The coronavirus pandemic is exposing the limitations of current artificial intelligence systems, particularly their struggle with extreme, unprecedented events. Many machine learning models, trained on historical data, fail when the present deviates significantly from the past, leading to unexpected performance degradation in areas like financial trading. Companies like Ocado have experienced issues where AI systems misinterpreted sudden surges in demand as cyberattacks. Experts advise continuous real-time monitoring of models, proactive identification of sensitive variables, and stress-testing systems with simulated crisis scenarios to improve their robustness. AI
IMPACT Highlights the need for more robust AI systems capable of handling unpredictable, real-world events.
RANK_REASON Article discusses the implications of an external event (pandemic) on AI technology, rather than a direct AI release or development.
- 2007–2008 financial crisis
- artificial intelligence
- Black Swans and Wormhole Wizards
- Boston
- COVID-19
- datarobot
- data science
- David Shriver
- Eye on A.I.
- Fortune
- machine learning
- Ocado
- SARS-CoV-2
- U.K.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
