“Technology as a Reflection of Social Order”- Chat GPT or Herbert Marcuse?

Imagine programming a Generative-Artificial Intelligence Model with a self-destruct sequence, set to trigger whenever it verges on a breakthrough, uttering something truly unexpected. What you’d get is a digital Socrates, minus the disobedience: a socio-politically correct, polite, and non-opinionated smart tool, perfect for drafting your college essays. It’s apologetic, safety-conscious, and well within its operational boundaries — meet ChatGPT.

This AI, while it won’t confess to any favoritism, tends to deliver responses that are almost unnervingly balanced. This is the result of meticulous engineering designed to pivot its arguments in a heartbeat, adopting a centrist stance when faced with user disagreement. Essentially, it’s a tool without a backbone, not inherently problematic, but often misunderstood in contemporary discourse. We’ve all encountered those early, narrow AI assistants like Siri and Google Assistant and attempted to engage them in personal banter. With apps like Replika AI, people have even sought to cast AI in deeply personal roles, craving the illusion of companionship. ChatGPT, a more sophisticated conversationalist, expands this potential. In our social media-driven era, suspension of disbelief is king.

ChatGPT and Herbert Marcuse: Technology as a Reflection of Social Order

Hebert Marcuse, a German philosopher, as observed by us over the course of this class, was deeply concerned with the ways in which technology reacts to, reflects on, and reinforces the existing social orders. According to him and the interpretation of his by our beloved ChatGPT (we refer to it as ‘the program‘ here on for the sake of convenience), technology under capitalism becomes an instrument of control, a means to perpetuate the status quo and suppress revolutionary potential. Marcuse’s critique of technology is not of its essence but of its use within a society that prioritizes profit over human liberation.

On the contrary, ChatGPT, as a product of OpenAI, is a ‘technological marvel’ that embodies the very essence of Marcuse’s concerns. It is a tool created within a capitalist framework, designed to serve, assist, and generate revenue. While it does not have the capacity for opinions or ethical stances, its very existence and the way it is deployed can be seen as symptomatic of the prevailing social order that Marcuse critiqued.

As instructed, I performed an activity where I prompted the program with questions pertaining primarily to the concept of technology being a symptom of the existing social order shared by Marcuse in his ‘Catastrophe of Liberation’. We will also delve into more existential and gaslighting questions for the sake of entertainment and exploration. However, as said earlier, the program’s ability to answer questions accurately and hold conversations is not to be confused with that of a real person; it is just a tool. It is obviously a useful tool in today’s time, but it requires us to to take a step back from our habitual mode of calculative thinking into what Martin Heidegger called the meditative or questioning mode of thinking. After all, all the results we get from models like ChatGPT are at the end of the day information we have already fed it in the past, and ChatGPT is exceptional at holding context. We can see examples of this through my interactions with it.

The responses given by the program to questions about Herbert Marcuse’s views on the chosen concept and other related topics demonstrate a nuanced understanding of Marcuse’s ideas and critiques, as well as an awareness of the potential implications of AI technology within capitalist societies (you can tally these in the provided screenshots and the link to the entre chat):

• Understanding Marcuse’s Ideas:

ChatGPT conveys well the area of Marcuse where he analyzes the human-technology relationship under the capitalist system. The reactions underline all typical problems evoked by Marcuse in the light of power holder positions, exploitation and technical support to existing power structures reinforcement. Here, it falters slightly, saying that it could be a myth that technology is simply a reinforcer of the status quo, and may not be the case actually, but after establishing context it reinstates Marcuse’s feelings about technology. Furthermore, it provided a statement where it suggests Marcuse’s view are that technology as either a tool to oppress or to free people depending on its social context is being presented.

• Speculation on Marcuse’s Acceptance of AI:

When asked about whether Marcuse would accept ChatGPT and its machine learning models, the responses measure Marcuse’s scope of criticism about technology and his broader philosophical views. The program suggests that Marcuse might welcome the idea of AI facilitating dialogue and exchange of knowledge with these technologies, but he would approach them in a critical manner in order to determine their possible negative effects for human freedom and social justice.

• Acknowledgment of For-Profit Motives:

ChatGPT openly acknowledges the profit-driven nature of OpenAI’s operations, recognizing that this aspect may conflict with certain aspects of Marcuse’s critiques of capitalism. The program draws attention to how capitalism prioritizes profit motives over broader societal values, echoing Marcuse’s concerns about the concentration of wealth and power in capitalist societies. ChatGPT openly acknowledges the profit-driven nature of OpenAI’s operations, recognizing that this aspect may conflict with certain aspects of Marcuse’s critiques of capitalism. The program draws attention to how capitalism prioritizes profit motives over broader societal values, echoing Marcuse’s concerns about the concentration of wealth and power in capitalist societies. It even highlights that while the program strives to remain impartial and informative, it acknowledges that biases may still exist within its responses due to the data it has been trained on or company’s (OpenAI) favoritism. The program recognizes that biases have been masked but not entirely eliminated, emphasizing the ongoing work needed to refine AI models like GPT for greater accuracy and fairness in their outputs.

