Onboard AI: Inside Independent Autonomous Devices - Forrester’s New Tech Forum
CEO and co-founder of Neurala Heather Ames Versace sat down with Senior Analyst Kjell Carlsson at the Forrester New Tech Forum: Automation Everywhere.
How far have we come in the artificial intelligence space?
A long way. 12 years ago, our images of robots were connected to wires and supercomputers. With recent innovation and hardware, we’re able to look at a different reality than what many people could have imagined then. We’ve made a lot of progress in making things smarter--back then, we relied on servers, but we’re now innovating on layers and layers of computing power, all working to understand what’s going on.
What’s holding us back now? Tech or perception?
Tech. To serve enterprises, AI needs to be able to learn and adapt on the fly. When you deploy a system that has to train on server, it’s a huge deterrence. Updating in real time, in a completely different environment, and retraining your network without connecting to the cloud—that’s the future of AI. The other thing we need is to improve users’ understanding of AI. A lot of people know that they need AI for 2018, but do not know what it’s going to do for them.
How does data fit into this?
There’s a need for an incredible amount of data. Right now companies are looking at 100 thousand - 2 million images, and have to annotate every one of them. The innovation we’re seeing is structured around requiring less data to generalize and learn. Is this feasible? Preliminary results are telling us yes.
Should we be worried about autonomous robots taking over?
No—these machines aren’t going to rise up anytime soon.
Why do you think North America is behind rest of world?
That’s a very real concern. When we identify the levels of training and education we look for in candidates, most of those people are international.That said, we are seeing so much innovation right here at home. Misconceptions are another I think companies need to think about the information they have and how to go about collecting the right kind of information—and then maybe we’ll see some improvements to the AI ecosystem here in North America. My advice to companies, though, is don’t wait for tomorrow. Do it today.
With this space moving so fast, what should companies be tackling now?
Don’t wait. You need to invest now. This work is not easy to do, so find the companies that have gone live and nailed it, like Neurala.