Book Review: Supremacy by Parmy Olsen
First off, I do recommend this book as it struck the balance well between the human elements and the technological jumps of the last decade or so in AI.
There were a few parts that stuck out for me, in both a general sense and when applied to higher education.
General thoughts
First off, my feeling is that particularly Sam Altman’s motivation being purely idealistic and around creating AGI to solve the world’s problems just doesn’t line up at all with his subsequent actions. Dario Amodei also comes across as a total hypocrite, as he apparently left OpenAI because of their lack of safety work and following profits, and then went out and started, checks notes, Anthropic and raised billions of dollars.
On the Google side, the accusation levelled at Demis Hassabis and DeepMind is that they were too academic, which for obvious reasons I enjoy. In this context it was their inability to actually deploy their innovation that made them ‘academic’, and resulted in them sitting on a ChatGPT style chatbot for two years before the launch of ChatGPT.
In the higher education front, it sparked a lot of thoughts:
Vendor Lock-in
AI is arriving in HE embedded in platforms we already depend on: Microsoft 365 (Copilot), Google Workspace, major LMS/VLEs etc
Sector risk isn’t “which chatbot should students use?” so much as “which 1–2 platform providers do we want to give structural control of our digital university to?”
Misalignment
Consumer tools designed for engagement and cross product integration, not critical thinking etc.
learning environments where the vendor motive is “maximise usage and stickiness”, not “cultivate disciplined, critical engagement with AI”.
HE as counterweight
Universities are also one of the few classes of institutions capable of building open, critical, transnational alternatives to these providers, be that through open models, public-interest data governance, and serious interdisciplinary work on AI’s social impacts.
The question for Higher Ed is are we acting as distribution channels and legitimisers for US big tech, or as a counterweight that can shape AI towards public, civic and educational purposes?