Dinosaur AI Data Centres

The news is full of stories of the massive build-out of AI data centres- focus being on issues such as energy, environment, community, resources, semiconductors and even the bond & capital markets.

Given massive capital costs coupled with scary depreciation, rapid technology obsolescence and high running costs, the economics seem to be based on wild optimism more than anything else.

Basically: you need to make a big pile of money within just a couple of years, after that, your DC asset is just a very large shed containing copper and a lot of A/C gear, with some very angry neighbours.

---

Bigger & bigger

Much of this growth is driven by ever-larger LLMs, which can do everything from write code to the best chicken jalfrezi recipe. We are beyond calling them ‘large’ language models.
I’d say FGLM is a better acronym, with the G meaning gigantic.

The prevailing approach is to throw massive amounts of processing power at the problem- measured in data centres using terawatts of power.

It’s a bit like thinking the way to win a car race is just a bigger engine.

Australian motorsport readers will remember the Morris Mini Cooper S humbled the American V8s at Bathurst. It was so embarrassing that they changed the rules to favour the muscle cars.
Then the Nissan GT-R did the same thing- but I digress.

WTF were we thinking ?

I think in a few years, these data centres will be seen as dinosaurs from an age of incredible enthusiasm, but with wildly misguided solutions.

Extending the automotive analogy, the future of AI is more the Colin Chapman philosophy of ‘simplify then add lightness’ rather than a brute-force approach of bigger models and more power.

This means the focus will shift to efficiency- so AI that runs on a sprinkle of energy on a mobile device, rather than in a datacentre needing it’s own nuclear power reactor.

Instead of relying on highly unpredictable and expensive AI token consumption- businesses will move towards cost-effective and efficient AI solutions, probably leveraging their own training data, rather than generic models which are available to anyone who pays. AI models and training data will be seen as valuable business assets and a key competitive advantage- not stuff you buy from someone else.

Rather than FGLMs that do more and more, there will be specialised, lightweight models that excel at specific tasks (a bit like humans), with a traffic controller than just sends the task to the right place.

After all, a microbe can carry out a certain amount of ‘compute’ using a miniscule amount of energy. The human brain uses something like 20 watts of power.