AI is increasingly creeping into modern business, as organisations seek to improve efficiencies and speed up their processes. It’s very easy to get this wrong, however, as the UK government recently found out. An AI algorithm designed to issue students’ final grades for A-Levels and GSCEs has caused uproar over the last few weeks as complaints of unfair and inconsistent results led to widespread protests.
The Department for Education was eventually forced to backtrack, announcing that the algorithm-generated results would be scrapped in favour of grades based on teachers’ predictions. But why did this AI system get it so badly wrong, and what can businesses learn from this debacle about deploying AI within their own organisations?
In this episode of the IT Pro Podcast, we look at what AI can and can’t do, whether it should be placed in charge of important decisions, and why it’s not a good idea to take your hands off the wheel. For the next four weeks, we’ll also be testing out a new format – instead of having a news roundup, each week we’ll be focusing solely on an in-depth discussion of that episode’s topic. Let us know what you think!
Footnotes
- Self-driving cars could be on UK roads by spring 2021
- UK gov's "unfair" A-Level algorithm faces several legal challenges
- Police use of facial recognition ruled unlawful in the UK
- Home Office to scrap 'racist' AI tool for visa applications
- Businesses put AI innovation on hold to avoid public backlash
- Why humans and machines work better together
- AI Weirdness: A neural net tries to write fanfiction
- Do we need an algorithm police?
- MPs claim AI 'doesn't need' a new regulator
- Why transparency is key to promoting trust in artificial intelligence
- How the driverless vehicles dilemma highlights wider system safety issues
- The A-level results crisis has once again shown the limits of AI
Subscribe
