In July 1956, Carnegie Mellon’s Herb Simon and Allen Newell presented what is considered the first AI computer program at the Dartmouth Summer Research Project on Artificial Intelligence. Called the “Logic Theorist,” it was designed to mimic human reasoning by proving theorems in math, a narrow goal indeed. Simon and Newell soon followed it with a broader and grandly named “General Problem Solver.” And so, the field of applied AI was born 70 years ago.
As a historical aside, I knew both Simon and Newell in the mid ’70s when I was on the CMU faculty. In fact, they ran an innovation grant program that funded my hand-done book on the structure and measurement of complex issues and reasoning.
See http://stemed.info/reports/Wojick_Issue_Analysis_txt.pdf
I am on the “I” side of AI, studying human reasoning because we cannot make computers emulate what we do not understand.
Simon got a Nobel Prize in Economics in 1978, even though he never did any economics. They do not have a prize for inventing new fields like AI, but they are flexible.
AI has made steady progress. Here are a few big milestones.
In 1996, IBM’s Deep Blue defeated the reigning world chess champion.
In 2011, IBM’s Watson beat two champions combined in Jeopardy. Question answering became the basis for today’s amazing AI chatbots.
Chat GPT was launched in 2022 and here we are.
So, what can these machines do well that is useful? This is the question I am not seeing explored very much, rather, it is drowned out by the raucous hype for and against something called “AI.”
As for a name I like “reading and reasoning systems” (RRS) over “chatbots” or “large language models” which give no hint what is going on. These RRS machines can emulate reading enormous amounts of stuff and reasoning about that stuff to an amazing degree.
These reading and reasons systems have already become standard use for a great many people, including me, simply because we do Google searches. They are no immediate job threat because few people just read and reason for a living.
It is the cognitive space between a search and a job that needs to be explored in detail. I have done a tiny bit of this over the last two years, so I offer these seven articles as starting points for further analysis and discussion.
“AI may bring a cognitive renaissance to human thinking”
https://www.cfact.org/2026/02/27/ai-may-bring-a-cognitive-renaissance-to-human-thinking/
“Using AI to understand big bodies of research”
https://www.cfact.org/2025/12/29/cfact-comments-on-using-ai-to-understand-big-bodies-of-research/
“AI emulates abstract thinking about Kipling, Lady Gaga, and The Rolling Stones”
“AI knows it is biased on climate change”
https://www.cfact.org/2024/11/04/ai-knows-it-is-biased-on-climate-change/
“AI’s key role in science education — grade level search”
https://www.cfact.org/2024/10/15/ais-key-role-in-science-education-grade-level-search/
“AI could take your computer from search to research”
https://www.cfact.org/2024/09/18/ai-could-take-your-computer-from-search-to-research/
“AI chatbots are automated Wikipedias warts and all”
https://www.cfact.org/2024/03/16/ai-chat-bots-are-automated-wikipedias-warts-and-all/
On a final note, reading and reasoning systems can greatly improve our cognition, just as air conditioning greatly improves our comfort, but both use a lot of electricity. This is an essential feature not a flaw. Both are working hard for us.