“AI-Wash” or Fake AI? AI hype confuses investors and media
New European research has revealed that up to 40 per cent of 2830 European-based AI firms who claim to use artificial intelligence (AI), actually don’t.
The research surveyed 2830 companies classified as an “AI company” to see if they met the characteristics and standards that you would expect. When lifting the hood, the research found no evidence that many of these companies used learning algorithms or anything else resembling AI.
Why have they classified themselves as an AI company then? The most common reason is to increase their chances for venture funding.
The research has thrown up many questions about what it means to be an AI company.
Here are three lessons we can take away from the research.
Lesson #1: You can’t always believe the name
First and foremost, the research proves that you can’t always believe that the letters “AI” in a company or startup’s name actually mean that the company uses learning algorithms.
Lesson #2: Greater classification is required
There is a need to classify between companies who develop their own algorithms, and those who use shelf products. And there is nothing wrong with the latter. In fact, shelf products and platform from Google, Amazon, Nvidia and algorithm market places such as Algoritmia have advanced significantly. They provide real value to commercial firms in need for AI deployment.
In fact, the Nasdaq CTA Artificial Intelligence & Robotics Index (see link to NASDAQ research unit document at the end of the article) is providing a clear market definition of publicly traded companies: AI Enablers, Engagers, and Enhancers.
AI Enablers are companies that develop the building block components for robotics or artificial intelligence, such as advanced machinery, autonomous systems/self-driving vehicles, semiconductors, and databases used for machine learning.
AI Engagers are companies that design, create, integrate, or deliver robotics and/or artificial intelligence in the form of products, software, or systems.
AI Enhancers are companies that provide their own value-added services within the artificial intelligence and robotics ecosystem, but which are not core to their product or service offering.
Lesson #3: Most future companies will be considered “AI”
This is perhaps the most important lesson: it is likely that almost every company will become an “AI company” in the near future. Not in order to attract investors, but because it will become standard practice to take company data and use it well. And at the moment, AI is the best way of doing this.
For example, almost every company today could be considered an “internet company”. It’s simply best practice, so it goes unnoticed in the company name. It will be the same for AI as technology progresses. With fewer deployment hurdles, AI will become more common. In the next decade, AI looks set to turn the entire economy on its head.