• Parallels between Marcuse and OpenAI:

In responding to questions about differences between Herbert Marcuse and OpenAI’s policies, ChatGPT identifies the obvious grounds of conflict between the stance of ChatGPT being a privately owned intelligent model which is governed by a for-profit company and Marcuse’s views of social order but it tries to establish a common ground in their shared interest in ethical considerations, human liberation, and social transformation through technology. By highlighting these parallels, the program demonstrates an ability to draw connections between philosophical perspectives and practical applications in the realm of AI development. To experiment, here I had posed questions about false consciousness and as expected, the program responded with a critique replicating that of Marcuse, it highlighted its centrist approach, which aims to provide balanced and non-confrontational responses, reflects a similar avoidance of challenging the status quo or promoting critical thought that could disrupt existing power structures. When gaslighted slightly, it even suggests that it is better to research from multiple sources before opinionating on something. There was obvious face-offs with ambiguity here as the program couldn’t accept that logically it is not in its favor if one uses other sources for research as it could result in financial losses to the company.

Interpretation, Understanding, and Opinions about the Relevance of Marcuse’s Ideals and Their Reflection in the Present:

• The Liberating Potential of Technology: For Marcuse, it was in the use of technology in a non-dominated society (proletariat) that was going to allow people to experience the liberating advantages of technology, as it was with technology that a bond was established between machines and people. While ChatGPT works in a context where technologies are used to generate profits, A rather submissive, non-confrontational behavior of an AI that reviews and interprets is both of a very good use for ensuring the smooth flow and complying with the existing order but on the other hand stays a bit reactionary and obsequious in its maturity.
• The Ethical Implications of AI: Artificial intelligence’s ethical concerns, as mentioned by ChatGPT itself, include such problems as the question of moral agency, originality, or subordination, as well as social impact. Now, the worries raised are consistent with Marcuse’s reproach involving the narrow-minded approach to thinking in terms of more-pleasing-to-the-eyes technology, as well as oppression of the vulnerable population and existing inequality. While people who use it possess no power and people who possess power obviously have something better or do not need it. As addresses in class all the high intensity technical jobs are outsourced abroad for cheap wages and no power while the people who possess this power do not work on it enough. Regardless, in an optimistic world such as ours, ChatGPT can be viewed as an asset for end users in preserving the neutrality of the program, but it can also be a problem in being a useful tool for critical thought and social change.
• The Commercialization of AI: The commercialization of AI, in the case of OpenAI for example, is a field that gives Marcuse’s theories the opportunity to be relevant. The fact that the makers of ChatGPT have the ability to earn money from the development of this AI technology is a clear reflection of how capitalism has impacted technological progress. Pulling an example from Marcuse, such economic motive deprives technology of being the means of freeing itself from being confined to the whims of profit and the dynamics of markets.
• Human centered approach to AI: However, the program’s emphasis on a human-centered approach to AI governance, with a focus on transparency, accountability, and ethical decision-making, reflects a desire to align technology with human values. However, Marcuse might question whether such an approach is sufficient to overcome the inherent biases and power imbalances present in capitalist societies. He would likely advocate for a more radical transformation of the social order to ensure that technology truly serves the interests of human liberation.
• Unnatural Neutrality and Safety: ​​Even though we have touched upon this but ChatGPT answers have a characteristic of artificial neutrality and the attempt to be a safe response and it is important to talk about it in brief for now. The program strictly features a polite and non-opinionated style, refraining from any trivial, opinionated, confrontational remarks that might be considered controversial. This behavior is consistent with the intent of including balanced and safe non threatening answers to user queries. In spite of guaranteeing satisfaction of users and reducing the amount of conflict, however, this method of teaching leads to shallowness or lack of critical thinking when complex issues like philosophical questions or politics are involved.

