The Future Promises and Challenges of AI

My interest in and passion for Artificial intelligence must be spanning over a decade now, but in comparison I have written very little on or about the topic. The problem was that despite continuously reading about the latest trends and development and while considering and reconsidering ethical issues and implications regarding this budding new technology, I was missing some relevant data and pieces of information. That was filled in beautifully at the Space Centre’s Cosmic Nights event I attended recently.

This was my first time attending the event, and I was lured in quite easily and effortlessly by its enticing title AI The Changing Face of Technology as well as its illustrious line-up of renowned experts, in particular Dr. Kevin Leyton-Brown, the Director of UBC’s Centre for Artificial intelligence Decision-Making and Action (CAIDA).

It was my first visit to the Space Centre, but I enjoyed its various parts and programs from trying to guess what responses were given by humans versus machines as a brief and subtle introduction to the Turing Test to a fun and interactive trivia section that included references to robots and popular films and culture building up to a planetarium show that combined robots with space technology and provided details and information about the physically and psychologically straining and exacting Mars Settlement Project.

But from an academic point of view – and academia being my bread and butter or rather wine and cheese - the highlight of the evening was the talk with the somewhat bland title of “Artificial Intelligence: The Journey so far, and the World in 2029” by Dr. Kevin Leyton-Brown. This well-spoken and humorous computer science expert adeptly tackled the potential upsides as well as downsides of AI technology and most of my blog post is based on information gleaned from his lecture.

The Frankenstein Complex: The Uprising of Robots

One of the biggest fears concerning AI is often named The Frankenstein Complex. When given a positive spin, this can simply be perceived of the student outperforming the master, but in its more negative sense, it implies that the creation has turned against its master, just like the monster / life that Frankenstein created slipped out of the scientist’s control and went against his desires and wishes.

This struggle between human and artificial minds has been depicted predominantly in sci-fi literature and movies. This ranges from 2001: A Space Odyssey (1968) with Hal taking active control and rebelling against his human masters to the dystopian world of Blade Runner (1982) in which androids run amok. More recently, we also have emotional and ethical entanglements, whether it is dealing with a robot boy propelled to become human by the love for his mother as in Spielberg’s A.I. Artificial Intelligence (2001) or a female entity as a man’s love object in Spike Jonze’s Her (2013).

Then, there is of course, the collection of stories I Robot  by sci-fi writer Isaac Asimov and his proposal of the three laws of robotics, which stipulate that a robot may not injure a human being, that it must obey the orders given by human beings and finally that the robots must try their best to protect their own existence. 

Evidently, the laws try to ensure protection by highlighting the higher status and position of the master human so that overthrow and rebellion do not occur. Yet despite the proposed safeguards, scientists and laypeople alike are particularly nervous about the appearance and existence of a super-intelligent artificial entity sometime in the near or far future.

The Transition from Narrow to Broad AI

There are two ways of perceiving AI: It can be either narrow and weak or broad and strong. Essentially, narrow AI tackles what it is programmed to do. It could consist of playing chess games or help factories and companies speed up production. It could help detect fraud or use face recognition as a means of increased measures of security. 

Yet with each passing day, the narrow line gets a bit wider as AI manages to get more and more skilled at doing specific tasks. For instance, the early chess programs like IBM’s Deep Blue could occasionally be defeated by the very best human players, but then it evolved to Komodo, which is virtually unbeatable.

Since that has become the norm, the AI industry has spread into other more complicated areas and scenarios ranging from checkers to Go and video games and even extending to games of poker. Poker games are more complicated than chess games as they involve not only strategy but also guess work, speculation and of course bluffing. 

In this case, it is more difficult for AI to crack the game as it is not so much about the plays and cards that are visible, but covert cards and manipulative plays often lead to wins in this type of game and competition. While it seems a given that computers would reign the world of chess – it is a game that would be natural for computers to excel at - it would be a somewhat different issue when it comes to manipulating and bluffing to gain the upper hand as in the case of games of poker.

Yet all of this is now possible because the skill set of AI has grown dramatically, not unlike Moore’s law and prediction that computer skills and abilities would increase exponentially over time. Our smartphones are more powerful than the best computers of old and now each of us (at least those who sport newer iPhone models) are even able to use face recognition, something that was unheard of a few years back. I was even told that my Vivino app not unlike Google images uses algorithmic image recognition software to find potential matches to queries.

As a result, AI is rising up to meet new challenges and because of its continuously growing impact and influence in society, more and more people are signing up for and flocking to this field. Enrollment in university programs and courses have sky-rocketed and many are not only interested in the field but also see it as a lucrative career choice for the future. It is decisively too late for me to switch careers now, but I can at least do whatever it takes to keep afloat with these technological innovations happening right under my nose and feet.

One of the upcoming changes but also challenges would be self-driving cars. Experiments have been rather promising despite certain occasional and necessary missteps and hiccups. However, there is a certain double standard here as it seems that humans are easily forgiven an accident or two but people and media pounce upon any errors committed by Artificial Intelligence. 

