What The Hitchhiker’s Guide to the Galaxy can teach us about Artificial Intelligence

We are repeatedly told that AI will replace millions of jobs.

Doctors. Lawyers. Accountants. Writers. Designers. Programmers. Salespeople. Managers.

But there is a problem with this prediction.

AI being capable of performing a task does not mean a business will eliminate the person who performs it.

A job is rarely one task.

A salesperson might use AI to research a prospect, prepare a proposal or summarise a meeting. But they still have to understand the customer, build trust, navigate the relationship and make good decisions.

A manager might use AI to prepare reports, analyse information or draft communications. But they still have to lead their people, make decisions, resolve problems and take responsibility for the outcome.

A trainer might use AI to research a subject or create learning materials. But they still have to understand the people they are training and adapt the learning when it is not working.

A consultant might use AI to research a market, analyse information or prepare a first draft. But they still have to understand the client, interpret what the information means, provide advice and accept responsibility for it.

AI can perform parts of a job without being capable of replacing the person who performs the job.

There is another question receiving far less attention:

What does AI really cost?

Today’s AI can appear remarkably cheap.

But behind a few pounds a month are:

• Chips
• Data centres
• Electricity
• Cooling and water
• Networks
• Data
• Initial and ongoing training
• Research and development
• Security
• Compliance
• Human expertise
• Replacement infrastructure
• Financing

And then there is the cost of actually using it.

AI produces the answer.

A human checks it.

A human corrects it.

A human validates it.

A human remains accountable.

Automation is not automatically productivity.

If AI makes producing reports ten times cheaper and the organisation responds by producing ten times as many reports, it has increased output.

It has not necessarily increased value.

Then there is the energy

Scaling AI requires not only more electricity, but additional generation and grid capacity where existing infrastructure cannot meet the demand.

So we need to ask:

Who gets the energy when demand exceeds available capacity?

Homes?

Hospitals?

Factories?

Transport?

Or AI data centres?

And there is another question.

Estimates suggest that around $1 trillion or more is being committed to AI infrastructure and development, with substantially more investment expected.

That money has to generate a return.

Today’s AI price tells us what providers are charging to build the market. It does not necessarily tell us what AI will need to cost when the market has to deliver a return on the trillions invested in it.

If AI costs 5, 10 or 20 times more, will businesses continue using it indiscriminately, or will usage become restricted to applications where the measurable value justifies the cost?

Someone ultimately bears the cost.

And today’s low AI prices should not be assumed to last forever.

We have watched the digital economy develop before:

Build users → create habits → create dependency → monetise → increase prices → introduce restrictions and premium services.

AI providers will ultimately have to operate within the same commercial realities as every other technology business.

So, should we panic about AI taking our jobs?

No.

But we should pay attention.

Some tasks will disappear.

Some roles will shrink.

Some organisations will need fewer people.

But many predictions confuse what AI can technically do with what businesses will actually find economically, operationally and strategically worthwhile to automate.

The real question is not:

“Can AI do it?”

It is:

“Should we use AI to do it?”

And then:

“Does it create more value than its total cost and risk?”

That is strategy.

AI is not a strategy.

It is a strategic capability.

So, borrowing the advice from The Hitchhiker’s Guide to the Galaxy:

DON’T PANIC.

Understand it. Question it. Use it. Measure it.

And above all:

AI should be an amplifier of human capability, not a machine that creates human dependency.

Do not follow the hype. Follow the economics.

Learn more at: www.mavenist.co.uk