glance a little more left, to those earlier technologies. Now, there's one in particular that I think you should pay attention to – electricity (light pink line). I reckon AI’s adoption curve will more closely match that of electricity than any of the consumer device technologies here like smartphones or VCRs. And the reason is that companies, not people, will be doing most of the adopting, at least at the start.

What does this mean?

When a new technology comes along that does things better, it tends not to be a difficult sell, at least from a consumer standpoint: people love to have the latest gadgets. But from a corporate perspective, there can be way more reluctance. AI will almost certainly do a lot of things better, but it’ll come with huge costs, like demolishing and rebuilding manufacturing facilities, or shelling out for expensive new software. (Not to mention the moral challenge of laying off droves of loyal employees). And when there are big costs involved, companies can drag their feet.

For investors, I think there are two conclusions. The first is that those who are first out of the gate – think: Nvidia or Microsoft – will enjoy a multi-decade period of potential rewards. There’ll be ups and downs along the way, sure, but these two firms’ AI offerings could look very different in 10 or 15 years. The second is that AI is going to throw up loads of exciting opportunities in the future, some of which haven’t even been dreamed of yet.

Few of us were around in the 1920s (please reach out if you were) but my guess is electricity’s early days would have been about as exciting (probably more) as AI’s. And there’s no doubt it sparked a multi-decade period of exciting investment opportunities.