But the fact remains that due to its inherent dangers, autonomous cars are more complicated and there are a host of factors that need to be considered before they can be launched on a global scale. Dr. Leyton-Brown believes that there will be imminent success for automatic hauling trucks as the highways tend to be more predictable than city streets; chaotic city traffic including maniac drivers during rush hour combined with unruly bikers, rebellious jay-walking pedestrians and variable weather conditions still pose many challenges for self-driving cars.

For autonomous cars to become a full-fledged ubiquitous reality, AI must be complex and broad enough to make difficult decisions ranging from traffic to weather conditions. For it to work successfully, AI must be moving away from simple programming to what is known and referred to as machine or deep learning; put differently, AI needs to be able to become more autonomous and independent by being (almost) able to think for itself.

The Economic Impact and Potential Threats of AI

All these changes and development will have economic impacts and consequences, but the situation is, according to Dr. Leyton-Brown, not as bleak as many make it out to be or want us to believe. AI will not come and replace all our jobs, but certain professions will be affected more than others. If you are active in specific professions, your future may not be too bright. Dr. Leyton-Brown jokingly claimed that he would not advise his son to enter the trucking business or start training to become a radiologist, for that matter.

In fact, AI will generally change and replace lower-skilled jobs. Any tasks that are repetitive in nature could be easily and simply replaced by machines and computers. In fact, Dr. Leyton-Brown believes that the assembly line was an invention that essentially considered and regarded humans as robots, and now in turn these actual robots would be indeed much better at replacing humans trying to act as robots.

These types of jobs and tasks of similar ilk will be handed over to machines simply because they will be much better and faster, hence much more productive at them than their human counterparts. Humans tire and need breaks to rest and eat, but machines will be able to work tirelessly even through the night. They will also not complain about working conditions and will not ask for a raise in salary, so from those standpoints and from the point of view of the employers, it makes sense that machines are often preferable to humans.

Due to powerful algorithms and processing power, AI will also work best when it comes to providing consumer choices. Whether this is in the advertising and marketing business or whether it is about choosing ideal and inexpensive travel arrangements and hotel accommodations, AI will be much more efficient, effective and less error-prone than humans. As a result, travel agents and call centers will be in peril as they simply cannot live up to the impossibly high standards set by computer technology.

Yet many other jobs will remain untouched by AI as the human touch is an integral part and of paramount importance to them. Health care, for instance, will be for the most part unaffected by these technological advances. People would not want to run to a robot for a diagnosis of their ills as they would like to share their medical issues with a (hopefully) knowledgeable, experienced but also caring medical professional.

On the other hand, radiologists whose jobs consists of processing data and information to detect anomalies could be replaced by the skill set of better trained and equipped AI. The latter would be able to compare and contrast healthy images and scans from those that pose potential health risks and problems in a much faster, more accurate, more detailed and expansive manner than humans possibly could.

That being said, most of us would shudder at the idea of having robots provide childcare. Robots would certainly be able to offer security and safety, but that is not the only reason we drop off our children at a daycare. When we leave our children at a day care, we want them not only to be safe but also to be in an emotionally safe and caring environment where they can learn and explore culture and society as well as engage in human interaction. This is an area in which robots would not be able to thrive and could not possibly replace humans.

The same can be said about education in general. Although translation services are going to be significantly improved with AI – goodbye to the challenging and arduous task of ad hoc simultaneous translations – and grade computation will be facilitated by the use of relevant and effective software, AI would still not be able to provide cultural information or individual feedback to students, at least not in the sophisticated manner and measure that well-trained and effective teachers and instructors are capable of. So, thank God, education is a field that ought to survive the onslaught of AI.

Similarly, counseling and psychotherapy are not careers, which can be easily taken over by automatons. They are based on empathy, understanding and personal experience as well as tact and feeling that will elude robots no matter how well they are programmed. Coaching whether it is regarding sports or more broadly life, will also be closed off to AI. In all these professions, the human touch is essential and absolutely necessary for those endeavors to be successful.

What would happen, however, is the fact that AI – not unlike technology since its inception - will provide us with even more freedom and free time. Routine tasks will be taken over and will fall under their domain. This means that we have more time and resources to focus on other more complex tasks. Basically, this would mean more free time for all of us and the opportunity to advance knowledge and skills by using technology to do tasks we did not want to engage in in the first place.

Nonetheless, the main threat of AI is tied with government and military expansion of missile range and capabilities. There have already been many advances that may be in the best interest of a given country, but they are definitely not in the best interest of humanity. Drones are not only able to “see” hidden or concealed weapons like AK-47s on humans, in cars and buildings, but they have also been programmed to make autonomous decisions, such as attacking them under certain circumstances. Killer robots, robot soldiers and autonomous aircraft all pose a danger because although machines are more precise and accurate, they lack the complexity of a human being, especially when it comes to ethical judgments or on the grounds of empathy and understanding. 

But before we panic over super intelligence, and have nightmares about AI taking over the world, we ought to also consider the fact that such super intelligence can also be used to our own benefit. It can provide us with novel information about how to save the climate or deal with famines, water shortage and overpopulation, while it also has the potential to offer peaceful and productive solutions to all of us and not only the parcel of land that we call our home. As usual and not surprisingly, it all comes down to how we use and harness the prowess and powers of technology alongside our underlying incentive and motivation, whether we choose to use it to the benefit or detriment of humanity.