Navigating through extreme events using Artificial Intelligence

Artificial Intelligence (AI) has the ability to free up employees’ time by automating tasks and forecasting future events, giving them time to focus on other work that requires human intelligence to complete. It does this by using the business data to predict what’s likely to happen next. However, in rare, extreme events like the global pandemic we’re currently experiencing, markets are sent reeling and all bets are off. Most AI systems do not cope well in these scenarios. In this article we’ll look at how AI systems have struggled in financial and retail industries, as well as how they can be adapted to cope with these extremes.

AI driven investment strategies

Most of today’s machine learning systems need to be trained on lots of historical data. But what happens when the present suddenly stops looking like the recent past? Most A.I.-driven trading algorithms, for instance, have only been implemented in the last five years. Their training data might not even have included the 2008 financial crisis. They almost certainly don’t include anything like the massive demand-driven shock we’re seeing across all industries right now. So, some A.I.-driven investment strategies that were supposed to do well in all kinds of different market conditions have actually performed much worse than expected in the past few weeks.

AI for retail

In the past month Ocado, the online supermarket, has seen traffic to its website spike four times higher than any previous peak the company has experienced in its 20-year history. This caused problems in itself –  so many visitors went to its website that the company’s cybersecurity software, which uses machine learning to detect unusual behaviour, assumed the site was experiencing a denial of service cyberattack and moved to block those connections. Luckily, human operations managers intervened to prevent that from happening.

How should AI systems be adapted to cope with these extremes?

  • It’s vital that businesses monitor their data models in real-time and lookout for anomalies that could cause problems. If a product is suddenly selling at 10 times the normal rate a human might need to step in and amend the processes in place.
  • Businesses need to be proactive about which machine learning models and which input variables within the models are most sensitive to extreme events. Anything that depends on human behaviour—from electricity demand to shopping—will be affected by major events.
  • The business’s data scientists should sit down with subject-matter experts and stress-test a system in simulation: What items might customers want in a crisis? And what will happen to your supply management algorithm if you do get thousands of people wanting to purchase six months’ worth of toilet paper in a week?
  • Although this exact scenario hasn’t occurred before, businesses should look for other major events that may have caused similar outcomes in the past. What can be learnt from these?
  • Finally, data scientists need to think carefully about whether they want the current coronavirus extremes included in future training data. For some systems, doing so might prevent the software being caught off guard in the future. But in a lot of other cases, it might have the opposite effect, leading the system to falsely expect that the crisis reflects a “new normal.”

The events we’re confronting today help us to remember that AI is still in its early days, learning from new data as it progresses and evolves. We’re a long way from machines replacing humans in the workplace, but AI can, and already does, play a useful supplementary role. Learn more on our data analytics and insights page, or contact us.

Related insights

Nothing found.

Enabling specialist UK businesses to unleash their true potential.

Get in touch