AI washing is a marketing effort designed to imply that a company's brands and products involve artificial intelligence technologies, even though the connection may be tenuous or non-existent.Content Continues Below
AI is currently the focus of the same kind of market hype that led to cloud washing and, before that, green washing. According to Gartner, over a thousand software vendors describe themselves as artificial intelligence vendors or claim that their products involve AI. Gartner analysts consider AI mischaracterization to be one of the top three problems that are impeding real development and adoption of artificial intelligence.
To determine whether or not a given product incorporates AI, it is essential to have a good understanding of what artificial intelligence is. Most of the products that actually do involve artificial intelligence technologies are examples of weak AI. Weak (or narrow) AI employs artificial intelligence technologies in a high-functioning system that replicates – and perhaps surpasses -- human intelligence for a dedicated purpose. In contrast, strong AI (also known as artificial general intelligence or AGI) represents generalized human cognitive abilities in software so that, faced with an unfamiliar task, the AI system could find a solution.
Bridget Botelho offers questions to ask to reveal AI washing in a product:
- Does the product have capabilities that are far in advance of current counterparts that are not designated as AI?
- How does the product's vendor define AI?
- What specific AI technologies are used?
- How does the vendor intend to stay current with the developing technologies of AI?
Gartner research VP Jim Hare maintains that you should look into products assuming that they aren’t AI and leave it up to the vendor to prove otherwise. In any case, Hare doesn’t believe that
AI is necessary for many of the uses it’s being touted for: “Really, what you need to look for is a solution to a problem you have, and if machine learning does it, great," Hare said. "If you need deep learning because the problem is too gnarly for classic ML, and you need neural networks -- that's what you look for."