Virtual and augmented reality. 3D printing. Natural language processing. Deep learning. The smart home. Driverless vehicles. Biometric technology. Genetically modified organisms. Brain-computer interfaces.
These, in descending order, are the top 10 most-invested-in emerging technologies in the United States, as ranked by number of deals. If you want to get a sense of which technologies will be shaping our future in the years to come, this probably isn’t a bad starting point.
The figures come from a massive new artificial intelligence forecasting engine built by the French intelligence firm, L’Atelier. “We make sense of tomorrow, today,” claims the website of the small company, which has been doing its smart technological guesswork (with humans instead of A.I.) since 1978.
“I call it the technology intelligence engine,” said Giorgio Tarraf, the bearded, yet boyish, 33-year-old who built the new model. “I think it’s a terrible name, but for now it’ll have to do.”
It certainly can’t be worse than much of what we have today. It’s no secret that most predictions are terrible. A famous 20-year study of experts, comprising 82,361 probability estimates about the future, were almost all wrong. As David Epstein’s 2019 article for The Atlantic, “The Peculiar Blindness of Experts,” notes of the study: “When experts declared that future events were impossible or nearly impossible, 15% of them occurred nonetheless. When they declared events to be a sure thing, more than one-quarter of them failed to transpire.”
Geoff Hinton, one of the Nobel Prize-winning pioneers of artificial neural networks, once described the future to me as being akin to peering through fog. “When you’re in fog, you can see short distances quite clearly,” he said. “When you look a bit further, it’s fuzzier. But then if you want to see twice as far as that, you can’t see anything at all. That’s because fog is exponential. Each unit of distance you look through fog, it will lose a certain fraction of the light.”
Technology is no different. We might have a reasonable idea of what the next six months will hold for tech, but this gets sketchier when we predict the year 2022 as a whole. Jump forward five, 10, 15, 25 years and it’s all but impossible. Venture capitalists have long hunted unicorns, meaning big billion-dollar companies, not just because they’re immensely profitable, but because they offset all the other mistakes they make. You only need one Google or Facebook, or to have guessed right about the potential of smartphones back in, say, 2000, for all the incorrect predictions to fade into the distance.
Tarraf was fed up with incorrect predictions. He wanted a more data-driven approach to forecasting that could help investors, governments, pundits, and anyone else to get a more accurate picture of the shape of tech-yet-to-come. Not only could this potentially help make money for his firm, but it could also, he suggested, illuminate some of the blind spots people have which may lead to bias.
Tarraf’s technology intelligence engine uses natural language processing (NLP) to sift through hundreds of millions of documents — ranging from academic papers and research grants to startup funding details, social media posts, and news stories — in dozens of different languages. The futurist and science fiction writer William Gibson famously opined that the future is already here, it’s just not evenly distributed. In other words, tomorrow’s technology has already been invented, but right now it’s hidden away in research labs, patent applications, and myriad other silos around the world. The technology intelligence engine seeks to unearth and aggregate them.
“We have 100 million publications from around the world that have come from dozens of journals,” Tarraf told Digital Trends. “We’ve got over a trillion dollars in grant funding. We have 14 million patents. In the next version, you’re going to have over 100 million, with a big focus on Chinese patents. And we have early stage investment data of tech startups from 2015 to today.”
The idea of having all these various metrics for assessing the future is that each gives a different perspective — and a differing timeline. Startup funding, for example, is typically focused on the next two or three years. That’s because it’s the speed at which investors want to see a positive cash flow and, possibly, an exit. Not every startup will be a success, of course, but broad trends in funding can show where the areas of interest are.
Research grants, meanwhile, are closer to the five- to 10-year range view of the future. Academic papers, especially theoretical ones, offer the longest view of them all, stretching off into the technological horizon. As Tarraf points out, there was a rush of journal articles about quantum computing published in the 1990s, but the field is only starting to take off (or, at least, to rumble at the launch pad) today.
There are also those technologies that receive an outsize focus in the news media, but are probably a lot more smaller than their large headlines suggest. “Dream technology captures a lot of attention,” he said. “Electronic contact lenses capture a lot of attention. But we don’t see them capturing a lot of academics’ attention. They’re just very cool [pieces of technology]. We all want to imagine a world where we can control our dreams while wearing these funny contact lenses.”
Another aspect of the technology intelligence engine is to look more broadly at technology from around the world. “Everyone’s obsessed with the U.S. or China,” said Tarraf. “We find that there’s innovation happening all around the world.” Consider, for instance, India. “No one’s talking about India’s growth in emerging technology,” Tarraf said. “It’s massive. It’s incredible. It has to be celebrated. We should be focusing on the amazing work that’s being done by academics there. But they don’t get the funding, they don’t get the attention. This is what the engine can do. It can take you away from the global conversation we are having, and into the global conversation that, maybe, we should be having.”
Based on the technology intelligence engine’s findings, there is no shortage of fascinating insights. For example, the U.S. dominates in most emerging technologies, although Canada leads in the number of deals related to carbon capture. There have been 2,000 deals on virtual, augmented, and mixed reality tech in the past five years — with almost $2 billion invested. During that same time frame, $1.1 billion has been invested in drone technology, while A.I. tech has captured $3 billion of investment dollars.
And what about the hot technologies of tomorrow? Cryonics, aka technology that enables super cold storage, is big. Long considered a niche technology, there has been a big bump in interest in the past two years. Cryonics-related tech was needed to transport and store some of the mRNA COVID-19 vaccines, underlining its real-world usefulness.
Post-quantum encryption is important, too, referring to new ways of encrypting data for the quantum world. In 2019, nearly $50 million was allocated to the topic in global research funding — more than double the previous year’s total.
And who can forget brain-computer interfaces, which benefited from the second largest growth in investment total in 2020, following satellite constellations. Once again, the U.S. dominates this field in terms of patents, capturing almost half of the nearly 200,000 BCI patents issued since 2015. China comes second, with a comparatively minor 35,000 patents to its name.
Affective computing, meaning computers that understand human emotions, is in demand worldwide — although the U.S. trails China and India as researchers in this field. In 2020, China has filed 589 patents on affective computing, compared to 37 filed in the United States. However, the U.S. is leading in the number of investments related to this field.
As comfortable as Gio Tarraf is talking about the future, he’s not yet able to share the future of the technology intelligence engine. The version I saw was clearly marked “demo,” and Tarraf notes that it is very much still a work-in-progress. As to exactly how this will be made available as a public tool (assuming that it will) has yet to be announced. One thing’s for certain, though: He’s predicting it will be big.
“I see this as a way to expand your vision of the world, and to reduce your bias and give a fairer view of technologies that are often overlooked, but that could have a significant impact on our lives,” he noted.
- Fighting football injuries with 3D-printed, hyper-personalized pads
- Read the eerily beautiful ‘synthetic scripture’ of an A.I. that thinks it’s God
- The future of A.I.: 4 big things to watch for in the next few years
- Emotion-sensing A.I. is here, and it could be in your next job interview
- Language supermodel: How GPT-3 is quietly ushering in the A.I. revolution