• by Roger van der Spek , Op-Ed Contributor, 3 hours ago

Humans won 2025 -- but only just.

Artificial intelligence (AI) is gaining critical momentum for next year.

Advertising executives may have afforded themselves a wry grin as 2025 draws to a close with the annual reminder that the "holidays are coming." 

At the end of a year where much of the daily discourse centered on whether AI really can replace humans, the sight of a fleet of red trucks in Coca-Cola's AI-generated Christmas ads was reassuring, as were the mistakes.

As observers have noted, the computer-generated vehicles seem to change in length as they approach the viewer. They also gain or lose wheels as they progress across the screen, and there is even concern that one appears to be about to hit pedestrians.

Coca-Cola's response is that it is testing out what AI can do and any inconsistencies are all part of its learning journey.

They are right, and it's only a matter of time before AI does get it right.

This is why the ad came toward the end of a year in which a flurry of agencies and brands have revealed that they are doubling down on AI investments.

Coca-Cola was one of the highest-profile names, as was Kraft Heinz. Mondelez went so far as to say that its $40m investment in an AI tool, developed with Publicis Groupe and Accenture, will slash ad-production costs by 30% to 50%.

It's the paradox that underlines advertising's year. AI is both so promising that it is attracting huge investments from brands and their agencies, and yet at the same time, executives are questioning its efficacy.

Emotion beats algorithms, humans say.

The Super Bowl ads summed up the mood well. Emotional messages that resonated topped the Ad Meter charts. The Clydesdales took top spot for Budweiser, as the horse written off as being too small for the job made sure a mislaid barrel of beer got to the bar in time.

A very similar story saw “The Little Farmer” planting a runaway potato that would one day grow to make Lay's potato chips.

Comedy also won big for beer brands Michelob Ultra and Stella as a pickleball game got serious and David Beckham met his long-lost brother, Matt Damon.

Conversely, the Ad Meter suggested that the losers -- such as Tubi's Z-Suite and Open AI -- had failed because they were built for algorithms rather than humans.

It's hard to disagree. Perhaps you have to be a Gen Z to "get" the Tubi ad and the Open AI has a great narrative, but the dot matrix graphics do make it appear that story yarn is being told by one computer to another.

I have to add that my own analysis of 2025's biggest ad story -- the "Sydney Sweeney has good jeans" campaign -- bears out the truth behind human creativity.

In AI predictive eye testing, the creative failed to attract the eye to the American Eagle branding because the gaze stays mostly with the face and body of a well-known celebrity. And yet brand sales went through the roof. We all know why. They went for a Benetton-style controversy by stirring the pot on eugenics.

The discussion that followed gave the ad momentum it never would have had unless a team of humans realized that sparking a huge furor was the best way to sell jeans.

What about the agency story?

However, here is the real paradox that cannot go unnoticed. Throughout the year, reporting suggested change was afoot in advertising. Big was out and little and agile is in. The David v Goliath fight is on. The reason: AI.

With technology that can create ads and even plan campaigns, as well as crunch data and run back-office systems, the logic has been that the small, digital-native firms can take on the huge holding groups more effectively.

Without the drag of huge office rents, massive payrolls and the weight of historic ways of working, the small firms can prevail as the playing field is leveled by technology.

The conundrum is that we have two very different stories here to sum up the year that has been -- and they both kind of ring true. Adland observers are crediting AI with giving smaller agencies a fairer shake and yet the same technology is being labelled as creating or inspiring ads that are less successful than those created in the old way, by humans telling emotional stories. Both of 2025's big stories are true.

So here's the rub -- the lesson we should all be picking up from the year that his just concluding: yes, when it comes to the big-hitter TV ads, a human can still think more emotionally than an algorithm.

But we are talking about the top end of the scale here -- not the vast majority of everyday advertising that we are increasingly seeing being created and tested by neuromarketing AI (including predictive eye tracking).

Some campaigns are now even being planned by AI too, at least in partial experiments to test how well they perform.

So we humans might take a look at Coca-Cola's trucks and choke up on the emotion of that "little farmer" or the small Clydesdale and think we beat the algorithms in 2025.

But the story of the year is far more complex. Behind the high-profile campaigns, AI is gaining significant traction and it is taking a lot of the cost -- the billable hours -- out of the industry.

So the holding giants are going to struggle to maintain earnings levels going forward and fleet-of-foot smaller agencies may appear to be doing better, if they have embraced slimmed-down teams and offices earlier.

This year has taught us two things, then. Yes -- we're better at emotion -- and that is likely to remain the case for high-profile campaigns.

But AI is coming for the rest, for the vast majority of the work adland creates, plans and places each year that doesn't get talked about in Super Bowl and holiday advertising discussions. And that's why the old school holding agencies are going to struggle to maintain earnings.

Even so, we’ll all still probably look at Super Bowl 2026 top ads and think of their creativity as reassuring rather than a high-level distraction from a deeper, underlying trend.