What if AI could rewrite one non-military decision in human history?
In this AI-swers Aftertalks episode, we examine how leading AI models approach counterfactual history. By excluding wars and military decisions, the experiment forces artificial intelligence to evaluate the quieter choices that shaped technology, public health, infrastructure, culture, and the environment.
The discussion explores decisions connected to Steve Jobs and the original iPhone, Alexander Graham Bell and communication technology, Thomas Edison and the War of Currents, Jonas Salk and the polio vaccine, and Thomas Midgley Jr.’s role in the adoption of leaded gasoline.
But changing history is never as simple as correcting one bad decision. Every intervention creates second- and third-order consequences. Faster electrification may increase fatal accidents. Patenting a vaccine may improve manufacturing while destroying public trust.
The deeper question is whether large language models genuinely understand historical causality—or whether they overvalue famous individuals because those people occupy more space in their training data.
CHAPTERS
00:00 What single decision would AI rewrite?
01:30 The six AI models and the no-war constraint
03:08 Great Man Theory vs. structural history
06:00 How counterfactual history is tested
07:25 Steve Jobs and the missing iPhone stylus
09:45 Why AI overvalues modern technology
11:30 Alexander Graham Bell and the telephone
15:00 Edison and the War of Currents
18:00 Faster electrification and hidden risks
22:10 Jonas Salk and the polio vaccine
24:00 Should Salk have patented the vaccine?
28:45 Thomas Midgley Jr. and leaded gasoline
31:30 The global cost of lead pollution
33:20 The unintended ethanol and food crisis
36:40 Does AI confuse fame with causality?
39:00 Why LLMs do not truly understand history
40:20 Comparing the five counterfactual choices
42:00 What happens when AI writes our future?
43:30 Who should control the historical “code”?