Jason Sosa talks like someone who knows what tomorrow brings. He is from San Francisco, where he managed a venture capital fund investing in new technology; and before that New York, where he sold a facial recognition start-up that proved well ahead of its time.
He has featured in Esquire and Authority Magazine. He spoke at the most recent Singularity University Global Summit, and has been tagged as a leader of diversity in the blockchain realm. At Melbourne’s 2019 Integrate Expo, Sosa is focused on the audiovisual industry’s most undefinable yet oft-mentioned topic: artificial intelligence.
“There are a lot of people talking about AI and they really don’t know what it is or the extent that it is used today,” Sosa says.
“When you go on Netflix and you’re looking for recommendations, every time you go on social media platform and use their feed, when you are using maps; these are all examples of using artificial intelligence.”
The artificial grass is greener
Sosa notes that our inability as a society to articulately define artificial intelligence is at least in part due to a very simple fact that Amazon’s Jeff Bezos, has touched upon: “In the AI era, we are the very first at bat.” By this, Bezos is articulating that our generation is first ‘on the field’.
In many ways, this makes us the guinea pigs of artificial intelligence’s imminent impact upon society. Bill Gates has compared the advent of this age as significant to humanity as the introduction of electricity. Like all technological advancements, this can prompt anxieties for users.
“If you think of AI, you might think of it in conduit with ideas of science fiction as a machine that can interpret new things, a bit like [Stanley Kubrick’s 2001]: ‘Dave, open the pod bay doors!’”
But while at least some of science fiction may prove prophetic, it is not the kind that should be feared.
A trigger-happy tomorrow?
The primary advancement discussed by Sosa is one that many will be familiar with from Stephen Spielberg’s 2002 movie Minority Report. When Tom Cruise’s character visits a shopping mall, the digital signage rolls customised content after the sensors discreetly scan his eyes to identify his demographics and tailor its ads and commentary accordingly.
In the real-world, this is now called “triggered content”. “We’ve reached the technological ability and infrastructure to be able to make this a possibility,” Sosa says. “There are some really interesting things that can be done here. If a man walks in front of a display, it’ll be a Gilette ad; if a female walks in front of a building, it’ll be a perfume ad.”
Still, even with this groundbreaking leap, inefficiencies still exist. It is important for those on the verge of implementing artificial intelligence in their products that they understand it is no shortcut to soaring sales.
“AI is not a one and done solution,” Sosa says. “It is a seed and grow [prospect]. It is something you have to plan for over the long term. The biggest thing is managing expectations with shareholders. It is a never-ending process.”
AI’s hand in the cookie jar
Such is the exponential growth of artificial intelligence that those who are willing to harness it harbour ambitions grander than merely establishing if you are male or female. The goal is to take that personalisation a few steps further.
“At the moment if you walk out of sight of the camera and then come back, it records you as another face. A new customer,” says Sosa. “But what if there was a solution like the cookie on the internet?”
“On the internet when we visit a website, it creates a web cookie. It has your browsing information so it can at least identify you when you come back.
“In the offline space, where an equivalent could be created, it wouldn’t actually identify your name or personal information, but it could know that you were the person that visited the store and that you came back a month later or that you came back a week later,” Sosa says. “That’s where you could measure repeat customers and see if you are generating the loyalty you want from your retail installations. It’s the next evolution of where this type of technology is going.”
While it is hard to dispel Big Brother-like notions to this idea, Sosa says the technology will recognise you not as a surveyed person, but as a simple hash. “This technology is made specifically for retail marketing analysis,” Sosa says. “It is not used for surveillance or security. It is using edge computing to gather information and optimise it within a millisecond.”
According to Sosa, this “face cookie” technology is already in production. So rest assured, it is something you will see in department stores someday soon. You can be even more certain that it will see you.