As we marvel at the capabilities of artificial intelligence, it’s amazing to think that technology has a long way to go to reach its full potential. With the positive benefits that AI and intelligent automation offer, there should also be a healthy acceptance of the probability that things could go wrong. I don’t mean to sound pessimistic, but excitement should not distort the potential risks that may come with the mass adoption of AI.
The advantages and disadvantages of artificial intelligence exist in equal measure, and in some circles, there is an acknowledgment of the potential disadvantages of introducing AI into certain areas of life. In this article, some of the potential disadvantages of artificial intelligence have been identified, explored, and addressed.
It should be noted that any of the disadvantages of artificial intelligence are not derived from technology, but rather from the way the technology is used. An example is how entertainment recommendation engines offer specific content types to a consumer to the extent that the consumer adopts a biased narrative. The problem here exists in that prolonged exposure to biased narratives can create polarization, radicalization or extremism.
Some Well Known Disadvantages of Artificial Intelligence
Deepfake and Identity theft
Whilst exploring the pros and cons of artificial intelligence, one notable innovation gaining popularity is deep fakes. AI deep fakes have been used for fun, inoffensive apps as well as malicious imitations of celebrities and politicians. Deep fake is a technique that uses artificial intelligence to swap faces in videos. The deep part of the word comes from deep learning, a subset of AI that uses neural networks.
Deep fake AI algorithms use audio or images of a person to recreate the person’s appearance, expressions, and voice. This innovation has the potential to take identity theft to another level.
Deepfake technology was invented in 2014 by Ian Goodfellow and works by using generative adversarial networks (GANs) algorithms not only to classify images but also to generate and create images. This is achieved when two GANs attempt to deceive each other that an image is real. By using just one image, a GAN can create a video clip from that image.
The ease with which a deep fake can be created raises concerns, especially in cases where representations of high-profile political figures can be used for misinformation or as a means to sway public opinion. This implies that as AI continues to evolve, there is a high probability that the technology may end up in the wrong hands or be applied without moral regulation.
Unethical use of Artificial Intelligence
A case comparable to these concerns highlighted is the Cambridge Analytica scandal. In the case of Cambridge Analytica, the political consulting company purchased data sets containing user profiles of 50 million Facebook users.
In early 2014, the company used personal information taken without authorization to build a system that could profile individual US voters in order to target them with personalized political advertisements. This strategy was used during the 2016 US election for Donald Trump’s presidential campaign.
The data sets obtained from Facebook were parsed into software programs and then used to predict and influence the choices of voters at the ballot boxes. This use case raises questions about the moral and ethical uses of AI.
The fallout of the Cambridge Analytica scandal resulted in public scrutiny, and the company was later banned from advertising on Facebook’s platform. Cambridge Analytica and its parent company eventually filed for insolvency and ceased operations in 2018. This case proves that as we move further into an AI ecosystem, we need regulations to safeguard humans from being violated by negative actors and their misuse of technology.
Youhua Bengio, an expert spearheading the growth of artificial intelligence and colleague of professor Geoffrey Hinton of Google, pointed out that his concerns are not with AI taking over humanity. His premise is based on his idea that if humans are able to build machines as smart as humans, then these machines will be smart enough to understand human values, which in turn will cause these machines to make decisions that will be beneficial for humanity.
Professor Bengio, also known as one of the Godfathers of AI, believes the real concern relates more to the use of AI to influence people’s minds, especially through political advertising, as in the case of the Facebook and Cambridge Analytica scandals.
As AI looks on course to influence many areas of our lives, Bengio believes a way to combat any perceived disadvantages of artificial intelligence is to introduce regulation that restricts its use in areas that are morally and ethically wrong.
The UK has started taking steps to address the need for the regulation of artificial intelligence. A select committee was appointed by the House of Lords in 2017 to produce a report titled: AI in the UK: ready, willing, and able? The objective of the report is “to consider the economic, ethical and social implications of advances in artificial intelligence.”
AI and the point of no return
Humanity has gone beyond the point of no return when it comes to machines automating the way we live, and AI has taken intelligent automation to a new dimension. What will happen when robots begin to replace humans in areas of society where humans find meaning, purpose, and sustenance for livelihood?
Some businesses have already started creating factories with semi-automated robots to pick and pack products. These robots are remotely controlled by human operators. Whilst being controlled, the AI is watching and learning how tasks are performed with the goal of one day becoming fully autonomous.
This form of automation and machine learning is predicted to eventually replace humans, and not only with regard to manual labor. It’s also threatening the jobs of white-collar workers too. David Autor, an MIT economics professor, pointed out that middle management jobs also face the threat of being absorbed by intelligent automation, especially in fields where humans are required to translate streams of data into solid business decisions.
