AI Trends in 2020, AI algorithm to sniff out disease, and more!

Last Week in AI #106

Mar 8, 2021

AI System Can Sniff Out Disease as Well as Dogs Do — Image Credits: Leon Neal / Getty Images via Scientific American

Mini Briefs

The 2021 AI Index: Major Growth Despite the Pandemic

Stanford Institute for Human-Centered Artificial Intelligence published the 2021 AI Index report, outlining recent AI trends and developments in 2020. Among their findings are significant increases in private AI R&D, especially in biotech and drug design, the share of AI PhDs going into industry continues to increase over those staying in academia, and advances in generative models are making synthetic media (text, speech, images) more realistic than ever. Covid-19 does not seem to have had a negative effect on AI hiring and investments, and attendance in AI research conferences might have actually gone up due to the accessibility of virtual presentations. The full report, although long, has many insights and interesting data visualizations - would recommend a skim!

AI System Can Sniff Out Disease as Well as Dogs Do

Dogs have long been trained to sniff out diseases from smells. While they can be highly accurate, dogs can get bored and tired, and they can't produce diagnoses consistently. Motivated by this, a recent study developed a system that developed an AI system to diagnose prostate cancer from the smell of urine samples. This is done by training a neural network to map observations from Gas Chromatography–Mass Spectrometry to known positive and negative samples. The system was able to identify positive cases 71% of the time, and negative cases around 70% of the time. While a promising first step, the researchers note the accuracies are not high enough for the system to be used in diagnosis. In the future, they hope to improve the system's performance and integrate it with hand-held smell detection devices.

Podcast

Check out our weekly podcast covering these stories!
Website | RSS | iTunes | Spotify | YouTube

News

Advances & Business

AI: Facebook's new algorithm was trained on one billion Instagram pics - "Dubbed SEER (SElf-SupERvised), the model was fed one billion publicly available Instagram images, which had not previously been manually curated."
Meet the AI algorithms that judge how beautiful you are - "Qoves started as a studio that would airbrush images for modeling agencies; now it is a "facial aesthetics consultancy" that promises answers to the age-old question of what makes a face attractive."
AI Tool "Deep Nostalgia" Lets You Reanimate Your Dead Relatives - "Finally! A way to make old family photos even creepier"
The Rise Of AI Voice Assistants In Clinical Documentation - "Medical decision-making must remain with clinicians, but why does cumbersome data entry work continue to bog down their time? Can AI be used to allow physicians to spend less time on administrative tasks and more on value-added care?"
Intel's 3D and AI tech now helps train athletes - "Intel today revealed that its 3D Athlete Tracking (3DAT) technology is being employed by Exos, a firm that focuses on human performance conditioning, to help train professional athletes aspiring to join the National Football League (NFL) and other organizations."
Pinterest details the AI that powers its content moderation - "Join Transform 2021 for the most important themes in enterprise AI & Data. Learn more. Pinterest this morning peeled back the curtains on the AI and machine learning technologies it's using to combat harmful content on its platform."
Multimodal Neurons in Artificial Neural Networks - "We've discovered neurons in CLIP that respond to the same concept whether presented literally, symbolically, or conceptually. This may explain CLIP's accuracy in classifying surprising visual renditions of concepts, and is also an important step toward understanding the associations and biases that CLIP and similar models learn."

Concerns & Hype

Tom Cruise deepfake creator says public shouldn't be worried about "one-click fakes" - "Weeks of work and a top impersonator were needed to make the viral clips"
Clearview AI Is Taking Facial Recognition Privacy To the Supreme Court - "OneZero's General Intelligence is a roundup of the most important artificial intelligence and facial recognition news of the week. Clearview AI plans to challenge an Illinois law guarding against secret private facial recognition databases in the Supreme Court, according to Bloomberg Law."
Facebook Weighing Up Legality of Facial Recognition in Upcoming Smart Glasses - "Facebook is reportedly weighing up the legal implications of building facial recognition technology into a pair of smart glasses that the company is currently developing and which it intends to launch later this year."
"This is bigger than just Timnit": How Google tried to silence a critic and ignited a movement - "Big Tech has used its power to control the field of AI ethics and avoid accountability. Now, the ouster of Timnit Gebru is putting the movement for equitable tech in the spotlight."
Timnit Gebru was fired from Google - then the harassers arrived - "Even three months after Gebru's controversial termination from the AI Ethics team, the sustained campaign of aggressive tweets and emails keeps coming"
AI ethics research conference suspends Google sponsorship - "The ACM Conference for Fairness, Accountability, and Transparency (FAccT) has decided to suspend its sponsorship relationship with Google, conference sponsorship co-chair and Boise State University assistant professor Michael Ekstrand confirmed today."

Analysis & Policy

As China Rises, the US Builds Toward a Bigger Role in AI - "After decades of staying out of industrial policy, a Pentagon-appointed commission recommends more spending on research and support for US chip makers."

That's all for this week! If you are not subscribed and liked this, feel free to subscribe below!

Last Week in AI