An Argument Against AI



1” While ideologies themselves could be considered beliefs and ideals that could be one way or the other, it is important to understand the system of ideologies at work behind accepting AI. Technology and science were not mainstreamed until the last two centuries. Post-enlightenment, “rationalism, or a belief that we come to knowledge through the use of logic, and thus independently of sensory experience,2” science is widely revered. Major fields like Western medicine, armed forces and the law depend on it.

“The medium is the message”, meaning that the form of communication shapes the medium and influences the message being communicated. McLuhan also believed that technology shapes us as much as we shape it, as it extends our abilities and influences the way we think and behave.3” The classic argument is that the designed object designs us in return. The problem might not be “technology” as much as the model it works on. Like most technology that can be considered an extension of our human capabilities,4 AI is an extension of human intelligence, i.e. the brain. This plays into anxieties that fear science fiction coming true. “The idea is that a truly intelligent AI would also develop consciousness, and with consciousness comes the capacity for emotions, agency, and all other aspects that we associate with subjects that deserve moral consideration5” However, it might be worthwhile to consider alternate models to AI as well. “While AI is being seen as a replacement for human intelligence, the emergence of IA offers a unique opportunity to augment human intelligence with machine learning. Yet, despite the immense potential of IA, there are still concerns about its potential impact on the workforce.6”

While we cannot say with conviction the IA model is better than AI, the difference between these two models reveals the nature of our current ideology. We may believe one technology could do better than the other because of our existing beliefs (perhaps in consumerism, empiricism, rationalism etc.) However, the AI model “is a computer system that performs tasks that usually need human intelligence. AI systems can learn, reason, and perceive. They automate tasks, such as data analysis and decision-making.7” And on the other hand, “IA is a human-centric approach to artificial intelligence. IA focuses on enhancing human intelligence by integrating technology with human intelligence. IA enables humans to make decisions based on data and insights provided by AI.8”

Another point of reflection must be our current faith in AI to get “better” with time. ChatGPT cannot still replace human creativity. In fact, there are emerging roles like “prompt engineers” who work on engineering pt prompts to generate interesting results. While we wait for dooms day and for AI taking our jobs, we are missing a far more concerning aspect. Tech leaders today have enough faith in AI despite there being several proofs that AI can be tricked. Corporations have already put AI into practice. We see job applications rejected by machines due to faulty assumptions and encoded bias. In fact, “Another interesting work, titled “Accessorize to a Crime: Real and Stealthy Attacks on State-of-the-Art Face Recognition”, showed that one could fool facial recognition software by constructing adversarial glasses by dodging face detection altogether. These glasses could let you impersonate someone else as well9”

Ultimately, a world free from AI need not be saved from it. There is nothing inherently wrong with technological advances in a purely academic sense. However, the power structures of our current society and how science is exploited to benefit some while oppressing others is the problem. Just like how the advancement in nuclear technology threatens to end the world, AI has left a resounding impact in our society, and it has not been fully developed yet. Companies like Microsoft have introduced Co-Pilot,  “Copilot combines the power of large language models (LLMs) with your data in the Microsoft Graph—your calendar, emails, chats, documents, meetings, and more—and the Microsoft 365 apps to turn your words into the most powerful productivity tool on the planet. And it does so within our existing commitments to data security and privacy in the enterprise.10” In essence, Co-Pilot has been used to summarise meetings by collecting all kinds of data like recordings, chats, emails etc. What about the inherent disparities glaringly visible during a meeting? Who does all the speaking? Is it always the person who knows the most or is responsible for all the research? Are other groups well represented? These are not questions that concern the machine because these are not questions that concern us yet. Even if they are, it is not profitable. And even if this question was to become profitable to answer, there may be others that may not. Our machines are just an imitation of our capabilities. In such a scenario, Artificial Intelligence remains to be just that…artificial.


References

1  Jaron Lanier and E. Glen Weyl, “AI Is an Ideology, Not a Technology,” Wired, March 15, 2020, https://www.wired.com/story/opinion-ai-is-an-ideology-not-a-technology/.
2 “Rationalism | History of Western Civilization II,” Lumenlearning.com, 2019, https://courses.lumenlearning.com/suny-hccc-worldhistory2/chapter/rationalism/.
3 OpenAI. "ChatGPT." Accessed April 9, 2023.
4 sytaffel, “Technology and Agency,” Opentextbc.ca (BCcampus, February 28, 2014), https://opentextbc.ca/mediastudies101/chapter/technology-and-agency/.
5  Sander Beckers, “AAAI: An Argument against Artificial Intelligence,” n.d., https://sanderbeckers.github.io/website/sander/articles/AAAI-preprint.pdf.
6Bob Hutchins, “AI vs IA: The Battle for the Future of Work- Can We Collaborate with Machines?,” www.linkedin.com, March 7, 2023, https://www.linkedin.com/pulse/ai-vs-ia-battle-future-work-can-we-collaborate-bob-hutchins-msc/.
7 ED BURNS, “What Is Artificial Intelligence (AI)?,” TechTarget (TechTarget, February 2022), .
8 Bob Hutchins, “AI vs IA: The Battle for the Future of Work- Can We Collaborate with Machines?,” www.linkedin.com, March 7, 2023, https://www.linkedin.com/pulse/ai-vs-ia-battle-future-work-can-we-collaborate-bob-hutchins-msc/.
9 Anant Jain, “Breaking Neural Networks with Adversarial Attacks,” Medium (Towards Data Science, February 9, 2019), https://towardsdatascience.com/breaking-neural-networks-with-adversarial-attacks-f4290a9a45aa.
10 Colette Stallbaumer, “Introducing Microsoft 365 Copilot,” Microsoft 365 Blog, March 16, 2023, https://www.microsoft.com/en-us/microsoft-365/blog/2023/03/16/introducing-microsoft-365-copilot-a-whole-new-way-to-work/.