Anicet Mbida 06:55, January 26, 2022

Every day, Anicet Mbida makes us discover an innovation that could well change the way we consume.

This Wednesday, he is interested for once in a technological failure.

IBM has just announced the resale of its super computer doped with artificial intelligence, Watson, to an investment fund when it was supposed to revolutionize medicine in particular.

For once, this morning you are telling us about a technological failure.

IBM has just announced the resale of Watson to an investment fund.

Watson is his super computer doped with artificial intelligence which was to revolutionize medicine in particular.

What happened ?

It did not work or very badly.

In any case, not as good as had been announced a few years ago.

Watson was a tech showcase: the heir to Deep Blue, the first computer to beat Gary Kasparov at chess.

It is also the first computer capable of beating champions in Jeopardy.

The first to debate with humans.

We therefore imagined him capable of diagnosing rare diseases, of helping to find new drugs…

Unfortunately, most of those who tried it came back (I'm thinking of the Pompidou Hospital next to our studios).

Why ?

Because his results weren't always relevant.

He also made several obvious errors and incorrect, sometimes even dangerous, treatment recommendations.

In short, it was the cold shower.

What does that mean for artificial intelligence that has been presented as a revolution for a few years?

This reminds us, once again, that there is no magic technology.

They are only tools and it is their use that can then change the game.

Look at the self-driving car.

It has been announced as a revolution for 15 years.

But in fact, the only ones that work are minibuses that run at 40 km/h on marked routes.

The few experiences of autonomous taxis have choppy driving, the car stops at the slightest plastic bag on the road.

They are unable to drive in bad weather.

Result: many car manufacturers no longer believe in it at all.

Are there still sectors where AI is changing the game?


Image recognition, finding defects on a production line.

This is where the most progress has been made.

For the rest, whether chatbots or natural language understanding, we are still far from the mark.

We measure it every day with our personal assistants.

So beware of all the magic technologies: Blockchain, Virtual Reality, Metaverse… The potential is always huge.

But let's wait for the real uses before talking about revolution.