
“If no one asks me, I know what it is. If I wish to explain it to him who asks, I do not know”.
These were Saint Augustine’s words about time but they could just as easily be mine about artificial intelligence.
If there’s one notion that causes widespread confusion it’s artificial intelligence.
Twenty years ago, artificial intelligence meant sat nav systems and video games (remember FIFA 98?) whereas today it means things like self-driving cars, Facebook’s newsfeed ranking and the Google research engine. What on earth has happened in just two decades?
To find out we should ask Yann LeCun.
Why him? Because he’s in charge of Facebook’s AI work and is considered to be one of the inventers of deep learning – the concept that relaunched research into artificial intelligence.
What is so revolutionary about deep learning? Well, until now, artificial intelligence was restricted to machine learning, i.e. learning based on statistical analysis.
Deep learning involves feeding a computer system a lot more “training” data, which it can use for conceptual learning to create models and correlations.
In short, it corresponds to a more open, less task-based algorithm, which directs itself in order to interpret data. So how are we different from machines now?
We still have two things that they don’t – instinct and common sense!
Share :
Something to say ?