Anthony Zador is Chair in Neuroscience and Professor of Biology at the Cold Spring Harbor Laboratory, wherein 1953, Watson made the first public presentation of DNA’s double-helix structure. Mr. Zador went to this year’s edition of the Web Summit to talk about “Using Artificial Intelligence for good”. He believes AI is “going to be eliminating tens of millions of jobs and it’s unclear what they’re going to be replaced with”.
“We’re going to be eliminating tens of millions of jobs and it’s unclear what they’re going to be replaced with”
Anthony Zador’s goal is to understand the neural circuits underlying cortical processing and he is currently developing a new high-throughput approach to determining brain wiring using next-generation DNA sequencing.
We met with him at the Web Summit to know the latest news on AI and Neuroscience research.
Neuroscientist Anthony Zador interview.
Miguel Salvado: How are you finding the Web Summit so far?
Anthony Zador: It’s been really a lot of fun. An amazing collection of people, it’s been really interesting.
How are you finding the Lisbon?
I love Lisbon, this isn’t my first time here. I’m a neuroscientist here at one of the leading Neuroscience Institute in the world, the Lisbon’s Champalimaud institute. I’ve been to Lisbon a few times to visit there.
Tell us a bit about what you do in Neuroscience.
My own research is focused on understanding brain circuits. The brain has a hundred billion neurons, a hundred trillion connections and our goal is to understand how they are all connected. The way they are connected makes you, you, and makes me, me.
We’re developing a technology that will actually allow us to unravel the wiring diagram of a brain, quickly and cheaply with the goal of understanding how your brain works and how it stops working when you’re depressed or schizophrenic. Eventually, it might help us build better machines.
You’re mapping, or framing, a system that you can implement technologically?
That’s right. It turns out that the current generation of Artificial Intelligence is based on taking insight from how real brains are wired up. The reason that neuro networks work today and didn’t work 25-years ago is that we built-in more about how we know brains work, and it is really those insights that have allowed for the current generation of neural networks.
In a way, is AI going back to basics?
That’s right. They used to think that you can just ignore the brain, you could just watch behavior, we could just seat down at a computer and program machines to do what people do, but that has failed. That really didn’t get us any further. That worked for chess, but what made computers able to recognize images and beat the world champion in Go was neural networks. Those are networks that came from understanding how real brains are wired up, and we’re just at the beginning of that. We’re just taking the tiniest bit of what we know actual brains do and put them into networks. We’re at the forefront of taking extra information of how real networks are wired up and putting those into machines.
How do you see AI going over the next years in relation to what you’re doing right now?
The good news for me is a lot of fun. There are tremendous opportunities for understanding how networks work and implementing them. The more troubling thing is that the consequences for society are going to be, in a lot of ways, devastating. Particularly in the United States where we don’t really have social services the way you do in Europe. We’re going to be eliminating tens of millions of jobs and it’s unclear what they’re going to be replaced with, if they’re going to be replaced by anything. That’s real challenges, but it’s a question of society, it’s not the fault of technology.