WHEN THE DATA TAIL WAGS THE COMMON-SENSE DOG

Posted on by

 

 

During the early part of the pandemic, the entire country was on lockdown.

Football matches were played, but crowds weren’t allowed into stadiums to watch.

One Scottish club, Inverness Caledonian Thistle, thought it had solved the problem by using artificial intelligence (AI).

They decided to broadcast their match against Ayr Utd, using AI controlled cameras.

They used a system called ‘Pixellot with inbuilt BTT (Ball Tracking Technology)’.

The data for following the ball was programmed into AI: the size, the shape, the colour.

So the cameras could lock onto the ball and track it throughout the game.

Which is exactly what AI did, it tracked the data that had been fed into it.

And the cameras spent the entire match tracking the linesman’s bald head.

His bald head was the same size, colour, and shape as the data fed into AI, so AI locked onto his bald head and tracked it.

In the 57th minute, Ayr Utd went ahead via a goal from Innes Cameron, but all their supporters saw was the linesman’s bald head.

4 minutes from time, Inverness Caledonia Thistle equalised, but again all the home supporters saw was the linesman’s bald head.

As far as AI knew, it was doing exactly what it had been asked to do: tracking data.

Consequently, fans didn’t see the actual ball, or the actual match, or the actual goals.

AI stuck rigidly to the data, without being able to interpret it, because quite simply AI has no common sense.

AI researchers call this the ‘brittle’ factor, they describe it as “Lacking human understanding of real world complexities”.

We might ask what is the point of AI if it “lacks human understanding”?

Especially when we’ve got billions of humans on the planet who are skilled at exactly that.

But this is the hypnotised fixation on data over simple common-sense.

The belief that if we find enough data we will automatically solve any problem.

I find this every time I use the same numbers to illustrate a point, I’ll say:

“£20 billion is spent in the UK on all forms of advertising. Of that, 4% is remembered positively, 7% is remembered negatively, 89% isn’t noticed or remembered.”

Every time I use those numbers I get the same response from planners (sorry, strategists):

“What is the source of those statistics?”

My reply is: “If I give you a source will it change anything?”

See, I’m only a creative and the correct response for me is: “Holy shit, we’ve only got a 1 out of 10 chance of even being seen, better change the way we do ads, quick”.

Whether or not you verify those numbers, or quibble about them, the simple fact is what we all know: most people don’t notice advertising. But you don’t need me to tell you that.

Every tube carriage is filled with ads, did you notice any this morning?

Every tube platform is filled with posters, how many did you notice?

Every freesheet and newspaper is packed with ads, do you remember any?

Your phone and laptop are crammed with popups and pre-rolls, notice any?

Your TV is wall-to-wall ads, remember any from last night?

According to media research, we each see over 1,000 ad-messages a day.

So 4% of that would be 40, do you remember 40 ads from yesterday?

The numbers I use were given to me by a media planner and a researcher.

So the real problem is probably many times worse that those numbers by now.

Either way, we all know the facts are true.

People ignore advertising, and your ads have hardly any chance of even being noticed.

Analysing the data, quibbling about the numbers, won’t solve the problem, it will just put off having to deal with it.

It will just kick the can a bit further down the street.

Data won’t replace simple, practical common sense.

Remember, according to data, a linesman’s bald head is the same as a football.