Humans have always had a suspicion of automation and have revolted against it in the past. Machines taking away jobs is a sensitive topic, and the 1800s uprising of the Luddites movement in England is a fine example. In this case, the Luddites (factory workers) revolted against machines replacing them.
The movement became increasingly violent, resorting to the breaking of machines, the assassination of William Horsfall, a factory owner, and the attempted assassination of another factory owner.
The rebellion occurred after attempts to negotiate wages and benefits for lack of work were ignored by the merchant factory and business owners. The Luddites were not so much against automation but rather held the belief that the profits from increased efficiency and productivity should not just benefit the owners of the machines. This debate didn’t end well for the Luddites, and the British government eventually crushed the movement. Dozens of Luddites were executed, and many more were exiled to Australia.
This suspicion of automation that has carried across centuries is relevant now, as AI, machine learning, intelligent automation, and robotics begin to replace the human workforce. It should be noted that, although automation eliminated jobs in previous cases, it also led to the creation of new jobs, although this outcome is only realized with hindsight.
Furthermore, the permutations of automation in past decades sought to replace manual and repetitive tasks, but this new generation of automation comes with a form of intelligence designed to stimulate the mind of a human as a minimum goal. Long-term AI seeks to exceed the capabilities of humans to reach a state known as singularity.
Commentators have mixed responses. Some highlight the fact that new jobs may be created as a result of this automation. Others believe that AI will give people freedom. But freedom at what cost? And freedom from what to do what? Will this freedom mean a future with only a few jobs for humans to perform?
Humans attach a sense of purpose and identity to their work. What will happen if people are deprived of this work? If the opportunity is taken away, how then will people make a living or have the opportunity to climb from poverty to wealth or even just remain content? One proposed answer to this concern is universal basic income (UBI), which Elon Musk and Mark Zuckerberg have been publicly vocal in supporting.
The Rise of AI and UBI
The increased adoption of AI and intelligent automation in the labor market should trigger governments, policymakers, and businesses to start exploring how people will make a living in the future. The reality of AI transitioning from science fiction to science fact will impact many industry sectors across a broad range of skills and expertise.
More importantly, if categories of job types that have in the past helped people to climb the financial ladder and build financial stability become obsolete, what impact will that have on society as a whole?
Disruptive technologies and disruptive innovation, as the word infers, will certainly cause disruption. However, I believe that AI and automation should augment human capabilities rather than relegate humans to becoming bystanders and observers.
Ok, let’s say that increased productivity at a higher velocity and reduced cost is a way of augmenting human capabilities. If this is so, then will the advantages of AI be outweighed by the disadvantages of artificial intelligence because humanity has generations of idle unfilled citizens?
Many humans need a sense of purpose to be fulfilled. This sense of purpose is sometimes found in work. Elon Musk shares this sentiment and echoed a similar point when speaking at the World Government Summit in Dubai.
Musk went on to say that “there will be fewer and fewer jobs that a robot cannot do better” adding that the output of business will be really high and “with automation will come abundance.” If the basic need for humans to realize a sense of purpose and meaning is taken away by automation, then “how will people derive meaning?” Furthermore, how will this impact humanity’s collective morale?
Challenges with the unknowns of AI
Some of the challenges revolve around the fact that AI is taking us into uncharted waters. This is analogous to early voyagers and explorers who sailed the oceans and seas in search of a new world and didn’t know what to expect.
Today we have tools that we can use to try to predict the effect of artificial intelligence and automation on our future. But the true impact of intelligent automation as a result of the rise of AI will only be realized after the fact. We can only make efforts to prepare.
AI is undoubtedly augmenting human abilities in ways that allow us to go faster, act smarter, and produce more for less. We, therefore, need to be considerate, thoughtful, and responsible with the power that is being bestowed upon us with AI and intelligent automation, especially because with all the benefits, there are also potential hazards that need to be navigated or managed in perpetuity.
We are stepping into the realm of the unknown and into a world where machines and their ability to learn, respond and decide are being woven into the fabric of everyday life. We should not overlook the fact that with the advantages come the potential disadvantages of artificial intelligence mass adoption in our world. With the pros of AI come the cons of AI.
Like two objects on opposite ends of a weight scale, advantages should be weighed against disadvantages, and pros should be weighed against cons. The outcomes should influence decision-making whilst providing a framework for the regulation of artificial intelligence.
The extent to which AI will change our world is not fully clear, especially because in the grand scheme of things, the potential AI offers mankind is in the early phase of discovery. Exciting times or scary times? The jury is still out. Either way, the scientific fact is that the increased adoption of AI will definitely be eventful.