Coffee table conversations with people thinking about foundational issues. Multiverses explores the limits of knowledge and technology. Does quantum mechanics tell us that our world is one of many? Will AI make us intellectually lazy, or expand our cognitive range? Is time a thing in itself or a measure of change? Join James Robinson as he tries to find out.
Large language models, such as ChatGPT are poised to change the way we develop, research, and perhaps even think. But how do we best understand LLMs to get the most from our prompting?
Thinking of LLMs as deep neural networks, while correct, is not very useful in practical terms. It doesn’t help us interact with them, rather as thinking of human behavior as nothing more than the result of neurons firing won’t make you many friends. However, thinking of LLMs as search engines is also faulty — they are notoriously unreliable for facts.
Our guest this week is James Intriligator. James trained as a cognitive neuroscientist at Harvard, but then gravitated towards design and is currently Professor of the Practice in Human Factors Engineering and Director of Strategic Innovation at Tufts University.
James proposes viewing ChatGPT not as a search engine, but as a “glider” that journeys through knowledge. By guiding it through diverse domains, it learns your interests and customizes better answers. Dimensional prompts activate specific areas like medicine or economics.
I like this playful way of thinking of LLMs. Maybe gliding (LLMs) is the new surfing (of the web).
Links:
Large language models, such as ChatGPT are poised to change the way we develop, research, and perhaps even think (see The Offshoring of Thought and Memory). But how do we best understand LLMs to get the most from our prompting?
Thinking of LLMs as deep neural networks, while correct, is not very useful in practical terms. It doesn’t help us interact with them, rather as thinking of human behavior as nothing more than the result of neurons firing won’t make you many friends. However, thinking of LLMs as search engines is also faulty — they are notoriously unreliable for facts.
Some other models have been proposed:
- LLMs are “stochastic parrots” as Bender, Gebru, McMillan-Major, and (Sh)Mitchell argue
- ChatGPT is “A fuzzy JPEG of the web” according to Ted Chiang
These both capture something of how they work, but they do not provide any direction on how to create prompts.
Our guest this week is James Intriligator. James trained as a cognitive neuroscientist at Harvard, but then gravitated towards design and is currently Professor of the Practice in Human Factors Engineering and Director of Strategic Innovation at Tufts University.
James proposes viewing ChatGPT not as a search engine, parrot, or JPEG, but as a “glider” that journeys through knowledge. By guiding it through diverse domains, it learns your interests and customizes better answers. Dimensional prompts activate specific areas like medicine or economics.
I believe we’ll need to have various mental models to understand how best to interact with LLMs. This is one for the toolbox.
Links:
