How long have you got?

Antony Slumbers / Midjourney

Meta, Microsoft, Alphabet & Amazon are set to spend $200 BILLION on AI infrastructure THIS YEAR.

The fundamental consequence will be that the price of intelligence will drop.

Leading to wider adoption & more innovation.

Leading to more efficient & effective AI solutions.

Leading to changing competitive dynamics in the tech industry & profound implications for the broader economy.

So best to pile in now, open up the big cheque book, and get ahead of the pack?

Err, no.

Because applying new AI to old operating models has a finite upside.

Electricity did nothing for the productivity of steam powered factories until they were redesigned to accomodate the capabilities of the new technology.

That process took forty years.

With AI you may have two.

Not two until you adopt any AI but two to understand how to really leverage it.

Essentially into three buckets:

1. Tasks to be automated by AI
2. Tasks to be augmented with AI
3. Tasks to remain solely human

Whilst looking for where value, skills, data, product and service will be commoditised and where enhanced by AI.

This has to be done, with the assistance of AI (of the $20 a month variety), BEFORE you go all in and commit to significant expenditure.

Sounds obvious but anecdotally there’s a lot of ‘What’s our AI Strategy - Why haven’t we launched anything yet?’ coming from C suites.

Bottom line is you don’t have much time but not so little you can’t ‘engage brain’ first.

This is as much a change management as a technology issue.