I refrained from discussing my own project during my second KIPP group meeting (as I had taken up a large amount of time during the first one), but my third meeting was much more fruitful. I explained that my project idea has become broader, with a focus on the implications of AI’s advent occurring within a capitalist system. My group undertook thought-provoking discussions on this topic and provided some helpful tips and pieces of feedback, which are summarised below.
- Our KIPP TA, Joe, kicked off the discussion by highlighting an article discussed by Nathan Dana on his blog which ties in to my own project idea. Khan (2018) discusses steering AI development in favour of workers over capital interests and underscores the importance of discouraging the use of AI for purely capitalist gains, promoting the idea that AI’s only purpose should be to drive social progress
- Joe also suggested the exploration of not-for-profit organisations e.g. Linux
- A peer expressed disappointment in OpenAI’s shift from a not-for-profit model to a for-profit entity, citing the significant Microsoft investment. They discussed the unprecedented board activity from November and expressed cynicism about the deviation from the original ambitions of OpenAI. Advocating for simpler, less expensive AI models with broader applications, they questioned the necessity of profit-seeking in LLM companies
- The peer also contrasted this viewpoint by providing potential justifications for the profit-seeking motive surrounding LLMs, questioning the high training costs and emphasising the significance of data collection for model value. They suggested the industry should focus on simpler, less expensive models to bring about positive societal changes
- Another peer expressed concerns about corporate motivations in AI, suggesting the prohibition of corporate AI development, and mentioning the UK national standard for transparency. They proposed making such standards compulsory. Emphasising public awareness, they supported policies which would make life harder for corporations, whilst acknowledging the necessity of standards despite potential law violations
══════════════════
The meeting helped to highlight some of the complexities within, and potential rebuttals to, my thesis. To state the obvious, AI is an especially new discipline, being a maximum of 68 years old (McCarthy et al., 2006) but economic theory has been much more thoroughly explored. I’m starting to become more comfortable in my project’s grounding within the existing literature, and I’m really excited to research a range of economic theories, from Marxism to late-stage capitalism, and apply these older ideas to 21st-century technology.
References
McCarthy, J., Minsky, M. L., Rochester, N. and Shannon, C. E. (2006) ‘A proposal for the Dartmouth summer research project on artificial intelligence’, AI Magazine, 27(4), pp. 12-14.
Khan, S. R. (2018) ‘Reinventing capitalism to address automation: Sharing work to secure employment and income’, Competition & Change, 22(4), pp. 343-362.
Curated using ChatGPT 3.5 by OpenAI. For more information visit chat.openai.com.
