Lex Fridman PodcastPamela McCorduck: Machines Who Think and the Early Days of AI | Lex Fridman Podcast #34
At a glance
WHAT IT’S REALLY ABOUT
Pamela McCorduck recalls AI’s mythic roots and scientific audacity
- Pamela McCorduck, author of *Machines Who Think*, reflects on the early days of artificial intelligence, when the field’s founders saw themselves as launching something historically important despite being scientific outsiders. She traces AI’s roots beyond math and computer science into ancient myths, religious tensions, literature, and long-standing human desires to create intelligence outside the human brain.
- McCorduck discusses the personalities and motivations of AI pioneers like Newell, Simon, Minsky, and McCarthy, contrasting their pragmatic goals with later spiritual and apocalyptic narratives about AI. She challenges the popular notion of “AI winters” as commercial phenomena rather than true scientific stagnation, emphasizing the continuity of basic research.
- The conversation also explores symbolic vs. algorithmic AI, the influence of complexity science at the Santa Fe Institute, and speculative futures including empathetic “geriatric robots.” McCorduck weaves in feminist perspectives, arguing that AI can undermine the idea that intelligence is inherently male and critiquing male-driven existential fears about superintelligence.
- Ultimately, she remains a cautious optimist about AI’s future, worried more about unforeseeable side effects than scripted doomsday scenarios, and grateful to have witnessed the birth and evolution of a field she believes will fundamentally change the world.
IDEAS WORTH REMEMBERING
5 ideasAI’s roots are as much mythological and literary as they are mathematical.
McCorduck argues that the impulse to create artificial minds extends back through Homer, religious stories, early automata, and Babbage, showing AI as a continuation of a deep human desire to ‘forge the gods,’ not just a 20th‑century technical project.
Early AI founders knew they were creating a new field and acted accordingly.
Figures like Newell, Simon, Minsky, and McCarthy saw themselves as launching a historically significant discipline, even when others in computer science treated AI as fringe; they pursued cognitive modeling, mathematical formalisms, and “cool” problems with long‑term societal impact in mind.
The AI “winter” is largely a commercial story, not a scientific one.
McCorduck contends that the supposed winters were mainly about funding and unmet commercial hype, while fundamental research in areas like expert systems and symbolic reasoning quietly laid essential groundwork that later technologies build on.
Symbolic reasoning remains an underdeveloped frontier alongside today’s algorithms.
While deep learning and algorithmic approaches dominate current applications, she notes that higher‑level symbolic processing and knowledge representation have not been fully realized, representing fertile ground for future researchers.
Common AI fears reflect deep psychological and cultural patterns, including religious and gendered biases.
She connects primal anxieties about ‘blasphemy’ and Frankenstein myths to modern AI panic, and argues that male technologists’ fear of being outsmarted by machines—the “male gaze”—colors existential risk narratives more than we acknowledge.
WORDS WORTH SAVING
5 quotesArtificial intelligence began with the ancient wish to forge the gods.
— Paraphrased by Lex Fridman from Pamela McCorduck’s *Machines Who Think*
The idea of intelligence outside the human cranium—this was a phenomenal idea.
— Pamela McCorduck
AI winter is, to me, a crock… Science has a much longer view than commerce.
— Pamela McCorduck
This is the first guys who have always been the smartest guy on the block, and here comes something that might be smarter. Ooh, let’s stamp it out before it takes over.
— Pamela McCorduck (on male fears of superintelligence)
I was so tired of hearing that intelligence was inside the male cranium… I wanted something besides male smarts.
— Pamela McCorduck
High quality AI-generated summary created from speaker-labeled transcript.
