This post made me sad.

I think anyone would deserve an over-the-top solo after all that

But I have to say I'm very intrigued by the possibility of machine learning tricks boosting heat pump efficiency. Am sure there's some headroom there.

Yh I've seen claims like this before. And even real world installs hitting these numbers: share.google/qQKKg4RgOXjx...

SCOP always more valuable than COP!

Great pics

Our trawl of methane data found the area near Drax has the UK's second highest levels:

share.google/kD4B26tD1Yl8...

2/2

Heh no worries at all Kevin!

AI is an admittedly imperfect umbrella term used to refer to lots of different forms of algorithmic processing. Deep learning, machine learning, large language models etc.

LLMs and machine learning are completely different things.

We're talking about completely different things! Also, my wife is a psychologist and therapist :)

I wouldn't say they've balanced out at all - I post all the time about the problems with large language models using vast amounts of energy and so on. But it's important to recognise the benefits that other, very different forms of AI (mainly machine learning) will bring.

I was writing about machine learning ten years ago and referring to it as AI. It's just what lots of people have long called this technology. It's never been a perfect term. It's even more divisive now.

Yeah I realise people are twitchy about AI for obvious reasons but the idea we're not going to need better models and smarter grids is for the birds.

E.g.:

"Since adoption, Quartz Solar has halved forecasting errors, according to Open Climate Fix, equivalent to saving at least £30 million in imbalance cost per year, with potential to increase to £150 million per year in 2035 if solar capacity targets are met."

share.google/j5kBHrSOPKYn...

A drop in the ocean maybe but I'm planning on installing a water butt this winter for our garden at least!