The Future of Work[Space]

If AI had a flaw (and let’s be honest, it has a few), humility tops the list. Ask it anything — even nonsense — and it answers with confidence. Never ‘I don’t know.

That’s a problem and an opportunity.

Current LLMs (and probably will be for a long time with current technology) are statistical engines – they calculate at extraordinary speed, the likelihood that the next word in this sentence will be ‘potato’.

You don’t often see ‘statistical engine potato’ together in a sentence, but intriguingly they do appear together in a number of paragraphs in a few places on the internet.

Up until I wrote this article, there are no known documents on the internet with these three words together, at least according to Google.

Statistical engine potato. It sounds like a Dadaist band name or the world’s least useful kitchen appliance. But in LLM-land, it makes perfect sense.

“Statistical engine potato” sounds delightfully absurd — like the title of an indie art installation or a whimsical AI experiment. Here are a few playful interpretations depending on what you’re aiming for:

⸻

🧠 If it’s a metaphor:

• Statistical engine = a powerful analytical or data-processing mind

• Potato = something humble, underestimated, or universally useful

Interpretation: “An unassuming powerhouse of data insight.”

(A bit like Mr Green with a couch and a calculator.)

⸻

🤖 If it’s a project name:

• Could be the codename for a quirky but brilliant AI that churns out statistical insights — maybe with a UI that looks like a potato?

• Or a statistical model designed to boil down complex data into digestible chunks.

⸻

🥔 If it’s literal:

• A potato wired up to power a tiny statistical processor (yes, a potato battery could technically light up a low-power microcontroller).

• A science fair project in the making: “Can a potato predict trends?”

It’s absurd — and revealing. Even the most ridiculous word combo will return something plausible. But that’s not a bug, it’s a feature of LLMs. Which brings us to the real opportunity…

When ChatGPT arrived in 2020, the rate at which content was being added to teh internet doubled in size. It’s the bump during 2020 in the chart below.

Since then, the total amount of content is doubling every four years — faster than we can build hard drives to store it all. In fact, the limiting factor wont be the ability to produce content, but the cost of storing it. (It’s a good time to have shares in disk drive companies).

When content is abundant, quality becomes the deciding factor. And the best quality will be measured not by the structure prose or gramatical correctness, but by the engagement that comes from truely creative ideas.

In a world where AI writes confidently about statistical engine potatoes… what will you create?

 Questions we think are worth exploring further

• Data is abundant, but what about knowledge or education?
• Knowledge and skill attainment is essentially free, but gaining competancy still has a net cost – this is an opportunity for AR (possibly the best one).