In the past year, AI has become the buzzword that everyone loves/enjoys waving around.

“Including AI” or “excluding AI” – has become the measure by which we evaluate any new solution, technology or method for good or bad.

But beyond the hysterical hype that sweeps us all, it is important to stop for a second and examine whether this is the magic solution we have all been waiting for or will it take a little longer before GPT-driven machines replace us all?!

Full disclosure

I am a technology freak!

Everyone who knows me knows that the thing that fascinates me most is new technology. New solutions that make our lives easier, simpler, smarter.

I love adopting new technologies, but over the years I have learned that beyond the hype, technology also needs to “deliver the goods” and not just be pretty in pictures and demo videos.

So where does this leave us regarding AI and market research?

So here’s where things get a little complicated.

I’ve been using AI for quite some time (relatively, relatively).

Geppetto and I have become very close friends over the past year. Close at the level of regular and continuous daily use in various fields and topics.

But when I get to the fundamental question “Can AI be a game changer in the field of research?” the answer is still no. At least as of today – 6/13/24 (next week everything could turn upside down because of GPT 4oo/5…).

And yet, there are quite a few places where AI can make life significantly easier.

Low and high hanging fruit

Based on the experience we’ve gained today in the combination, I would distinguish between places where AI’s contribution is immediate and clear and those that require more complex adjustments and adjustments.

What is it?

Coming up with new ideas, types of questions to ask, formulas, etc. – the partner I always wanted for coming up with ideas, no matter how flowery they may be. Ai is “the man”.

Text construction and coding (keys) and long text analysis (who said transcriptions?) – If there’s one thing Ai is good at, it’s coding and analyzing texts. Just let it generate insights from text or tell us whether something exists or not in the text.

In all these places, Geppetto and his friends are pretty outstanding. Especially when you don’t have to connect or confront them with outside reality.

What’s (a little) less?

General (historical) knowledge – not bad, but sometimes creates a parallel reality for itself. For example: brings me examples of a groundbreaking campaign developed using Ai in 2011 (!!) – guess who likes to fly on its own?

Quantitative data analysis – getting closer but still a bit problematic. Especially his tendency to take artistic liberties in some cases. We still need to simplify and organize the data for him in a way that is easy to operate, and even then we need to check him and make sure he hasn’t taken creative liberties. Yes, even if we define him to stick to the data.

Arranging and organizing data – not bad, but still sometimes tends to invent things (who said that the Esod HaMaala is located north of Kiryat Shmona??!!).

So where does that leave us?

There is no doubt that Ai is (probably) the future. Good or bad? That’s another story that we won’t touch on right now.

What is certain is that as things develop and progress, Geppetto and his friends will be able to take on more and more tasks in the field of marketing research.

The main problem I see today is mainly related to the added value of AI in the process of creating insights and the paradox inherent in it.

On the one hand, we want and expect that using Ai will help us discover unique things that we couldn’t discover without it. Those hidden “corners” in the data that AI will illuminate and reveal for us.

On the other hand, how do we make sure that it hasn’t invented all those unique insights and those “corners” that it has discovered?

So what does this mean for the continuation of our relationship?

For me at least, right now it means that we will continue to experiment and test, a lot of trial and error until we find the “golden formula” for work and mutual trust. One that won’t require us to run every second and verify the insights it brings. It will take a little longer but I have a feeling that it will arrive, shooting faster than we think.

With me

(This blog was not created by AI. The image actually is)