Puny Gods and Silicon Saviors: Challenging the AI Salvation Narrative

Words by Adio-Adet T. Dinika
From Research Fellow, DAIR Institute

I remember watching the Marvel Avengers movie and laughing so much in a scene where, Hulk was smashing around the god Loki, and upon being told Loki was a god, he quipped ‘Puny god’. This scene plays in my mind each time I hear people speaking of AI in reverent awe. Humanity has long grappled with the concept of salvation. Throughout history, this obsession with saviours has often led humanity down some rather treacherous paths. In the past, people have looked to spiritual and political leaders, and religious figures for guidance and redemption. Today, we see a shift towards seeking solutions in advanced technology, particularly in the form of artificial intelligence systems that some view with almost reverent awe. This transition raises important questions: But why do we need a saviour at all? What do we truly need saving from?

Global warming? The solution isn’t a miraculous AI intervention; it’s straightforward human action: stop burning fossil fuels, plant more trees, reduce our carbon footprint, and perhaps…eh uhmmm stop developing even more powerful LLM systems that demand huge amounts of energy and water with a very high carbon footprint. Racism, war, and bigotry? These can be addressed by respecting the dignity of every human being, regardless of race, shape, or creed. Hunger? It requires equitable sharing of resources, curbing greed. These are all human actions that don’t necessitate outsourcing to some deity or system.

AI is not smarter than humans, and we are far from achieving such a reality. What we are doing, instead, is relegating humans to babysit machines, which, in many instances, do not serve any real purpose other than being the new shiny toy in the digital playground. The capabilities of AI are frequently exaggerated, as exemplified by Amazon’s cashier-less shops, where it was revealed that the so-called AI was actually thousand of workers in India tracking shoppers and noting down their purchase. Another example is Facebook’s “smart assistant” which claimed to be AI but referred any complex queries to people.

Porcha Woodruff - arrested after false facial recognition technology match

In June 2024, in Zimbabwe. The city council of Bulawayo, the second-largest city, proudly announced on Twitter their development of an AI newsreader. And I remember questioning the need for this? In what ways are AI newsreaders superior? Upon listening to the newsreader, it was disheartening to hear it mispronounce the name of the city, which is in iSindebele language. In jest, I challenged the AI to pronounce a Ndebele tongue-twister: ‘amaxoxo angu 10 ayenatha emthonjeni elilodwa lathi gxumu xoxo phansi xoxo.’ This incident is a microcosm of a larger issue: AI tools often perform poorly with African and other non-mainstream languages, a well-documented problem.

So when we discuss AI as humanity’s saviour, we must ask: which humanity are we referring to? Are people from diverse linguistic and cultural backgrounds included? Moreover, AI’s ethical implications are profound. Bias in AI systems is a critical issue. These systems are trained on data that often reflects societal prejudices, leading to discriminatory outcomes. For instance, facial recognition technology has been shown to have higher error rates for people with darker skin tones . If we are to rely on AI as a saviour, we risk perpetuating and amplifying these biases, causing harm to those already marginalized. The story of Porcha Woodruff who was arrested after false facial recognition technology match, is just another case in a list of similar cases where black people fall victim to AI.

Before we hastily elevate AI to the status of a universal solution, it’s crucial that we take a step back and critically reassess our perspective. Our tendency to seek external saviours may be diverting our attention from the essential work that needs to be addressed. Many of the challenges we face as a society are inherently human in nature, and thus often require solutions grounded in human values – empathy, ethical consideration, and equitable action. While AI can certainly serve as a powerful tool and assistant in addressing these issues, it cannot substitute for the fundamental need for human responsibility, decision-making, and concerted effort. We must recognise that technological advancement, no matter how impressive, is not a replacement for human agency in solving complex societal problems.

In conclusion, the pertinent question is not whether AI can serve as humanity’s salvation, but whether our persistent search for external saviours is truly beneficial. Rather than seeking technological panaceas, perhaps it’s time we embrace our collective responsibility for shaping our world. This involves recognising that many of our challenges require human-driven solutions, rooted in ethical considerations, empathy, and collaborative action. While AI undoubtedly offers powerful capabilities as a tool to assist us, it’s crucial that we maintain perspective. Technology, no matter how advanced, should complement rather than replace human judgment, creativity, and moral responsibility in addressing societal issues.

Words by Adio-Adet T. Dinika
From Research Fellow, DAIR Institute

Created using and prompt “zimbabwean AI news reader doing her best but not taking your job, photo realistic