This is the first in a series of six posts on compliance issues with various online platforms. Today’s post is about YouTube. Next Tuesday’s post will be about Facebook. The third post in the series, on March 22, will discuss Instagram. The fourth post in the series, on March 29, will focus on Twitter. The fifth post, on April 5, will be about Snapchat. On April 12, the sixth and final post in the series will discuss Reddit.
The video hosting and sharing service YouTube was created in 2005 and is now owned by Google. YouTube contains content from both individuals as well as media corporation partners. This content is extremely diverse, ranging from short clips to entire television shows and films as well as music videos, video blogs, live streams, and educational presentations. YouTube also makes use of the advertising program Google AdSense and includes targeted ads on its content; most of the videos on YouTube are free to view but some ad content will appear before, during, and/or after the video plays.
Black Mirror is a very popular US-UK television science fiction series. It originally aired on Channel 4 in the UK and is now released and broadcasted by the subscription video streaming service Netflix. The series is anthology-style, with short seasons of stand-alone episodes which are like mini films. Most of the episodes of the series touch upon the dominance of and overreach into human life by technology, such as social media, AI, and other advanced, immersive systems and devices. The take offered is quite dramatic, often delving deeply into adverse psychological and sociological effects on modern society, taking a dark and even dystopian perspective.
While all the episodes of Black Mirror do depict a future reality, it is an immediate and accessible reality impacted by technology exceeding that which is currently possible but not so much as to be unthinkable. Indeed, the title of the show, Black Mirror, refers to current technology which is increasingly ubiquitous and addictive – television screens, computer monitors, and smartphone displays. The show both entices with the idea that many of these technological advancements could be convenient or novel or life-enhancing, while also warning that the obsessive and addictive aspects of technology could cause great harm and disruption if not developed and managed thoughtfully and carefully with the risks well in mind.
“The Entire History of You” (Series 1, Episode 3): In this episode, a couple struggling with mistrust and insinuations of infidelity make disastrous use of a common biometric – a “grain” implant everyone has that records everything they see, hear, and do. The recordings on the implants can be replayed via “re-dos.” This is used for surveillance purposes by security and management, as the memories can be played to an external video monitor for third parties to watch. Individuals can also watch the re-dos from their implants directly in their eyes, which allows them to repeatedly watch re-dos, often leading them to question and analyse the sincerity and credibility of people with whom they interact. People can also erase the records from their implants, altering the truthfulness of the recordings. This troubles the status of trust and honesty in society which has already in contemporary life been eroded by the influence of the internet.
“Be Right Back” (Series 2, Episode 1): In this episode, Martha is mourning her boyfriend, Ash, who died in a car accident. As she struggles to deal with his loss, her friend who has lso lost a partner recommends an online service that allows people to stay in touch with dead loved ones. The service crawls the departed person’s e-mail and social media profiles to create a virtual version of the person. After the machine learning advances enough by consuming and trying enough communications, it can also digest videos and photos by graduating from chatting via instant message to replicating the deceased’s voice and talking on the phone. At its most advanced, the service even allows a user to create an android version of the deceased that resembles him or her in every physical aspect and imitates the elements of the dead person’s personality that can be discovered by the online record. However, in all this there is no consideration given to the data privacy of the deceased person or to his or her consent to be exposed to machine learning and replicated in this manner, including even the physical android form.
“Nosedive” (Series 3, Episode 1): This is one of the most popular, critically-acclaimed episodes of the series, and one of the obvious reasons for this is that it focuses on social media and how it impacts friendships and interactions. The addictive aspects of social media in current times are already a hot topic in design ethics, driving people to question whether social media networks like Facebook or Twitter are good for the people who use them, and where to locate the line between entertainment and a fun way to connect and share, versus a platform with a potentially dark and abusive impact on users. In this episode, everyone is on social media and is subject to receiving ratings from virtually everyone they encounter. These ratings determine people’s standing both on social media and in the real world as well – controlling access to jobs, customer service, housing, and much more. Anxieties and aspirations about ratings drive everything people do and all the choices they make. “Addictive” has been met and surpassed, with social media having an absolutely pervasive impact in everyone’s lives.
