Gen AI - Week 6 - Google Slides
This week we continue with reading and discussion of Bender et al.’s influential Stochastic Parrots paper – in part because it brings up criticisms that remain relevant even with the changing face of AI today.
In addition this week – and as time permits – we broaden still further the scope of our inquiry. One of the objects of Bender et al.’s critique is the implicit assumption that bigger means better: the larger the language model, the closer we reach to Artificial General or Super Intelligence.
We might ask a follow-up question: what does it mean if this goal, or something like it, is realised? What sort of society do we have to look forward to? It is common to imagine a dystopic or a naively utopic future, but the readings I’ve picked here instead pull from different perspectives to think of – arguably – more nuanced positions. They broadly come under the title of “accelerationism”: the idea that this car we are driving into the future needs no breaks, but instead all force applied to the accelerator pedal.
As always, not all of the readings are required. Scan or read what you can. We start with one of the most famous statements of optimism in our technological destiny, from perhaps an unlikely source: Karl Marx. We then examine an important statement of futurist Marinetti, who more than any other captured – in the 1910s – the enthusiasm for near-future. We then – excuse the pun – accelerate toward the recent past, where accelerationism has enjoyed a voluminous, if largely subcultural, life on various heterodox platforms and publications.
It might be asked why this emphasis on what seems a strange philosophical offshoot (sidenote: you’ll see references to Deleuze & Guattari, two hugely influential figures of the 70s and 80s). However the core issues are ones that underpin an enormous amount of the discourse on both left and right side of politics around AI today. Peter Thiel, for example, is an influential Silicon Valley investor, with strong ties to Elon Musk and J.D. Vance. Alex Williams and Nick Srnicek, on the other hand, were at one time connected to Jeremy Corbin, leader of the British Labor party for several years. Meanwhile, of the thousands of articles and op ed pieces written by the producers of AI, I’ve selected a relatively modest and careful article by Anthropic’s CEO, Dario Amodei.
As we will conclude next week, how we think about generative AI for education depends in part on how we think about technology in relation to “big picture” social topics: the future of work, the possibility of abundance, the prospects of wealth redistribution, the geopolitical matrix assembling around AI, and the destiny of the human.
Question for the week: You’ll notice this week is heavily sprinkled with “manifestos” of different kinds. Why? What is it about technology that inclines people to declare to make things manifest?
Readings:
Amodei, D. (2024). Machines of Loving Grace: How AI Could Transform the World for the Better. https://darioamodei.com/machines-of-loving-grace
Marinetti, F. T. (1909). The Futurist Manifesto. Le Figaro, 20, 39-44. https://www.societyforasianart.org/sites/default/files/manifesto_futurista.pdf
Marx, K. (1993). Grundrisse: Foundations of the critique of political economy (rough draft) (M. Nicolaus, Trans.). Penguin Books. (Original work written 1857-1858). https://thenewobjectivity.com/pdf/marx.pdf
Thiel, P. (2007). The Straussian Moment. Politics and Apocalypse, 199. https://mindseyemag.com/magazine/wp-content/uploads/2024/07/2007-thiel.pdf
Williams, A., & Srnicek, N. (2013). # ACCELERATE MANIFESTO for an Accelerationist Politics. Rhuthmos. https://criticallegalthinking.com/2013/05/14/accelerate-manifesto-for-an-accelerationist-politics/
Haraway, D.J., ‘A Cyborg Manifesto: Science, Technology, and Socialist-Feminism in the Late Twentieth Century’, Manifestly Haraway (Minneapolis, MN, 2016; online edn, Minnesota Scholarship Online, 19 Jan. 2017), https://doi-org.proxy2.library.illinois.edu/10.5749/minnesota/9780816650477.003.0001.
Supplementary Material:
A useful recent overview of accelerationism: https://xenogothic.com/2019/03/04/a-u-acc-primer/
Mackay, R., & Avanessian, A. (Eds.). (2014). # Accelerate: The Accelerationist Reader. MIT Press.
A recent discussion of AGI and abundance on Ezra Klein’s podcast – https://www.youtube.com/watch?v=Btos-LEYQ30
Assignment(s):