Artificial intelligence leads the top tweeted terms are the trending industry discussions happening on Twitter by key individuals (influencers) as tracked by the platform. The steps being taken to integrate artificial intelligence (AI) into healthcare and the use of AI techniques in the detection and management of various diseases were popularly discussed in Q2. Rafael Grossmann, a surgeon and clinical innovator, shared an article on two new companies namely Anumana and Lucem Health being launched by healthcare company Mayo Clinic that will collect and analyse patient data gathered from remote monitoring devices and tools to enable early detection and diagnosis of diseases. Mayo Clinic will launch a remote monitoring platform that will enable clinicians and physicians to make quicker and better decisions with the help of the collected and analysed patient data thereby speeding up the diagnosis before symptoms appear. It will also allow patients to take more control of their health and related decisions.

US federal agencies are planning to expand use of facial recognition systems, according to a report by the Government Accountability Office (GAO) despite continuous backlash over the technology's application for more than a year. The report, published Tuesday, assessed the use of facial recognition systems by federal agencies, and how they plan to expand the use of the technology in the future. Eighteen of the 24 surveyed agencies, including the US Departments of Justice, Defense, Education, Housing and Urban Development, reported using facial recognition technology (FRT) for one or more purposes, the GAO report said. The survey also found that 10 of the agencies plan to broaden their use of the technology by 2023, with two of them investing in its research and development. While most of the facial recognition systems used by the federal agencies are government owned, the report says six such systems come from commercial vendors like Clearview AI, and Acuant FaceID.

Excelling at a test often does not translate into excelling at the skills the test purports to measure. This is true not only of humans but also of AI systems, and the more so the greater the claims of the test's significance. This became evident less than a decade after the introduction of the Winograd Schema Challenge (WSC),3 a test designed to measure an AI system's commonsense reasoning (CSR) ability by answering simple questions. An example would be, given the information: The sculpture rolled off the shelf because it wasn't anchored, answering: What wasn't anchored? There are multiple AI systems2 that achieve human performance on the WSC but are not capable of performing CSR.

A team of researchers at the University of Cambridge, in collaboration with the Alan Turing Institute, constructed sophisticated machine learning tools capable of detecting dementia in its early stages. Published in the journal NeuroImage: Clinical, researchers took brain scans of numerous patients who subsequently developed a neurodegenerative illness, like Alzheimer's disease. With their machine learning algorithm, researchers were able to spot structural changes in the brain. According to the study, the AI-based algorithm was more than 80 percent accurate in predicting the onset of dementia among the participants. The algorithm also predicted the speed at which cognitive function declined.

After a dense presentation about the undeniably impressive work Tesla is doing with AI, the company's self-anointed Technoking, Elon Musk, capped the evening by bringing out a dancer in a spandex suit. Behold, said Musk: my Tesla Bot. The dancer in the suit, he said, was the model for a new humanoid robot Tesla will produce in the near future. After the dubstep and applause had faded, the vaguest of briefing slides promised that the Tesla Bot will stand five feet, eight inches (1.7m), weigh 125 pounds (56kg), have "human-level hands," and eliminate "dangerous, repetitive, boring tasks." Musk said that building a human-replacement robot -- something no company in the world is close to achieving -- was a logical step forward from Tesla's work developing self-driving cars.

Early this year, the US Federal Trade Commission forced facial recognition startup Paravision to delete a collection of improperly obtained face photos and machine-learning algorithms trained with them. FTC Commissioner Rohit Chopra praised that new enforcement tactic as a way to force a company breaching data rules to "forfeit the fruits of its deception." Roth and collaborators from Penn, Harvard, and Stanford recently demonstrated a flaw in that approach, showing that the unlearning system would break down if submitted deletion requests came in a particular sequence, either through chance or from a malicious actor. It will take virtuoso technical work before tech companies can actually implement machine unlearning as a way to offer people more control over the algorithmic fate of their data. Binns says that while it can be genuinely useful, "in other cases it's more something a company does to show that it's innovating."

The robots are coming, and Elon Musk wants to usher in their arrival. The Tesla CEO revealed the Tesla Bot this week during an event devoted to advances in artificial intelligence. Described as "the next generation of automation" on the Tesla website, the humanoid robot is "capable of performing tasks that are unsafe, repetitive or boring." That makes sense as there are many jobs needing done but humans don't want to do, Musk says. At the foundation it is labor," he said during the event.

The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. Many systems like autonomous vehicle fleets and drone swarms can be modeled as Multi-Agent Reinforcement Learning (MARL) tasks, which deal with how multiple machines can learn to collaborate, coordinate, compete, and collectively learn. It's been shown that machine learning algorithms -- particularly reinforcement learning algorithms -- are well-suited to MARL tasks. But it's often challenging to efficiently scale them up to hundreds or even thousands of machines. One solution is a technique called centralized training and decentralized execution (CTDE), which allows an algorithm to train using data from multiple machines but make predictions for each machine individually (e.g., like when a driverless car should turn left).

After spending years looking into 30 separate Tesla crashes, this week federal safety officials finally took a step toward cracking down on the electric carmaker. On Monday, the National Highway Traffic and Safety Administration announced an investigation into Autopilot, Tesla's driver assistance system, which allows the vehicle to manage certain highway tasks like changing lanes and moderating speed, and which numerous drivers have treated like a fully autonomous driving system (sometimes for the entertainment of their social media followers). NHTSA's new investigation has a narrow focus: It will seek to determine why Teslas with Autopilot engaged have crashed at least 11 times into stationary first-responder vehicles. Depending on what the agency concludes, NHTSA could declare a "defect" in Autopilot, insisting that Tesla correct it or else face a hefty fine. NHTSA's power over the automotive sector shouldn't be underestimated; the agency's investigation in Takata's faulty airbags helped push the multi-billion dollar company into bankruptcy in 2017.

The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. Machine learning is undergoing a paradigm shift with the rise of models trained at massive scale, including Google's BERT, OpenAI's DALL-E, and AI21 Labs' Jurassic-1 Jumbo. Their capabilities and dramatic performance improvements are leading to a new status quo: a single model trained on raw datasets that can be adapted for a wide range of applications. Indeed, OpenAI is reportedly developing a multimodal system trained on images, text, and other data using massive computational resources, which the company's leadership believes is the most promising path toward AGI -- AI that can learn any task a human can. But while the emergence of these "foundational" models presents opportunities, it also poses risks, according to a new study released by the Stanford Human-Centered Artificial Intelligence's (HAI) Center for Research on Foundation Models (CRFM). CFRM, a new initiative made up of an interdisciplinary team of roughly 160 students, faculty, and researchers, today published a deep dive into the legal ramifications, environmental and economic impact, and ethical issues surrounding foundation models.