Once you start using the Nero 1, it captures and prompts you to tag the faces it identifies. This lets the Nero 1 remain passive in the event that it sees a known face while alerting you to new faces who could represent potential intruders.
What’s interesting about the facial recognition feature is that Butterfleye developed it using the Amazon Rekognition API. In other words, Amazon built the general image recognition algorithm, and Butterfleye is now using it for the specific case of facial recognition by applying the algorithm to the facial data that it collects.
This is a great example of the commoditization of AI algorithms. As more people have access to these algorithms, the source of value increasingly shifts from the algorithm itself to the data to which the algorithm is applied.
I recently clicked on a link to a Wall Street Journal (WSJ) article behind a paywall. I was shown a $277 annual subscription price and, in an attempt to see if there was a free alternative source that covered the article’s contents, I closed the page and performed a Google News search. I clicked on an article on the same topic but it turned out to be just a brief summary of the WSJ article I was looking to access, and it linked to the WSJ article.
On my second visit to the WSJ, I had already made up my mind to buy the annual subscription to read the article. However, to my surprise, this time I was quoted a $177 annual subscription price rather than the original $277. Since I had not purchased the subscription on my first visit, the WSJ offered me a lower price to convert me into a customer on my second visit.
In this case, the WSJ lost money. This is because I was already prepared to pay the $277 annual subscription price on the second visit independent of the additional discount. However, I imagine that the second visit price is optimized across all of WSJ’s potential subscribers. In most cases, the reduced price helps the WSJ convert a visitor who otherwise wouldn’t have paid to access the WSJ. So, in the aggregate, the WSJ wins. It’s an example of customer segment-based pricing.
However, in the future, as these algorithms gather more data points about individual visitor behavior, they will grow smarter. And this will let them move from customer segment-based pricing to customer-based pricing. At the limit, companies who use data to inform their strategy (pricing or other) will not only win in aggregate, they’ll win every time.