In mid-2018, Bernard Marr told Forbes that “there are 2.5 quintillion bytes of data created each day”. To help one grasp the immensity of this gargantuan number, that is the equivalent of 2.5 million terabytes. Divide that number by 2 and that’s how many $60 hard drives you’re looking at – you know, the ones that fit in the palm of your hand, assuming you had 1.25 million palms. In other words, there’s a lot of data on the internet. However, this is only a very recent phenomenon. When computers were but a novelty and only those with half a hockey rink of space available could own them, there was not a whole lot these machines could do. While Alan Turing’s “Bombe” computer, responsible for cracking the enigma code, was certainly a sight to behold, it pales in comparison to what a simple $500 laptop is capable of processing today; and thank goodness for that! Because, as appreciative as we can all be for Alan’s first Turing machine, our modern problems require solutions a little more advanced than what was available at Bletchley Park.
As the age of information took off in the late twentieth century, humankind began to rely more and more on the computational prowess of contemporary machines. Although with some criticism, computers were largely seen as keys to the future. Yet, early computers were ill-equipped for the tasks required of them. Before we could come to rely on them for the wide variety of tasks we require today, they needed to undergo a fundamental transformation. No longer are these machines expected to act mindlessly, performing the calculations we need them to do, only to put them away until the next project. Now, they need to respond to a wide variety of inputs, each arriving from different sources, in order to produce a dynamic range of outcomes. Indeed, in our world of big data, devices cannot be counted on to compute solutions through the sheer strength of their hardware; instead, we have imbued machines with the power of cognition, or in the words of Karek Capek – the man who first conceived of the concept – artificial intelligence (AI).
To help illustrate the transformation heralded by AI, I will use American author L. Frank Baum’s novel The Wonderful Wizard of Oz to compare the Lion and Scarecrow. In wanting a brain, Scarecrow yearns for a time in which he will be able to make sense of his environment. In many respects, that is what computer processing helps us achieve: by calculating inputs from its surroundings to deliver an intended output, the computer is able to make sense of that data. Conversely, Mr. Lion can already think; in essence, he already has a computer “installed”. However, as a cowardly feline, his computer takes input from his environment and outputs fear! Because fear is not a very useful emotion (output) to interact with the world, it needs to be corrected. But by what? By courage of course! By receiving courage, Mr. Lion is overcome with valour and is ready to brave the many frightening complexities of his world. Put differently, the “courage” program helped Mr. Lion rationalize the data collected from his surroundings and output a “functional” character, someone who is ready to confront adversity free from angst. This ability to process data selectively and in a manner that delivers the desired outcome (i.e. overcoming fear), is at the heart of what makes AI special. Rather than just understanding information and leaving users with a sophisticated jumble of data, AI extracts what is necessary and does so without the programmer having to intervene.
So what are these desired outcomes? First of all, AI automates processes, making them more efficient than if they were performed by humans. Indeed, by eliminating the need for people, AI-powered machines can execute complex tasks without breaks or delays. Amazon takes advantage of this benefit by integrating “more than 100,000 AI-based Kiva robots in their fulfilment centers”. These smart robots are capable of exerting themselves physically to displace cargo more massive than any human ever could and at a speed that would leave workers in a pant. As a result of their increased efficiency, AI-powered devices and services are also less prone to error. Pharmaceutical leader Novartis implements Machine Learning algorithms – code that teaches computers to think – to identify the efficacy of different substances under study. This eliminates the need of human scientists hovering over microscopes for hours, testing the reactivity of certain drugs. Due to the automated process, Novartis is able to output more accurate results at a faster rate.
These helpful qualities are not only reserved for big pharma’s R&D. Consumers such as you and I take advantage of AI’s helpful attributes every time we request help from Siri, Cortana, or Google Assistant. In a similar fashion, Bank of America “employs” AI to provide customer service. Its virtual assistant Erica is an “AI-enabled chat-bot” that has already helped more than “50 million clients” with requests such as bill payments, investment planning, and budgeting formulation. Apart from rendering helpful services, AI is being tested as a tool to help keep humans safe. In India, Microsoft is experimenting with HAMS (Harnessing Auto-Mobiles for Safety) to assess drivers’ state of alertness while on the road. By scanning their posture and facial expressions, HAMS is able to advise drowsy car operators on how to best prevent an accident.
While these are but a few of the benefits that AI bestows upon us, the demand for more advanced programs has created a plethora of opportunities for aspiring programmers. In academia, young scientists can dedicate themselves to unravelling the secrets of AI and study how two minds – one made of carbon, the other silicon – can coexist harmoniously. In robotics, opportunities are even more numerous. Not long ago, we witnessed the launch of NASA’s latest space exploration vehicle – Perseverance. Aiding NASA to make sense of Mars’ peculiar environment, this brave little rover was empowered by AI (for those interested in learning more about Perseverance, you can read about it here). However, if Earth is more a la mode, roboticists can have the time of their lives giving cars the ability to sense their surroundings with the help of LIDAR and AI. By giving automobiles the ability to understand their environment with the help of these technologies, companies such as Zoox and Waymo will continue to pave the way for the deployment of autonomous vehicles. As a final benefit for the hard work AI scientists dedicate to its creation, individuals in this field can hope to earn between $93,625 to $135,260 per year.
However, that compensation comes at a cost greater than just having to work longer hours. As AI proliferates throughout the world and its many markets, it brings with it numerous dangers. One that is already being felt is human obsolescence. McKinsey reports that AI-based robots stand to “replace 30% of the current global workforce”. Estimates by the World Economic Forum are even more damning as it predicts that by 2022, “artificial Intelligence automation will replace more than 75 million jobs”. These figures should not be surprising. If one were to direct their attention only at Amazon, they would see how its 100,000 Kiva robots already displace a roughly equivalent number of warehouse workers. The second danger introduced by AI is the encroachment on one’s privacy. As new devices are added to the internet of things and databases containing personal information are amassed online, pattern recognizing algorithms are becoming popular methods of making sense of the millions of terabytes people output daily. One such technique involves facial recognition, a ubiquitous tracking process instituted by national governments. Whether it be at airports or on the street, facial recognition infringes upon the privacy of regular citizens every day. Finally, there exists the danger of AI behaving erratically. Already, we have witnessed this occur when Microsoft’s Twitter chatbot began outputting racist messaging. Luckily, the plug was pulled on this failed experiment before the situation escalated. Unfortunately, AI’s racist relationships with the criminal justice system have not been dealt with in an equally responsible manner in the US. It is said that the Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) is an “artificial intelligence algorithm” that profiles criminal suspects and exhibits “[bias] against the black community”.
Thankfully, these dangers, and existing problems, are by no means incorrigible. In fact, through deliberate debugging, many of the software errors present in AI can be mediated. To achieve this, GSGF’s mission to train future generations of digitally literate professionals includes teaching them about civic responsibility. For us, it is not enough to simply train young students in the ways of technology. Rather, we seek to teach them how to become socially responsible global citizens. Citizens that can tackle the ethical dilemmas that artificial intelligence presents vis a vis racial profiling and infringements on privacy, while also preparing them for highly remunerative wage opportunities that will help them climb the social ladder, allowing them to sustain themselves, their families, and the environment. To realize our mission, we work ardently to fund the construction of green schools and deliver quality education to students. By supporting us, you will be actively contributing to the creation of a prosperous new world where we can all benefit from responsible technologies, built in no small part by the youth we educated. We welcome any donation, no matter how small. If you want to empower families in precarious positions and help their young ones develop high paying skills, here’s where you can make a difference!