Inclusivity, Values to Shape Future Scenarios When AI Outwits Humans
Originally published in The Yuan on June 27, 2023, as part of its "quest of intelligence discovery" series. Republished here by the author. Read the archived original.
Editor's note, The Yuan
Day 7. On day seven of its quest of intelligence discovery, The Yuan sets course for the Gulf of Guinea and the delta of the River Niger - called by the nomadic Tuaregs of its northern reaches Egereou N-Igereouen (River Among Rivers) - to the nation of Nigeria that bears its name, which suffered grievously from the slave trade that accompanied the Age of Exploration, where ML engineer Gift Kenneth argues that, only if humanity can devise AI systems that are inclusive and align with its shared values, just as the Niger draws many streams into one mighty flood, may this avert the AI apocalypse.
Shifeng Wang, Chief Editor, The Yuan
As technological advances in the age of artificial intelligence (AI) continue, with ever more integrations to spice up regular applications, and improved access to AI products, to hold a thorough discussion of the future of humanity and the role AI must play, both now and in the long run, is essential.
Throughout history, humans have had a penchant for attempting to predict the future, despite their inherent fallibility in doing so. This can be seen in attempted predictions about natural disasters, economic crises, scientific discoveries, and technological disruptions. Such forecasts are wrong more often than not, but that does not stop them from trying.
This article will discuss some of the possible scenarios that can be expected as AI continues to expand and grow.
Future scenarios
When exploring future scenarios of what might happen when AI outsmarts humans, several come to mind. Three of the most common are:
AI proves its destructive potential: This is one of the most famous portrayals of the future of AI and humanity - especially in science fiction, which has highlighted concerns that humanity might be building towards a dystopian future. This depiction is seen often in popular movies such as The Matrix, Ex Machina, and The Terminator. This could possibly happen if AI becomes intelligent in such a way that it sees humans as a species not worth saving from global warming, overpopulation, disease, or famine, or if it decides to wipe humanity out.
Harmonious coexistence: In contrast to a dystopian future, this is a far more positive and utopian outlook where AI is used to enhance human potential and a symbiotic relationship takes root between humans and AI. This sort of relationship is also portrayed in sci-fi, notably in works such as Her and Interstellar. Kai-Fu Lee's book AI 2041 also explores this positive outlook on the future of AI in different cultural contexts.
AI falls into the hands of a select few: Another possibility to consider - one which has been the concern for most current AI policies and strategies - has been the misuse of AI in a way that leads to the consolidation of power by select entities, i.e., the economic elite, large tech companies, and governments. This could result in biased decision-making processes and hidden agendas perpetuated by a few significant players who have outsized influence in large populations.
Another perspective worth pondering is the possibility that humanity could destroy itself before AI has the chance to morph into artificial general intelligence or superintelligence. When exploring these various scenarios, one must still depend on sources such as sci-fi, influential leaders, and the news, all of which are flawed sources of guidance and information. This means that, while all these scenarios are plausible, they are also still highly speculative: there is still no way of knowing which scenario is most likely, or whether a wild card one that no one has yet thought of may come to pass.
AI's moral compass
AI and human learning may differ, but they still share similarities, such as the need to make sense of the world around them, learning by trial and error, and building on previous knowledge and experiences.
A child's moral compass is determined by his or her environment and the amount of information and training s/he has been exposed to. Like AI, a child's moral compass - which may be pivotal in determining the end result of humanity's future with AI - can be heavily influenced by the amount and type of data it has been exposed to.
This also leads to further questions: What data are used? How are data collected? Who is collecting these data? Who will benefit from the data collection? How are these models being built and optimized? Who is building them? There are also questions about humanity's moral compass and ability to instill ethical considerations within AI systems in a way that aligns with different cultural, social, and economic perspectives.
The way these models are built will significantly impact future outcomes - especially in the so-called Global South, since it is critical that these models not be biased against certain people or leave out groups who ought to be represented.
Embracing uncertainty
When contemplating the future of humanity if and when AI outsmarts humans, one can draw inspiration from Nassim Nicholas Taleb's famous book The Black Swan, which explores rare and unforeseen events with significant consequences. The upshot of this is that one must acknowledge that the future is inherently uncertain, and the past is by no means a reliable guide to it.
Reflecting on the COVID-19 pandemic, the world witnessed the power of the capabilities of collective intelligence and technological advancements in healthcare when addressing unprecedented challenges. People can shape the future of AI by establishing more ethics boards, continuously reviewing policies, and ensuring companies abide by them, while also adopting them to address new developments. By asking critical questions about representation, understanding the potential biases within datasets and underlying algorithms, and embracing ethical practices across the entire pipeline, humanity can guide the development of AI systems that align with its shared values. Doing so may be the best way to avert the possibility of a dystopian future, or at least reduce its probability.
Even now, one can already see a growing dependence on AI, as people become more attached to it both consciously and unconsciously for various tasks, from relying on ChatGPT for work to depending on autocorrect to fix misspellings. Such developments should also raise questions about the function of human agency and the potential risks that may arise as a result of surrendering research and essential decision-making entirely to AI systems.
Responsible development
Going back in time or trying to erase technological progress is clearly no solution. AI is here to stay, and as humanity transitions into a future dominated by it, everyone must make a collective effort to nurture it so that it embodies the best aspects of humanity as much as possible. Ethical considerations must be woven into the very fabric of AI by embracing uncertainties, addressing unethical behaviors, asking the right questions, and actively shaping its development. The world can forge a future where AI and humanity can coexist, empowering humans to face whatever challenges lie ahead for them.
This also raises the need for more African governments in particular to address issues like data privacy and AI ethics more proactively. As of the time of writing, only three African countries - Mauritius, Egypt, and Rwanda - have official national strategies, while the other 51 are either still developing theirs or have no coherent strategy at all. In this age of AI, this is an urgent problem that must be addressed soon in order to safeguard the interests and rights of Africans and ensure that their values are also represented in the AI space.
Conclusion
The future remains both unknown and unknowable, but as AI advances and outsmarts humans in more and more domains, it is crucial that humanity navigates this future with informed deliberation and a solid ethical framework. By exploring differing potential scenarios and addressing biases, the world can shape a future where AI is a tool for human empowerment, progress, and coexistence, not something to be feared.
Predicting the impact that AI will have in the distant future in Africa depends on many factors, such as the continent's investments in AI education, technology, and infrastructure, as well as the nature of the policies that will be put forward and updated.
While humans may not know what lies ahead for them as a species, that is no excuse for inaction: if people want to achieve a better tomorrow, they must focus their attention on what can be fixed today and work on that to the best of their abilities.
