Future of Text in AI in academia special session
At the end of April, I played a small part in discussing AI and academia. The biggest thrill, as always, was talking with Ted Nelson and Vint Cerf. Here’s the talk and snippet of the talk (edited for clarity):
- Comics as a bastard medium, a product of newspapers and journalism
- Robert Caro’s writing process and the natural affordances of writing on paper
- Sheet music
- How the fluency engine could help take back some of the lost affordances of our current medium
- Examples of how the fluency engine could be used to explore ideas and create compositions
- The potential for the fluency engine to be used as an art form
- The limitations of our current text editors and the potential for ChatGPT to bring back semantic geometry in writing and drawing
Venkatesh Rao recently wrote about the physics of intelligence.
I think it’s been adequately demonstrated in the last six months that at least intelligence, if not subtler notions like consciousness or sentience that may not, may or may not be well-posed is not a substrate, is not substrate dependent. The physics of intelligence is no more about silicon semiconductors or neurotransmitters than the physics of flight is about feathers or aluminum. These low-level substrates constrain but do not define physics. In either case, when you analyze the flight characteristics of an airplane or a bird, you might quickly check the strength-to-weight ratio of bone and feather composites or aluminum. But then you move on to talking about wing geometry, lift versus drag, and so on, concepts that still belong in physics but at different levels of abstraction. Flight is actually a very good reference phenomenon for thinking about intelligence. Since it is, it too is a property of biological organisms that reproduce with non-living machines that work on similar but not identical principles. Understanding the physics of flight in a way that’s agnostic to the differences between birds and aircraft is a similar problem to that of understanding the physics of intelligence. Whether realized with silicones or neurons. Interestingly, even though aerospace engineering is a mature discipline today, the physics of flight is actually still quite mysterious.
One of the things I’m personally interested in, and this might apply to Ted; it’s certainly in the spirit of Xanadu, is that even the most conservative critic of GPT would probably concede that GPT (LLMs and transformers) is as big a deal as newspapers. I was a cartoonist in my previous life and comics are a bastard medium. They are an accident or byproduct of technology, in this case, newspapers. And then there was this appetite for images and whimsy or something else that kind of glommed together and comics spilled out. Now comics are a new art form or new medium. called ‘comics.’ And I think that we’ll see the same thing with GPT. I think a bastard medium will pop out of GPT. This is why I’m excited about GPT. Not for the productivity boost, the better search, or the virtual assistant. But the silly side effects that lead to a new way to express ourselves.