Andreas Gal | Machine Learning and the Internet of Things | Singularity University

Andreas Gal | Machine Learning and the Internet of Things | Singularity University


(music) – I couldn’t be happier to have
a very special guest today. Andreas Gal, former CTO at Mozilla. Today, the founder and CEO
of Silk Labs, where you apply artificial intelligence
to this vast area of IoT. Like, I’ve got all these
connected devices and, mostly, they don’t talk to each other, or talk to each other in very basic ways. Where do you see this going? Because there doesn’t seem
to be a common language for these devices at the moment, right? – I think that’s a really
good point and I think there’s a lot in the history of
the web that we can use to learn here. If you think back, the
web wasn’t really solved and evolved into what it
is today by sitting down and doing a grand committee
of plans and coming up with this 10-year plan of technology and then implementing that. It was really just like people
were adding random pieces to the web, and some piece
of it worked, others didn’t. Multiple companies evolved the
web, and then at some point after the fact, they would get
together and try to kind of harmonize these standards a little bit. And that’s a very powerful
approach, and I think that’s really what will happen
in the IoT space as well, is that we have to focus first
on really good experiences and great products that
will solve real problems. Once you have two devices that
solve real problems for you, then we can find a way
how they work together and do this even better. We start the other way around by gathering a large committee of people and
come up with some standards. There’s five or six of them already, and nobody’s really using them because it’s very hard to take the first step. And I think this is
one of the big mistakes that a lot IoT device
manufacturers are making, is that they kinda slap
connectivity onto a device and I have a connected toaster, and it’s just as bad or
worse than an actual toaster. So it’s really around
bringing intelligence to these devices, and that’s
what we are focused on. So at Silk Labs, really it’s specializing, we are bringing the latest
advances in the iTechnology directly into these devices. There’s a lot of privacy,
but also usability benefits if the intelligence comes
to the edge, versus trying to put all these intelligence
functions into the cloud. – Do you think, on a macro
trend, do you think that we are seeing a shift away
from people going into software engineering and
more into the data science? So is the job of the
future the data scientist versus the software engineer? – Absolutely. I firmly believe long term,
software engineering is really going to be de-emphasized
as kind of the crown jewel of making, especially
software-based products. The last twenty years, software
engineers were basically fairly high up in the
decision-making process because they had so much influence how software experiences turn out. This will change. If we measure about data
scientists and how to understand data and how to
form these new networks, there’s really a big shift happening. Now it’s important, this
is long-term trend, right? I think it’s comparable
to like self-driving cars. If today you can still get a
driver’s license, it will also be useful to you for many
years, but in the longterm trend at some point, you probably
want to think about a different long-term
career than a driver. I’m not discouraging people
that go to school right now to learn how to program,
but in a 20-year time frame we will see a lot more emphasis
being put on data sciences, and understanding how to use
data to train your network to solve problems. And so you use software
engineering, manual skills, to design algorithms to
solve similar problems. – I’m curious, you’ve been
in technology forever, you’ve got a really deep
background in technology. What are you most excited
about, about the future outside of your field? – I think one of the things
that I find very exciting right now about just the way
technology itself is evolving is that, the most powerful
applications of technology have now left the direct
technology sector. You’re seeing, especially
software technology, but too actually as a hardware technology, we need transcend software engineering, and so they can really end it. The coolest companies have
solved the coolest problems, and not only narrowly speaking, like just technology companies, This goes everything from a
Lyft and an Uber that transport. Yes, there’s technology involved,
but at the end of the day, first of all it’s a magical experience. It’s the closest we have gotten so far to the start of Transporter. You step outside a
building, get out your phone and push a button, and half an hour later you show up in a different place. It’s magic, right? And technology is involved,
but really in a secondary role. It transforms the way people
interact with each other and makes new experiences possible. But also if you think about agriculture. I started helping a company
that does indoor agriculture using big data to kinda see how we can improve the efficiency of agriculture. So here’s technology
making agriculture better. So I think this is very exciting
for me as a technologist is that I can see what we
have done the last 20,30 years in technology, what we have
learned, really impact broadly the field around the
narrow technology sector. – That’s fantastic. (bubbly electronic music)

Leave a Reply

Your email address will not be published. Required fields are marked *