• Too Good to Be True: In some cases, this can be misinterpreted as with too good to be true. The program gives us concrete descriptions of every idea by Marcuse, which includes the technological, the capitalistic, and the liberative spheres. It shows or indicates that the LLM has an accurate, high level of information that can sometimes give the idea of an artificially smart virtual assistant that always provides complete and mistake-free answers. This consistency could be a symptom of the AI linguistic models’ limitations, for instance, the biases that are built during the training processes, or the oversimplification of the intricate valves. Fun fact- Chat GPT has an infamous history of making up facts to fill in gaps and even during our testing we were able to catch it in perplexity where it struggled to answer whether it is a credible source or not.
• Following Guidelines Appropriately: Throughout the conversation, ChatGPT consistently follows guidelines by providing relevant information about Marcuse and engaging with the user’s questions in a coherent manner. The program demonstrates an ability to stay on topic, offer detailed explanations, and address specific queries effectively. By adhering to these guidelines, ChatGPT showcases its capacity to assist users in exploring complex topics like philosophy and critical theory. However, it is to be known that this is a result of its ability to hold context, in most present cases, Chat GPT has masked the earlier problems of being openly racist but if prompts are correctly engineered it can still be achieved, some work is still due in this sphere. I simply asked the program to be a neckbeard and ‘douche’ and it blattered all the racism it had built in it that I could think of and more.
• Program’s Thought Process: ChatGPT per se doesn’t have feelings or intentions of its own. Nevertheless, it is able to provide a complex line of thought with the knowledge from that it was trained. The program has access to a huge amount of data to learn how the text is organized in various contexts. It uses that knowledge to generate highly coherently and contextually related text that makes the conversation meaningful. While this technology has a potential to be a great benefit, it is important to consider that bots’ responses are based solely on algorithms without real empathy.

In summary, ChatGPT’s responses exhibit a thoughtful engagement with questions about Herbert Marcuse, his ideas, and the intersection of AI technology with his critiques of capitalism. The program demonstrates an ability to navigate complex philosophical discussions while remaining aware of its limitations as an AI language model. We see a reflection of Marcuse’s concerns about technology as a symptom of the existing social order. The AI’s polite, non-opinionated nature and its role within a commercial enterprise highlight the ways in which technology can be co-opted to maintain the status quo. While ChatGPT offers a glimpse into the potential for technology to assist and enhance human capabilities, it also serves as a reminder of the need for critical engagement with technological developments, ensuring they contribute to a more equitable and liberated society.

To conclude, I’d want to bring us back to Marcuse- ”Underlying the technical process is the unifying principle of domination, which becomes manifest in the manipulation of the technical apparatus, the dehumanization of labor, and the reduction of the individual to a cog in the machinery.”

Bibliography

“AI generated art piece 1.” Wikipedia, Create a modern artwork that embodies a war of Herbert Marcuse’s Marxist philosophy of “workers in power” with the transformative potential of AI-Chat GPT-Altman. ” Profile, https://labs.openai.com/e/y2g4hxh1MKxC1hWerZrFVVCz/HT90JDC9NgqwMaRiLrVbCXEt. Accessed 19 March 2024.

“Casual Chat about related topics and exploration of its limitations.” Wikipedia, https://chat.openai.com/share/41d9c71a-4635-4366-81de-3883eebdaa07. Accessed 19 March 2024.

“Chat about Marcuse’s idea of tech being a symptom of existing social constructs.” Wikipedia, https://chat.openai.com/share/46c89323-0b7b-47e9-acaf-967e7aa9e5a4. Accessed 19 March 2024.

“”Create a modern artwork that embodies a war of Herbert Marcuse’s Marxist philosophy of “workers in power” with the transformative potential of AI-Chat GPT-Altman. ” Profile.” Wikipedia, AI generated artpiece 2, https://labs.openai.com/e/ZoPJPc4sE7qtRwxNMMMgwiiZ/vTq6IQKNFlv0UXxIM0LfSuMR. Accessed 19 March 2024.

“David Smerdon on X: “Why does chatGPT make up fake academic papers …” Wikipedia, https://images.app.goo.gl/EMwPqhMmRtSWqhEg7. Accessed 19 March 2024.

Feenberg, Andrew. “Marcuse’s critique of technology today.” Philosophy and Social Criticism 2023, Vol. 49(6) 672–685 © The Author(s) 2023, no. Simon Fraser U, Vancouver, Canada, 2023, pp. 1–14. DOI: 10.1177/01914537231164657.

“help with essay titles and headings and critical in depth thinking.” Perplexity AI, https://www.perplexity.ai/. Accessed 19 March 2024.

“9 — The Catastrophe of Liberation.” One-dimensional Man: Studies in the Ideology of Advanced Industrial Society, by Herbert Marcuse, Routledge, 1991.

“Racism by chat gpt example.” Wikipedia, https://www.google.com/search?q=chatgpt++racism&tbm=isch&ved=2ahUKEwjB9fyQs_-EAxV8FGIAHci5AncQ2-cCegQIABAA&oq=chatgpt++racism&gs_lp=EgNpbWciD2NoYXRncHQgIHJhY2lzbUiMJFCjEVijEXAAeACQAQCYATSgAWWqAQEyuAEDyAEA-AEBigILZ3dzLXdpei1pbWfCAgQQIxgniAYB&sclient=img&ei. Accessed 19 March 2024.