AI has the potential to play an important role in supporting public health efforts in the event of a future pandemic. Last week, Dr. Marzyeh Ghassemi (University of Toronto Professor and a Scientific Advisor at Radical Ventures) discussed AI’s important role in combatting the pandemic. Dr. Ghassemi notes that AI will have the most impact in resource-constrained areas. For example, when it comes to symptom surveillance in communities where there is limited access to testing, AI systems can better predict how the virus is spreading and may help with early diagnosis.
While Dr. Ghassemi is confident in the opportunity presented by AI, she acknowledges its limitations. For example, the data used for training models in healthcare can be skewed when data is comprised of significantly more examples of unhealthy people versus healthy. That said, machine learning’s predictive capabilities should be viewed as an essential tool in our larger decision-making toolkit when it comes to tackling the next public health crisis.
AI News This Week
-
Andrej Karpathy on the AI in Tesla’s autonomous driving (The Robot Brain Podcast)
Pieter Abbeel, renowned AI researcher, professor and entrepreneur, launched a new podcast focused on AI Robotics. In the premier episode, Andrej Karpathy, Tesla’s Director of AI, discusses what it is like working with Elon Musk, training driverless cars, and the time he had to sleep on a yoga mat at Tesla HQ. Pieter is a co-founder of Covariant, a company that has built a universal AI software brain for robotics. Radical Ventures is an investor in Covariant.
-
If your company uses AI, it needs an Internal Board Review (Harvard Business Review)
Adopting an AI strategy is increasingly important for large organizations to stay competitive. This strategy should also address AI risk mitigation. Although the challenges around AI feel unprecedented, companies can adopt internal review boards (IRBs) – a process already well established in the medical profession – as a powerful tool that is well suited to address complex ethical questions.
-
Surfacing women in Smithsonian history (Google Arts & Culture)
The Smithsonian is using machine learning to mine two centuries of metadata to uncover new stories about women in the Smithsonian’s history. The project has confirmed “some of the Smithsonian’s most exciting early scientific research was undertaken by women.” The model uses a structured data set to identify “named entities” (such as people, places, or dates) in both the collection’s metadata and the texts of annual reports to uncover misplaced or overlooked stories.
-
Biden unveils $2tn infrastructure plan and big corporate tax rise (Financial Times)
President Biden’s announced commitment to improve US infrastructure comes as part of a multistage effort to reshape the US economy. The main goal for the transportation industry is to “modernize 20,000 miles of highways and roads; repair 10,000 bridges; and, by 2030, build a network of 500,000 electric vehicle chargers.“ This plan also includes a $180bn investment in AI and biotech R&D aimed at improving competitiveness with China. The combination of these investments in infrastructure, AI, and climate change may be reflected in the future design of our cities. Radical continues to invest in smart infrastructure technologies that are transforming our built world, such as Pittsburgh-based company RoadBotics and, this year, Promise Robotics.
-
AI for reducing food waste (Tech Xplore)
The Fraunhofer Gesellschaft (FhG), the largest organization for applied science in Europe, is working with 30 partners to design an “AI ecosystem” that reduces food waste at all points of the food product value chain. In the long-term, the ecosystem will enable a virtual marketplace for AI-algorithms in an effort to maximize the impact and scale of successful implementations.
Radical Reads is edited by Leah Morris (Senior Director, Velocity Program, Radical Ventures).