Enthusiasm abounded at the World AI Cannes Festival

By now, we’re well-accustomed to waves of tech-based fervor. You don’t really have to touch the industry to have become a cynic. Perhaps you remember “peak blockchain” in 2019, when the technology was integrated into a toothbrush for reasons that no-one will ever begin to understand?

And of course, we’ve endured fanfare and furor over the metaverse, XR, Big Data, crypto, Web3, NFTs, quantum computing and, most notably, we’ve seen AI grow legitimate roots as the defining technology of its era (while its name continues to be frequently used and abused by sneaky bandwagon jumpers…).  

On those firm foundations, generative AI is the term at the center current hype-cycle. In Cannes last week a reported 16,000 attendees swarmed on the famed Palais des Festivals et Des Congrès to learn about it, talk about it, and – for many – showcase the tentative steps they’ve taken towards real-world application.

As with previous seasons, there was an urgency (perhaps even a whiff of desperation…?) in the air as companies from industries as diverse as hospitality, finance, entertainment, marketing, and pharma joined this latest gold rush as hopefuls. And, unsurprisingly, both snake oil and substance could be found.

I was lucky enough to host the festival’s Applications Stage, and here are three broad-brush observations I made:

1) Generative AI feels different

In contrast with some of the aforementioned waves of near-hysteria, it’s evident that generative AI is more than just the hype. For one, it’s neither premature nor entirely novel. That’s not to say that it isn’t immature, but it’s worth noting that the most impressive applications were timely augmentations of pre-existing AI models. In many cases generative AI adds the magic fairy dust that really makes those systems sing.

2) But it’s still a wild child (and humans are needed more than ever)

With the significant risks attached to hallucinations, drift, and other potentially catastrophic bloopers, most presenters were keen to stress the importance of keeping a “human-in-the-loop.” Generative AI is still an intelligent and well-intentioned Bamm-Bamm-type character that currently requires the support of a vigilant adult to ensure it colors within the lines.

3) More thought is needed on the “so what?”

Though generative AI will undoubtedly allow us to do many new things, most applications this year showcased enhanced speed and efficiency for existing processes. But we sometimes labor under the mistaken idea that speed and efficiency are ends in themselves. They’re not. True, compelling benefits center on what speed and efficiency allow us to do.

For example, does it gift individuals the “brain space” to be more creative (if so, what’s the output)? Does it allow you to build more widgets (and how will more widgets improve the human condition)? What about new initiatives, projects, or products? Is it simply that this technology reduces stress-inducing pressure on human workers – how might that be measured and celebrated? (If it enables the company to do more with less, how might ultimately benefit customers, clients, users, or even investors?).

While there is a natural tendency to focus on the mechanics of how very clever the AI is, at the end of the day it’s a tool. A tool’s job is – and always will be – to foster human happiness and flourishing.  Let’s have more illustrative examples that draw a clear line from the system to this objective.

———-

Last year at WAICF we were all still reeling from the loud and audacious arrival of ChatGPT. This year, we saw some suitably cautious but robust attempts to harness and direct its power. Next year, we shall all expect full evaluations of the mostly-baked pudding, complete with metrics and (perhaps!) just a hint of how it’s making the world a better place.