For reasons that range from world domination to academic dishonesty, artificial intelligence has somehow managed to tick every box on critics’ hit lists, placing it in an awkward position. Just about every technologist out there seems excited about the prospects of AI yet many of those same people continue to express caution about its use. For now, the number of uses for the technology keeps climbing regardless of any dissenting voices.
One of those uses relates to apps, those little smartphone programs we use to entertain ourselves. Advances in the space mean that Netflix can deliver high-resolution streaming on entry-level devices and Twitter can keep feeding us warnings about artificial intelligence on the go. Similarly, the Betfair website and app can provide exclusive thrilling online slots without the need for much technological horsepower.
In the latter case, even live-streamed casino lobbies can be rendered on a mobile phone, something that had previously only been possible on full-size computers. What all this means is that apps are becoming increasingly complicated behind the scenes, even as they become more user-friendly up-front. This is where AI comes in, specifically, its unmatched ability to collect and sort data.
Toward the end of May, the Microsoft Build conference in Seattle began with a series of live sessions on various topics. Inevitably, several of these were about artificial intelligence, including talks on how to build AI with component company NVIDIA, get started with generative AI in Azure, and work with AI in the cloud. Stuffed in with the robots, a spot about securing APIs seemed rather lonely in comparison.
Returning to apps, the future use of generative AI, i.e. AI that can create something, could have significant ramifications for online shopping. Microsoft plans to use its AI models to summarise the (human) reviews in the Windows Store. This would produce a short piece of text describing an app’s usability, performance, and other important things. Overall, it could reduce the influence of the loudest voices, giving less biased reviews.
The good thing about this seemingly innocuous use of AI is that it takes full advantage of the technology’s biggest strength, namely, making sense of a large volume of information. Rival uses at present, like chatbots, tend to produce disappointing results, whereas AI reviews can easily slot into any e-commerce site, whether that’s Walmart or something more niche, like a sports equipment store.
The slightly more disappointing news is that this use for generative AI seems to have been mentioned in little more than a passing comment, suggesting that it’s either not ready for use yet or is only meant to serve as a tech demonstration. The idea that a site dealing with billions of reviews, such as Jeff Bezos’ Amazon, wouldn’t be interested in AI tools seems unlikely though.
Overall, AI is as much of a buzzword as a genuine solution but it’s these practical everyday uses that can elevate it above the question of its own value, rather than the pursuit of thinking and feeling machines.