Skynet Today Last Week in AI News #58
The robots are here. Not to kill us, or to save us for that matter, but to help us… sort of.
With social distancing and remote working becoming the norm, robots have come in handy in various ways, including disinfecting hospitals, aiding doctors and nurses, and performing deliveries. But our current catastrophe has blown up the myth of an “AI takeover” in which human workers will be entirely replaced by such robots. Because machines are still far, far away from matching human intelligence and dexterity, robots are being used to augment humans, rather than replacing them. However, a number of roboticists, who argued their case in an editorial for Science Robotics, are optimistic that with the right amount of funding and development, advancements in medical robotics will allow robots to aid us with everything from social distancing to facilitating interaction during times of pandemic.
We’ve already seen that AI systems are capable of exhibiting different types of bias, resulting from factors such as the availability of data. Biased systems are known to disproportionately impact different groups when employed in contexts such as policing. In a study published in the journal Proceedings of the National Academy of Science, researchers found that even speech recognition systems have trouble understanding certain groups of people. In particular, the study found that speech recognition systems from Amazon, Apple, Google, IBM and Microsoft “make far fewer errors with users who are white than with users who are black.”
If you’ve been following the news about Clearview AI recently, you’re aware that the company has scraped the internet to construct profiles of as many people as possible, using photos of them from various different sites. You might also know that by snapping a picture of you and performing a search, any of Clearview’s clients can access the pictures from that profile to find out more about you. Thomas Smith went ahead and made the effort to get a hold of Clearview’s profile on him, and he found that anyone who had his profile could discover where he went to school, what line of work he’s in, and the region where he lives.
Advances & Business
Forecasting COVID-19 - Forecasting pandemics is harder than many people think. In my book with George Athanasopoulos, we discuss the contributing factors that make forecasts relatively accurate.
Peer-to-peer AI-tracing of COVID-19 - Social distancing works but in its simplest form it is brutal and economically very damaging. We have already seen how tracing and testing can greatly improve it by applying social distancing at the right places, around infected people. Technology can help further.
Musical Robot Learns to Sing, Has Album Dropping on Spotify - Georgia Tech’s marimba-playing robot can now compose and sing its own music.
The GANfather: The man who’s given machines the gift of imagination - By pitting neural networks against one another, Ian Goodfellow has created a powerful AI tool. Now he, and the rest of us, must face the consequences.
An AI Ranked Gaming’s Most Toxic Communities - Many have wondered whether their gaming community is one of the most toxic, and a recent study now gives insight into which games trigger some of the most offensive comments.
New Rules Could Finally Clear the Way for Self-Driving Cars - For the first time, the National Highway Traffic Safety Administration is weighing in on autonomous vehicles with no driver behind the wheel—or no wheel at all.
Can artificial intelligence fight elderly loneliness? - Voice technologies using AI are being used to help combat loneliness in countries including Sweden and the UK.
A.I. Versus the Coronavirus - Advanced computers have defeated chess masters and learned how to pick through mountains of data to recognize faces and voices. Now, a billionaire developer of software and artificial intelligence is teaming up with top universities and companies to see if A.I. can help curb the current and future pandemics.
Can your company’s A.I. handle the coronavirus pandemic? - The coronavirus pandemic is providing a real-world test of how robust many companies’ new-fangled A.I. systems really are.
Analysis & Policy
Research summary: The Deepfake Detection Challenge: Insights and Recommendations for AI and Media Integrity - Top-level summary: Accurate and trustworthy information plays a key role in the development of a well-functioning society.
Expert Opinions & Discussion within the field
A debate between AI experts shows a battle over the technology’s future - On March 26 at MIT Technology Review’s annual EmTech Digital event, two prominent figures in AI took to the virtual stage to debate how the field might overcome issues of fragility, bias, and inability to generalize.