“San Junipero” (Series 3, Episode 4): One of the most universally loved episodes of Black Mirror, San Junipero depicts the titular beach town which mysteriously appears to shift in time throughout the decades. Kelly and Yorkie both visit the town and have a romance. San Junipero turns out to be a simulated reality which exists only “on the cloud,” where people who are at the end of their lives or who have already died can visit to live in their prime again, forever if they so choose. In the real world, Kelly is elderly and in hospice care, while Yorkie is a comatose quadriplegic. Both eventually chose to be euthanized and uploaded to San Junipero to be together forever, after getting married first so that Kelly can give legal authorization to Yorkie to pass over. The bioethical considerations of such a reality are clear – in this society, assisted suicide is a legal normalcy, and part of patient care is planning one’s method of death and treatment path after death, which digitalization being a real option. All of the San Junipero simulations exist on huge servers, and judging by how many lights are flickering in the racks this seems to be a popular practice – but what about cybersecurity and information security of the simulations? What if the servers were hacked or damaged? This gives a new meaning to humanity and places an entirely different type of pressure on making sure that technology is used safely and the data stored on it is protected.
“Men Against Fire” (Series 3, Episode 5): This episode concerns the future of warfare in a post-apocalyptic world. Soldiers all have a biometric implant called MASS that augments reality, enhances their senses, and provides virtual reality experiences. One soldier’s implant begins to malfunction and he soon learns that the MASS is in fact altering his senses so that he will not see individuals he is told are enemy combatants as people. It turns out that the soldier is part of a eugenics program practicing worldwide genocide and the MASS is being used to deceive the solders and turn them into autonomous weapons who murder on command due to the augmentations and alterations to reality by the MASS. This storyline falls cannily close to many current concerns about the adoption of autonomous weapons that are not directed or monitored by humans, which are nearly within technological capability to be created and are the subject of international calls for appropriate supervision of and restraint in their development.
Black Mirror offers many interesting scenarios for analysis of and study by compliance and ethics professionals considering risk management related to the use of technology in organizations and society. As described above, surveillance, data privacy, consent, design ethics, autonomous weapons and other AI, bioethics, and cybersecurity are just a sampling of the issues invoked by episodes of the series.
Artificial intelligence (AI) describes the cognitive function of machines through technology such as algorithms or other machine learning mechanisms. The very definition of AI places technological devices with this “artificial” knowledge in comparison to and opposition with humans possessing “natural” knowledge. This discipline within technology has been around for more than sixty years and in recent years, is gaining consistent enough momentum that many of its once outlandish ambitions – such as self-driving cars, for example – are current or imminent reality. As computing power advances exponentially and uses for and types of data are ever-growing, AI is becoming ubiquitous in the news of the newest and emerging technological innovations.
As AI sustains and draws on its now considerable basis of achievements to make even more advancements in research and development across many business sectors, ethical and existential dilemmas related to it become more prevalent as well. Returning to that initial dichotomy between artificial or machine intelligence and natural or human intelligence, the design ethics and morality of bestowing human-like thinking ability on devices and networks raise many philosophical questions. Certain uses of AI, such as for autonomous weapons, could even pose safety risks to humans if not developed and directed thoughtfully.
These questions can go on and on; practical ethics represents the attempt to navigate the broad social context of the workplace by reconciling professional rules with moral expectations and norms. This, again, is highly pertinent to a corporate compliance program, which seeks to encourage an business culture that respects legality, approaches business competitively yet thoughtfully, and also sets standards for employee and organizational integrity. It is imperative for compliance professionals to understand practical ethics and use dilemma sessions or open discussions with the businesses they advise in order to encourage a common comfort level with this sort of thinking throughout their organization.
