Traveling back in time to discover the machines of the future

Psychology from the 1920s, 1960 statistics, today’s smartphones and tomorrow’s energy-efficient houses. As surprising as it might be, the common denominator of all these things is the Artificial Intelligence. Long before AI was an official research topic (for the VUB this was some 36 years ago), researchers were already unknowingly laying the grounds for it. Who knows which other subjects turn out be a contribution in the future?

 

Have a chat with a robot

In 1956, John McCarthy – the father of AI, proposed that with just two months of a detailed research, an elite group of researchers would have “found how to make machines use language, form abstractions and concepts, solve kinds of problems now reserved for humans, and improve themselves”. 60 years later, this is yet to be done. Nonetheless, the advances in the area are stunning – incredibly hi-tech developments in the field have been based on elegantly simple concepts from seemingly completely unrelated fields of science.

How do you think your smartphone knows when to tell you it’s time to leave asap to arrive on time for you scheduled meeting, and you actually do, because it even “thought” to take into account the current traffic? Or how does Siri or Google understand when you ask information using simply your voice with its specific accent and mannerisms of speaking? And while these incredible functionalities have been developed and are being improved on a daily basis, there are things machines can’t learn on their own,  but need to be thought. One of the most interesting examples is sarcasm. Computers, like small children, do not understand sarcasm. Unlike children however, they will not learn it on their own. This is where machine intelligence struggles to catch up with its creators.

The Smart House that does it all for you

In the strive for a more sustainable development and a stable environment, one of the biggest playgrounds of the current day for AI researchers is the smart machines. In the AI lab of VUB, researchers are working on ways to make your house so that it provides you with the comfort you’d expect from a home. An actual smart house knows you like your living room at 22 degrees for the Saturday movie night, that you are at the gym every other day and you come home late, and it would adjusting the heating cycle based on your initial input and learning from your interactions with it. The “iHouse” does the most efficient laundry and dish-washing cycles, so that it ultimately saves you some on your monthly bills and lowers your carbon footprint. While the actual execution of such a machine might be a little more challenging, the concept on which it is originally learning involves the same rules by which you teach your dog to wait to do its business until you are in the park– punishment for mistakes and rewards for positive attitude. A principle described in detail already in the 1920s by a team of psychologists (the famous dog of Pavlov might ring a bell).

Where is AI going?

Among the comforts of everyday life with smart houses and machines, AI is being developed to also monitor and help elderly people, learning from their habits and observing and recording variations. And while today still it proves to be a borderline impossible achievement to create a running robot, we are already surrounded by learning machines which become more and more sophisticated by the day. So instead of worrying about being taken over by vengeful race of machines, plug in and stay tuned for the latest improvements coming out of the VUB AI lab and their colleagues around the world. The industry is definitely full in on those already, so don’t fall behind!

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s