Are smart machines as smart as they claim to be?

Are smart machines as smart as they claim to be

The terms Business Intelligence, Artificial Intelligence, smartphones, smart watches, smart cities, smart buildings, and a long list of objects and concepts that are supposed to be intelligent or smart are on everyone´s lips these days. Although they are labelled as ‘smart’, is there real intelligence inside them?

To start with, from an IT point of view, can we say that smartphones are really smart? We cannot deny that it is truly amazing how we can do practically anything with our phone just by talking to Siri. “She” “listens” and immediately performs any action in the same way a human being would do. So, is our smartphone smart because its behavior is exactly what we could expect from a human being? For those who don’t know, what’s behind Siri or any personal assistant like “her” is a natural language processor that captures and interprets human voice and delegates the requests to a wide range of web services which have the answers to the questions or are able to process the corresponding requests. Nothing that a software engineer or a domain expert would call actual intelligence.

Smart watches are no more than an extension or another kind of smartphones: wearable computers with a reduced size, equipped with sensors, internet connection and a GPS signal offering full functionality.

More recent terms claiming to be smart are: Smart Cities, Smart Buildings, Smart Cars, Smart Healthcare, Smart Tourism, Smart Agriculture and Smart Transportation, among others. There is no universally accepted definition for them but they all share a common feature: the application of the Internet of Things (IoT) or related technologies. Environments such as cities, cars or buildings count on networks of sensors and cameras embedded in everyday objects that generate, process and receive data that is used to make the environment more livable, workable and sustainable. According to the IoT institute, Singapore claims to be the smartest city in the world, followed by Barcelona, London, San Francisco and Oslo. These cities are differentiated from normal cities because they have installed systems such as dynamic traffic lights (which adjust the timing of lights and the maximum speed allowed depending on the traffic), smart parking (assisting drivers to find a parking spot, thus reducing time, fuel, pollution and space), intelligent lighting (turning street lights on only when needed), Wi-Fi access points, etc.

When it comes to talking about Business Intelligence (BI), its purpose is clear: to support decision-making. Business Intelligence tools and techniques use the raw business data (i.e., data from transactions) to extract valuable and meaningful information that is then transformed into knowledge. A business can take advantage of the data from its sales, social media, its human resource department, its inventory etc. and find patterns and trends to customize marketing campaigns, decide where to open a new office, add new products or services to its catalogue or hire new teams with certain profiles. Business Intelligence decreases the risk of making decisions and there is no doubt about how powerful and extended it is. Nevertheless, although BI helps make intelligent decisions in businesses, there’s no actual intelligence in its processes or tools. BI success depends on their users applying their knowledge and skills to make the most of it.

SmartwatchWhen talking about Artificial Intelligence (AI), chess is one of the oldest, most popular and well-known case where it is applied. Back in 1996, Deep Blue – a super computer designed to play chess – beat the current world champion for the first time. Some people were, and still are, afraid of artificial intelligence and fear a future where robots take over the world just like in sci-fi movies. But machine intelligence is neither similar nor comparable to ours. For instance, what Deep Blue does is calculate a certain number of possible moves (both for the human player and the machine) and to select the one that leads to the best situation for the machine to win. This means that it is not more intelligent than we are, but just faster than we are at making calculations. Without any time restrictions, a human being could do the same and win.

One of the foundations of Artificial Intelligence is Data Mining. Just as BI does, it explores data in order to find hidden relations. It was successfully applied in 1996 when it was discovered that men in their thirties and forties shopping on Friday evenings and purchasing diapers were more likely to also buy beer. Stores then placed beer next to diapers and the sales of beer instantly increased 35%. Of course, Artificial Intelligence goes further than exploring a set of possible states or finding data relations, it is composed of many techniques and algorithms applied to different domains. But still, AI solutions are problem-specific and there’s no human intelligence in them.

In my opinion, there is much more smartness than deserved in all these ‘sexy’ terms. Independently of the name they receive, the truth is that technology is everywhere and is changing the world. And, like it or not, we are changing with it.

By Lola Burgueño