Azeem's commentary: My second robot brain
Three ways in which ChatGPT improves our quality of thinking.
I’ve been playing around with ChatGPT and other AI tools based on large language models at your favourite wondermissive. It is too early to say which of these behaviours will actually stick. Based on my experience so far, I think that using Elicit for research, Midjourney for image generation and various summarisiation products will become part of our workflows.
I’ve also realised that ChatGPT is quite a helpful second brain. In fact, I’d go so far as to say that I’ve been able to use it to improve my critical thinking.
Working with, one of EV’s team members, we’ve come up with three ways in which ChatGPT improves our quality of thinking.
The divergent thinker,
The challenging critic,
Let’s step through each of these with an example.
The divergent thinker
When I’m evaluating a problem, I’ll throw light at it from different angles. I might view it from the lens of declining cost. I might use a structured analysis (think Porter’s Five Forces), or I might historicise it (“was this like cyclicality in the DRAM business?”), or look at it from the perspective of individual motivations or incentives.
Putting the question to ChatGPT with the instruction to summarise key arguments often fleshes out a different way of framing the problem.
Asking ChatGPT to apply a framework to a new subject is also helpful; take for example a Weberian analysis of the Protestant ethic in 21st century capitalism, 117 years after the publication of The Protestant Ethic and the Spirit of Capitalism.
Finally, it can help explore counter-factual situations, and force the prompter to ask the correct questions.
The broad range of ideas the chatbot comes up with is, well, what it is. Some are trivial. Some are relevant. Some are simply wrong. But like the Post-Its gathered during a brainstorming session, it can, occasionally, throw up interesting avenues for further research.