The below TED/TEDx talks emphasize the connection between AI and human life, commonly invoking questions about bioethics, practical ethics, and morality.
Artificial intelligence: dream or nightmare? (Stefan Wess) – Stefan Wess, a computer scientist and entrepreneur, provides a helpful primer on the history and current state of artificial intelligence in the contemporary movement of machine education. Big Data, the Internet of Things, machine learning, speech recognition – all these technologies and AI-related topics are already part of daily life. But as this continues to develop, how will organizations and individuals interact with the technology? How should it best be controlled and is it even possible to do so? The many risk implications of AI must be considered as more advanced creations become stronger and closer to reality every day.
Can we build AI without losing control over it? (Sam Harris) – Neuroscientist and philosopher Sam Harris is well-known for his commentaries on the interaction of science, morality, and society. Advanced AI is no longer just theoretical stuff of science fiction and the very distant future. Superintelligent AI – completely autonomous, superhuman machines, devices, and networks – is very close to reality. Technologists, the organizations in which they work, and the communities for which they create must all be conscientious about the development of these technologies and the assessment of the risks they could pose. Contending with the potential problems that stem from creating this very advanced AI needs to be done now, in anticipation of the technology, not later – when it may no longer be possible to control what has been designed and brought to “life.” Planning, careful control frameworks, and regulatory supervision that balances openly encouraging innovation with soberly considering safety and risk consequences are all necessary to conscientiously embark upon these amazing technological endeavors.
What happens when our computers get smarter than we are? (Nick Bostrom) – In the same vein as the previous talk, one of the consequences of extremely “smart” artificial intelligence is that machine learning could be just as smart as a human being’s knowledge – and then, of course, eventually overtake humans in intelligence. This is alarming because it suggests the potential that humans could introduce their own subservience or obsolescence via machines created to make machines smarter. Again, all participants in developing this technology, including the consumers to whom it is ultimately directed, need to consider their intentions in bestowing machines with thought and balance the various risks carefully. With the ability for independent thought may also come the capacity for judgment. Humans must make an effort to ensure the values of these smart machines are consistent with those of humanity, in order to safeguard the relevance and survival of human knowledge itself for the future.
The wonderful and terrifying implications of computers that can learn (Jeremy Howard) – The concept of deep learning enables humans to teach computers how to learn. Through this technique, computers can transform into vast stores of self-generating knowledge. Many people will likely be very surprised to learn how far along this technology is, empowering machines with abilities and knowledge that some might think is still within the realm of fantasy. Productivity gains in application of machine learning have the potential to be enormous as computers can be trained to invent, identify, and diagnose. Computers can learn through algorithms and their own compounding teaching to do so many tasks that will free humans to test the limits of current inventions and to extend human problem-solving far beyond where it already reaches. This is certain to change the face of human employment – already bots and androids are being used for assisting tasks in diverse fields from human resources recruiting to nursing patient care. Again, the extension of these technologies must be carefully cultivated in order to neutralize the existential threats to human society and life that may be posed by unchecked autonomy of machines and artificial learning. The time to do this is now, as soon as possible – not once the machines already have these advanced capabilities with all the attendant risks.
What will future jobs look like? (Andrew McAfee) – Picking up on the theme of the changing nature of human employment as machines get smarter, Andrew McAfee draws on his academic and intellectual background as an economist to unpack what the impact on the labor market might be. The fear, of course, is that extremely human-like androids will take over the human workforce with their advanced machine intelligence, making humans mostly irrelevant and out of work. The more interesting discussion, however, is not whether androids will take away work from humans but how they may change the kinds of jobs that humans do. Considering and preparing for this reality, and educating both humans and machines accordingly, is imperative to do now.
Check back here in the future for continuing commentary on AI and its impact on human life and society, including technology and the ethics of knowledge acquisition, as well as more insights on specific AI innovations such as self-driving cars and machine learning.