AI is one of the hottest topics of our day. The discussions concerning AI have long surpassed the limits of the scientific and entrepreneurial communities and became the main pastime topic on taxi rides, twice a year phone catch-up calls with acquaintances, workplace small talk and more. However, when people talk about AI, they tend to focus on the sexy version ; the version that boils the blood with fear, excitement, or hope. Despite its becoming mundane a long time ago with everyday technologies from chatbots to product dispersion machines, from visual classification tools to popular applications such as Shazam, people love to discuss the juicy sides of the topic:
Today, at the furthest – publicly available – point we have come so far with AI, there are two applications belonging to Google; Duplex AI and AlphaGO. Duplex AI is an NLP-based virtual assistance and booking application that is expected to be integrated into Google Assistant. What distinguishes Duplex AI from other virtual assistants is its partial autonomy that allows the application to go beyond finding the opening hours, addresses, or phone numbers and manage conversations with staff to make appointments and reservations. In addition, Duplex AI can mimic the sounds created by a human larynx including pauses and gap-filling sounds by just analyzing the sound bites fed to the learning module. This makes Duplex sound considerably more authentic than other virtual assistants on the market. AlphaGO, on the other hand, is the product of a more experimental effort and is therefore less market or service-oriented than Duplex. GO is a board game rooted in Ancient China. Playing GO requires a sharp mind, considerable amount of strategic thinking, and players’ full attention. Much like chess, mastering the game usually takes years. Through playing itself, AlphaGO came to beat the worldwide champions of the game in just a year. However impressive, these examples are far from the robots that will take over the world. So why are our discussions of AI concerned more with robot armies, sentient machines, and all that doom and gloom than what’s on the table? The immediate answer is the drive inside all of us that carried us forward through ages; our imagination. Other answer include popular culture, expectations carved and shaped by history, and popular culture. Throughout this series, we will devote special attention to each of these factors and answers.
The Principle of Duality and The Food Chain
From Asimov’s infamous I, Robot to the Droids in the Star Wars franchise or to later examples such as Her and Ex Machina or HBO’s Westworld, the big screen loves to draw a grandiose and futuristic image when it comes to AI. Few exceptions aside, the cinema also tends to merge the discussions about AI and robotics, in other words, it tends to give a body to all those minds and voices we are set to create. Two cognitive factors in particular can be claimed to run the machinery of this logic. The duality of existence or self is an idea that has its roots in ancient philosophy. According to this view, self is a two part entity consisting of a body and a soul. To put it more plainly, the body is seen as a vehicle that reflects the willing and workings of the soul. Souls that are powerful enough to not need a vehicle, such as the souls of the ancestors in many animistic belief systems, angels and demons, gods, and ghosts deliver their messages through shivers and whispers. Their non-material existence works to their advantage and gives them to invade without raising alarm, possess without being seen. In the 1992 film The Lawnmower Man, Job who has been trained basically as an AI module in a laboratory transcends the limits of body and tells the Professor that he will be everywhere, every phone ring will be his birth cry, and that he will run through the cable veins of the world. This means that assigning metal bodies to these minds made of bytes aims to limit them and draw lines for them. This fear of penetration is materialized as the unseen actors behind our screens started to collect and analyze our data to profile us not just as consumers but also citizens and voters. This correlation shows although our discussions of AI are mostly fiction-inspired and happen within a semi-real sphere, the core concerns find themselves a place in reality. Another reason for our tendency to dress AI with metal is that we attribute a supernatural quality or divinity to non-bodily existence. Therefore our desire to include robotics in the discussion of AI, or our desire to embody AI if you will, is an attempt towards making this technology that we were too hasty to personify on par with ourselves. We are afraid of losing dethroned and thrown down from our den at the top of the food chain. In short, we both want to enhance our reach and increase the capabilities of AI for our own good but at the same time are afraid of creating our own masters. This worry reveals itself in our rush to balance and control the advance of AI. According to the experts, there is a long way ahead before we consider AI a rival. Scientists speculate that GAI or generalized artificial intelligence; a module that can perform a wide variety of tasks, recognize correlations and contexts and make deductions on a human level, is at least half a century away. SAI or super artificial intelligence; a module that can surpass humans in all humanly activities such as learning, performing tasks, and scientific discovery is a dream the odds for whose realization are slim, optimistically speaking. Lawmakers and regulators are already hard at work to create a balance and although some companies started to chip their employees or produce Crispr kits for commercial use, these stand as rare eccentrics within the sector and are expected to remain on the sidelines. However, within a cultural context, it is clear that our hurry to see this technology as a rival and strive towards a nuanced, reserved approach can be read as a lack of trust for the ones who are in the wheel of this.
So although the general opinion on AI is shaped largely by mutually inspirational expectations, fears, hopes, and popular culture, some of our historically embedded concerns correspond to reality and have the potential to guide the regulators and policymakers as well as to generate a healthy public discussion. When it comes to the reality of AI, like every other scientific field, it is moving slowly on its course, there are decades between big leaps and both our imagination and our survival reflexes seem up to speed.
In the second installment of this series, we will discuss the more material concerns created by AI such as unemployment or dependence on cold calculations for crucial decisions