Where is Artificial Intelligence today?

Artificial Intelligence (AI) is no more restricted to science fiction movies, regardless of the popularity of Spielberg’s eponymous feature film of 2001, and how much you believed or not that it could come true one day. AI is now one the biggest subjects in high-tech in particular and science in general, and already is finding concrete, daily applications.

According to Chris Smith from the University of Washington “AI has been studied for decades and is still one of the most elusive subjects in Computer. The term was first coined by John McCarthy in 1956.”

There was a tremendous amount of literature on the subject in the early 1980s, and the Japanese government, in particular, invested billions to research it intensively back then, only to drop it in the mid-1990s, feeling that actual, tangible results would take many more years to materialise.

But of course the subject was never completely dropped and research continued in most other industrialised countries, although at a more reasonable, slower pace. Japan, like the other technologically advanced countries, also returned to working on it.

The fact is computer programmers now want computers not just to calculate (or to compute, another word for it), which is what they have been doing since the beginning, but to take decisions, to “think” in a certain way, to act on behalf of us human beings, in other words to act more intelligently.

Now, whether the expression AI is justified or not, whether it can be compared to our own intelligence is another topic and it remains a debatable matter. It would entail an accurate definition of what intelligence is. Nevertheless, the fact that machines are acting in a manner that is closer to our own behaviour makes no doubt, and robots that are now doing things that seemed impossible only 15 years ago. Examples abound.

Thanks to AI methods, “a system identifies violinists’ bowing techniques with 94 per cent accuracy. [… ] In playing music, gestures are extremely important, in part because they are directly related to the sound and the expressiveness of the musicians”. (sciencedaily.com). 

Writing about the eagerly awaited camera system in the upcoming (April-May this year) Huawei P30 Pro smartphone, artificialintelligence-news.com says: “AI turns anyone into a skilled photographer”. Whereas the artistic side of such a statement is subject to discussion, it makes no doubt that the very smart post-processing done by advanced smartphones cameras is helping talented and less talented photographers take shots that no one thought were possible with a mobile phone just a few years ago.

AI is also on its way in the judiciary world to replace human beings. According to the French lefigaro.fr, soon in Estonia AI systems will be pronouncing sentences instead of real judges. The Estonian government is working on the system, adding that it will be a world’s first, that at the beginning of its implementation it will be limited to treating minor cases, and that the sentenced party will still have the option to appeal.

Needless to say that computers working on AI will still act based on data collected and stored inside them, as it has always been the case. However, the algorithms put in place to analyse that data and then use it, is what AI is about and what differentiates it from “dummy” traditional programming.

To use a simple example in banking: a traditional programme will check your available balance and if found positive will let you withdraw money within the limit of that available balance, or if found negative will issue a message telling you there are not enough funds in your account to make the withdrawal. Instead, a programme with AI would tell you that based on your excellent history with the bank you could easily apply for a loan, or would recommend you use your credit card for a cash advance payment or would advise you wait a few days till your monthly salary or pension is credited to the account.

The field of application is huge, and so are the implications. One of the domains that are to be significantly affected by good application of AI is driverless vehicles. Such cars will need to do much more that to suddenly break if they detect a pedestrian crossing the street in front of you.

Acknowledging the impact of AI and the eventual consequences, Google has set up an AI Ethics Panel. Mentioning it bbc.com writes: “The group will meet to discuss how emerging technologies should — or perhaps should not — be used.”

One thing is certain, we have actually entered the age of AI applications. Computers can now be considered to be post-graduate fellows, thanks to AI methods and algorithms, and also thanks to technological advancements without which AI would not be functional at all today: ultra-fast processors, huge storage memory and gigantic wireless networks.