Matthew: There's a lesson in there about your own
conviction and perseverance, and that helping to transform the
perspective of someone who was a mentor or someone who professionally
was ahead of you. And I think that's powerful. And that's a
powerful lesson for many, many people.
Rana: Yeah. You run into a lot of naysayers and a
lot of skeptics, and you have to either decide that you are going to
forge ahead anyway, with or without their buy-in; or you want to get
their buy-in. If you want to get their buy-in, it's a process, and they
have to build a lot of trust. You have to work on it. And I think that has definitely been the case in my careers, especially
with the company.
Matthew: Hey, Shaping the FutureTM listeners, do you know where you can get the best digital curriculum all in one place? If you're a school or district leader, team up with HMH today to help all students and teachers reach their full potential with our new HMH Anywhere solution. HMH Anywhere offers educators seamless access to core curriculum for language arts, math, science, and social studies, paired with AI-driven, personalized practice, and online professional development for teachers. And all supported by a unified assessment system that precisely tracks students’ growth. HMH Anywhere also integrates with tools like Zoom, Google Meet, and Microsoft Teams for easy remote teaching and learning. Check it out at hmhco.com/anywhere. Now, back to the episode.
So I mentioned at the outset, you're a
pioneer new branch of AI. We have mentioned casually this thing, Emotion
AI, but for our listeners, could you just do the sort of basic
explanation? What is Emotion AI, and how did it come to be, and how is
Rana: Sure. So I'll start with human intelligence.
In human intelligence, you have your IQ, your cognitive intelligence,
but you also have your EQ, your emotional intelligence. And we know from
over 60 years of research that people who have higher EQs and certainly
all educators would resonate with that. They tend to be more productive
and personable and persuasive in their work and in their lives. So, I
believe this is true for technology as well. Especially technology that
is so deeply ingrained in our everyday lives like AI is. So my
hypothesis is that all these devices that we interact with on a daily
basis, it's not enough for these devices to have IQ. Like Siri, has a
little bit of IQ because you can talk to Siri and talk to Alexa. But
they have no EQ. They have no emotional intelligence.
Emotion AI, or artificial emotional intelligence, is this idea
that we build technology that can quantify and capture these enriched
nonverbal signals like facial expressions. Or sometimes, we do vocal
intonations of how fast are you speaking. How much energy is in your
voice. All of these signals, we're now able to capture them using
technology. And it unlocks a lot of use cases.
Matthew: And the face is important in the expression of emotion. Why?
Rana: Ooh, the face is one of the most powerful
canvases for expressing one's...not just emotions, but your cognitive
states or social state. There's about 45 facial muscles that drive what
expressions we make, and in the seventies, this guy called Paul Ekman—there's a show that was modeled after his work. It's called Lie to Me.
And basically, he mapped every single facial muscle to a code. So, when
you smile, it's code 12. When you do a brow furrow—angry—it's like
action unit four. And he built a system so that people could become face
readers, like certified face readers. But it's still laborious. It
takes a hundred hours of training, and it's just so time-consuming.
So instead, we use machine learning and computer vision to train
machines to do that automatically. So when it sees your face, Matthew,
it can very quickly say, "Ooh, I can see the 12 plus four plus three
plus 17." And just map out your expressions. And then if there's a
different level of coding that takes it to an emotional state or a
cognitive state. Do you look tired? Do you look confused? Do you look
excited, interested, happy, sad, et cetera.
Matthew: And so, you really took this idea and your
research here. You went with Rosalyn to MIT Media Lab where you
started there as a researcher, but then you guys got together and
co-founded this company Affectiva that you're now the CEO of. And spun
it out of the university. So, you're not only a pioneer in this whole
new branch of AI, but the fact that you are a founder CEO of a venture
capital finance tech startup is also a pioneering part of your journey.
And I'd love for you to talk a little bit more about that aspect of it.
Taking this thing from being a research project to becoming a company.
Rana: Yeah. So when I got to MIT Media Lab in 2006,
the Media Lab is very unique as an academic institution in that it's
very interdisciplinary. It's where the misfits are, but also, we were
very tied to industry. So twice a year, we would sponsor all of these
Fortune 500 companies and invite them to the lab. And it was actually
called a "Demo or Die." You couldn't just show up with a PowerPoint, or
you just talk. You have to show a prototype of what you were building.
So, leading up to these weeks, twice a year, these weeks we would just
work overnight. We would spend the nights in the lab building these
prototypes. So, for a few years in a row, all these companies wanted to
buy the technology. So, Procter & Gamble wanted to test their new
products using this. Bank of America wanted to track customer
We had Pepsi that wanted to test their ads with this, right? We just
had this list. Toyota wanted to track driver drowsiness. And I literally
kept a log of all these different use cases. And I had no way of giving
them the technology because it's an academic colors, no mechanism to do
that. And when the list got to about 20 different companies, Roz and I
went to the Media Lab director at the time, Frank Moss, and we said,
"Hey, Frank, we need budget. We need more researchers on this because
we're ignoring our sponsors." And he said, "This is not research
anymore. This is a commercialization opportunity." My knee-jerk
reaction, my initial reaction was like, "Wait, Frank. I'm about to apply
to faculty at MIT. Don't mess with my plan." But then I thought about
it. And it was a really unique opportunity, as you said, to take
something I'm deeply passionate about and now bring it to the world at
scale. Which often, academia isn't really set up to do well.
Yeah. So we started at Affectiva in 2009. We are venture-backed.
We've raised over $15 million of venture in strategic funding. And it
has been a roller coaster, an emotional roller coaster,