Table of Contents
Artificial intelligence (AI) has become an integral part of our lives, with its applications ranging from virtual assistants to autonomous vehicles. However, as AI continues to advance, so does public concern about its ethical and social implications. While some may argue that it is time to slow down and address these concerns, tech companies seem to be moving in the opposite direction. The race to develop AI technology is heating up, and there is a sense of urgency to move fast and break things, as famously quoted by Mark Zuckerberg.
n
The Concept of Technical Debt
n
In software development, the term “technical debt” refers to the cost incurred in the future as a result of choosing faster, less careful solutions in the present. When developers rush to release software that is not fully ready, they often rely on fixing bugs and issues after it hits the market. This approach can lead to technical debt, as the need for future fixes becomes apparent.
n
However, the concerns surrounding AI technology go beyond mere software glitches. The negative news stories surrounding generative AI often revolve around the amplification of harmful biases and stereotypes, deceptive use by individuals, privacy concerns, and the potential replacement of human jobs. These are not issues that can simply be fixed with a software update. They represent a different kind of debt – ethical debt.
n
Ethical Debt: Considering the Consequences
n
Ethical debt arises when developers fail to consider the possible negative consequences or societal harms associated with their technology. Unlike technical debt, ethical debt is not solely incurred by the developers themselves – it affects society as a whole. The repercussions of AI technology reinforcing oppression or bias are far-reaching and impact marginalized groups disproportionately.
n
As a technology ethics educator and researcher, I have dedicated considerable thought to these types of “bugs.” Ethical debt accumulates when technologists fail to consider the ethical implications of their work. To reduce ethical debt, it is crucial to take the time to anticipate potential risks and consider the ways in which technology might go wrong.
n
Rushing to Market: The AI Race
n
The race to develop AI technology has intensified with the release of OpenAI’s ChatGPT in November 2022. Google and Microsoft quickly followed suit with their own generative AI programs, seemingly rushing to keep up with the competition. However, these hasty releases have not been without their consequences.
n
Google’s stock prices took a hit when its chatbot Bard provided an incorrect answer during a demo. Microsoft, despite its previous misstep with the Twitter-based bot Tay, saw its AI-powered Bing raise concerns among users for repeating misinformation. These instances highlight the social debt incurred by rushing AI technology to market.
n
When the full impact of these rushed releases becomes apparent, we can expect to hear about unintended or unanticipated consequences. Ethical guidelines alone cannot predict all possible societal problems that may arise from AI technology. However, it is disingenuous to suggest that technologists cannot speculate on these consequences. Countless examples have shown how AI can reproduce bias and exacerbate social inequities, and it is often external researchers who identify these harms.
n
Speculating Responsibly: Thinking Like Science Fiction Writers
n
To address the ethical implications of AI technology, designers need to learn to think more like science fiction writers. Isaac Asimov once said that science fiction authors “foresee the inevitable, and although problems and catastrophes may be inevitable, solutions are not.” The responsibility of developing solutions to mitigate these risks falls on technologists themselves.
n
Encouraging ethical speculation is a crucial step in this process. By teaching students to imagine the potential negative consequences of technology, we can prepare them to consider the ethical implications of their own work. One teaching exercise, the Black Mirror Writers Room, allows students to speculate on the negative consequences of technology like social media algorithms and self-driving cars. This approach fosters a skill that can be applied immediately and encourages students to envision harm to others, particularly marginalized groups disproportionately affected by technological harms.
n
Slowing Down the AI Race
n
In March 2023, an open letter called for a pause in training AI systems more powerful than GPT-4. The authors of the letter voiced concerns about the potential risks associated with unchecked AI development, such as the loss of control over civilization. While some critics argue that the focus on hypothetical risks ignores current harms, there is a consensus among AI ethicists that the development of AI technology needs to slow down.
n
Addressing ethical considerations cannot be dismissed as an afterthought. Developers cannot simply cite “unintended consequences” as an excuse for neglecting ethical implications. The debt incurred by rushing AI technology will eventually come due, and history has shown that Big Tech executives and investors may not be the ones paying for it.
n
Conclusion
n
The ethical implications of AI technology cannot be ignored or brushed aside. As the race to develop AI intensifies, the focus on speed and competition has led to the accumulation of ethical debt. To navigate the ethical challenges posed by AI, technologists must anticipate and address the potential negative consequences of their work. By thinking like science fiction writers and fostering ethical speculation, we can prepare future technologists to consider the ethical implications of their work and avoid the pitfalls of rushing AI technology to market. Slowing down the AI race is essential to ensure that the benefits of AI are realized without sacrificing ethical considerations.
