Home PC News Gary Marcus: COVID-19 should be a wake-up call for AI

Gary Marcus: COVID-19 should be a wake-up call for AI

The world pandemic has been cited as a “wake-up call” for a lot of points — the environment, economic and social rights, and regular global inequalities. However, scientist, author, and entrepreneur Gary Marcus thinks that the COVID-19 catastrophe additionally must be thought-about a wake-up identify for AI too.

Speaking on the digital Intelligent Health AI conference yesterday, Marcus lamented a few years of missed alternate options to assemble a further sturdy artificial intelligence, arguing that an extreme quantity of AI helpful useful resource has been put in direction of utilized sciences that don’t truly help the world in any important method.

“We would like AI that could read and synthesize the vast, quickly growing medical literature, for example, about COVID-19,” he acknowledged. “We want our AI to be able to reason causally, we want it to be able to weed out misinformation. We want to be able to guide robots to keep humans out of dangerous situations, care for the elderly, deliver packages to the door. With AI having been around [for] 60 years, I don’t think it’s unreasonable to wish that we might have had some of these things by now. But the AI that we actually have, like playing games, transcribing syllables, and vacuuming floors, it’s really pretty far away from the things that we’ve been promised.”

One of the underlying factors, in accordance with Marcus, is that we’re putting an extreme quantity of focus on deep finding out.

“To understand how to bring AI to the next level, we first need to understand where we are, and where we are right now is in the era of deep learning, where deep learning is the best technique, and the dominant technique, and maybe one that’s getting too much attention,” Marcus acknowledged.

Marcus has a PhD in cognitive science from MIT, and has been a professor of psychology and neural science at New York University for the earlier 20 years. Throughout that interval, he has moreover written a variety of books, and in 2015 he cofounded Geometric Intelligence, a stealth AI startup which was swiftly snapped up by Uber to perform the muse of its new AI Labs. Marcus stepped down as head of Uber’s new unit after solely a few months, and he later went on to found Robust.ai to assemble an “industrial-grade cognitive engine” for robots.

The downside

Deep finding out is a division of machine finding out based mostly totally on artificial neural networks that try to mimic how the human thoughts works. Deep finding out isn’t in want of critics, and the inherent weaknesses are properly understood. Large swathes of information (footage, audio, textual content material, shopper actions, and so forth) observe the deep finding out system to acknowledge patterns, which might be utilized to help Netflix counsel video content material materials or autonomous autos decide pedestrians and freeway indicators. But slight modifications to the data enter, modifications {that a} human would possibly (or won’t) be able to spot, can confuse even most likely essentially the most superior deep finding out strategies. An occasion that Marcus makes use of is you’ll be able to observe a deep finding out system to find out elephants — nevertheless current it a silhouette of an elephant, one {that a} human would merely acknowledge, and the AI would most likely fail.

“The reality is that deep learning works best in a regime of big data, but it’s worse in unusual cases… so if you have a lot of routine data then you’re fine,” Marcus acknowledged. “But if you have something unusual and important, which is everything about COVID since there is no historical data, then deep learning is just not a very good tool.”

Marcus moreover reiterated elements from his Rebooting AI e book which was printed closing 12 months, noting that the AI world should refocus its efforts on a further hybrid “knowledge-driven” methodology. One that comes with deep finding out, which is good at some kinds of finding out nevertheless “is terrible for abstraction,” and classical AI, strategies in a position to reasoning and encoding info.

Whatever probably the greatest path forward is, Marcus’s principal takeaway as far as COVID-19 is apprehensive, is that the pandemic should serve to encourage the AI world to rethink the problems that they’re in the long run making an attempt to resolve.

“COVID-19 is a wake up call, it’s motivation for us to stop building AI for ad tech, news feeds, and things like that, and make AI that can really make a difference,” he acknowledged. “With better AI, we might have computers that can read, digest, filter, and synthesize all the vast growing literature [around COVID-19]. Robots could take on a lot of the risks that human health care workers are facing. To get to that level of AI, that can operate in trustworthy ways even in a novel environment, we’re going to need to work towards building systems with deep understanding not just deep learning.”

Most Popular

Baidu’s smart home group seeks to raise capital at a $2.9 billion valuation

Baidu today announced that it will seek to raise financing for its Smart Living Business (SLG), the internal group that maintains Baidu’s DuerOS voice...

D-Wave’s 5,000-qubit quantum computing platform handles 1 million variables

D-Wave today launched its next-generation quantum computing platform available via its Leap quantum cloud service. The company calls Advantage “the first quantum computer built...

Google’s Cloud TPUs now better support PyTorch

In 2018, Google introduced accelerated linear algebra (XLA), an optimizing compiler that speeds up machine learning models’ operations by combining what used to be...

Square adopts QR codes to bring self-serve ordering to restaurants

Square has introduced a new self-serve ordering feature for restaurants that allows dine-in customers to order and pay for their food through their phones,...

Recent Comments