“Say AI again, I dare you, I double dare you!”
Four in ten of Europe’s artificial intelligence start-ups demonstrate little evidence of actually using artificial intelligence in their products and services, according to a new report that explores the hype around the technology.
London-based investment firm MMC Ventures, in association with Barclays, analyzed 2,830 AI start-ups in Europe and based their findings on public information (mostly the companies’ websites and available collateral) and interviews with executives.
The findings should come as no surprise to anyone involved in the tech or startup scene — AI is mostly hype, and in many cases (40% according to this study), it’s nothing more than a buzzword. I recently sat through a demo day of a highly respectable UK accelerator and not surprising all of them had AI “miracle” built into their solution.
But the buzzing activity around AI doesn’t stop on the marketing end- it rolls over to patent offices. AI-related patent applications in Europe grew at a 54% annualized rate from 2014 to 2017. Luckily- patent officers are more stringent than your ordinary folk, meaning that, many of these AI-based patent submissions were rejected: “The European Patent Office (EPO) recently issued fresh guidelines last year examining the patentability of AI and machine learning based on the principle that all patentable innovations must not be obvious.”
When we analyze these finding we can come to the following conclusions:
1. AI isn’t considered a novelty anymore — if everyone’s using it, then there’s nothing new or exciting about it. The only exception, in my mind, is companies, like Binah.ai, etc., that is all about “pure” advanced AI.
2. AI could even be considered a commodity — AI components can now be purchased from OEM software developers and can be integrated with practically any product. Just as the fact that your product is “An App” or works “in the cloud” or has BI build into it, and no one expects you to develop these components all by yourself (or gives you credit for doing so).
3. AI isn’t a product. AI is a technology and should be viewed as such. Many companies that have tried selling AI (or any algorithm for that matter- like anomaly detection) are no longer with us. Lack of a robust business case and customer understanding led to their demise. Customers don’t buy an algorithm, as they have no budget item titled: “Algorithm”. They can buy a network product, an endpoint, a monitoring product. All these can and should use AI. But selling “AI” to these buyers just won’t make them reach for their wallets.
4. When marketing a product that relies heavily on AI, like any other product, focus on tangible benefits, or actual use cases. It is futile to argue “Who’s AI is bigger”, or even to try to explain the inner workings of your AI. People who trust you will trust that the product will deliver what it promises. No customer is capable of “looking under the hood” anyway. Instead- focus on what your product (and awesome technology) will do for them — more speed, less churn, better prevention, increased ROI, etc. These are tangibles both technical, and business people can appreciate.
5. Don’t fall for the “next gen trap”: Machine learning and deep learning are more sophisticated forms of AI, and as such, could not and should not be marketed as a product.
In a sense, you can say that AI fell victim to its own success. It’s now something everybody wants and expects to have built-into their products. But companies and entrepreneurs now realize that people look beyond the hype cycle, so they should too.
At ForceNock, our product used advanced, machine-learning algorithms. But we never sold it as such; we merely stated it as a technical feature in our product. I was somewhat proud of my “buzzword free” presentation. The fact that the product was effortless to install, a breeze to use, and generated real measurable security benefits, meant so much more to our potential clients than what was “under the hood”. Yes, AI may be the engine driving your product, but it is not THE product itself.
Keep Safe, Dotan
