AI Saved You Time. What Did You Do With It?

I get that AI isn’t for everyone and I understand that some people don’t use it out of principle. I also know that AI carries a lot of flaws – I talk about them often.

Users and non-users used to have a two-way conversation, but it’s definitely turned into a one-way lecture now.

A knife has flaws, and used in the wrong way, a knife can cause deaths, but we don’t stop people using it to chop their vegetables in the kitchen.

All that said, I recognise that there’s something slightly odd about asking AI to come up with a completely original idea.

And I do it…a lot. Most people using AI regularly probably do it. Give me ten ideas for this. Find another way of doing that. What haven’t I considered? Sometimes the answers are surprisingly good. Every now and again one appears that makes you wonder why you hadn’t thought of it yourself.

I’m increasingly interested in whether we’re confusing the ability to generate ideas with the human urge to have them.

AI has access to an extraordinary amount of what we’ve already written, discovered, designed and argued about. It can find connections across information that no individual could possibly hold in their head and it can recombine existing thinking in ways that can feel genuinely inventive.

What it doesn’t have is a stake in what happens next.

It doesn’t get irritated by the ridiculous process everyone at work has tolerated for six years. It doesn’t overhear a customer complaining and continue thinking about it on the drive home. It doesn’t wonder why nobody has ever tried something differently or have the slightly inconvenient idea that takes somebody away from what they were supposed to be doing.

People do.

This matters because businesses are currently investing heavily in AI with the expectation that it will make people more productive. In many cases, it will. Something that took four hours might take two. Research can happen faster. First drafts appear in seconds. Information that once took an afternoon to pull together can be analysed before somebody has finished their coffee.

But then what?

I think this is one of the more interesting questions around AI adoption.

We tend to assume that time saved becomes value created, but there’s nothing automatic about that. If somebody saves six hours a week using AI and those six hours are simply filled with more emails, more meetings or a higher expectation of output, the business has certainly increased its capacity. Whether it has become more innovative is another matter.

It may simply have become faster.

There’s a human behaviour problem sitting underneath this. If I find a way of completing my workload much more efficiently and my reward is even more work, I’m likely to learn something fairly quickly about announcing my next ‘efficiency’.

That doesn’t make employees resistant to AI. It makes them human.

This is one reason why I’m fascinated by what AI could mean for young people who choose to work for themselves, which is what I explored in this article.

I don’t think self-employment is an easy answer to a difficult jobs market, nor do I think everybody should become an entrepreneur. But AI is giving individuals access to capabilities that once required considerably more money or several different people.

A young freelancer can research a market, learn unfamiliar software, create materials, analyse information and develop an idea with support from AI. Increasingly capable agentic tools will extend what one person can manage even further.

Which makes me wonder about the reverse of that equation.

If one person with AI can increasingly operate with capabilities that once belonged to a much bigger organisation, what should a team of people inside an established business now be capable of?

The difference is that the freelancer has an obvious incentive to do something with the time AI releases. They might find another client, develop a new service or investigate an idea that could become another source of income. The benefit of becoming more efficient belongs, at least partly, to them.

For an employee, that relationship can be very different.

Businesses often say they want their people to innovate, but innovation can be inconvenient. It means questioning processes that already exist and spending time on ideas whose value isn’t yet known. Occasionally it means trying something that doesn’t work.

AI can create more room for those behaviours. It can’t create permission for them.

That’s why I’m not convinced that measuring AI effectiveness purely through time saved or increased output tells us very much about whether an organisation is becoming better.

Perhaps there are more interesting things to look for.

What happened to the problem everyone had been working around because nobody had time to investigate it properly? What happened to the employee who had been sitting on an idea for months? Has anyone used their new capacity to understand customers better, rethink a service or challenge an assumption about how the work needs to be done?

There’s another complication here. The more we ask AI to supply the thinking as well as help execute it, the more likely we are to circle around what already exists.

AI is exceptionally good at helping us explore established knowledge and patterns. It can challenge an idea, develop it and expose weaknesses we haven’t noticed. I use it for precisely those things.

But there’s a difference between bringing your own strange, half-formed idea to AI and saying, ‘Help me work out whether there’s something in this’, and asking AI what your next idea should be.

That distinction matters for young people entering work, but I think it matters just as much for the people already there.

Perhaps the most valuable AI skill won’t ultimately be knowing how to generate more with it. It will be having enough curiosity, comprehension and judgement to bring something worthwhile to it in the first place, then being prepared to disagree with what comes back.

That leaves businesses with a rather more complicated AI challenge than choosing the right technology.

They need people who still notice things.

People who become curious about why something isn’t working, who understand enough of the bigger picture to recognise consequences and who have retained the confidence to suggest something that isn’t already in the plan.

Then they need to create a working environment in which doing that feels worthwhile.

Over the next few years, businesses will undoubtedly become better at measuring how much time AI saves. I’m not sure that will tell us whether their investment has worked.

I’d be more interested in what their people did with the time they got back.

If that’s something your organisation is beginning to wrestle with, let’s have a (human) chat.