Dartmouth Summer Research Project on Artificial Intelligence
A group of pioneering scientists and mathematicians gather at Dartmouth College to formally propose and debate the creation of artificial intelligence as a new field of study. John McCarthy presents h
Setting
A spacious seminar room in Dartmouth College's mathematics department, filled with wooden desks arranged in a semi-circle facing a chalkboard covered in equations and diagrams. Large windows allow summer sunlight to stream in, partially filtered by leafy trees outside.
Characters
John McCarthy
primary
A lean man in his late twenties with sharp features, thick dark hair combed neatly back, and intense brown eyes behind round wire-frame glasses. His posture carries the quiet confidence of a mathematician who knows his proofs are sound.
Marvin Minsky
primary
A lean man in his late 20s with a sharp, angular face dominated by thick black-framed glasses. His dark hair is slightly unruly, suggesting he's been running his hands through it during intense thought. There's an energetic restlessness in his posture - a coiled intellectual energy.
Nathaniel Rochester
secondary
A bespectacled man in his mid-30s with a lean, wiry build and short, neatly combed brown hair. His sharp facial features are accentuated by a prominent nose and a clean-shaven jaw. His hands bear faint ink stains from frequent work with blueprints and schematics.
Claude Shannon
secondary
A lean man in his early 40s with sharp features and a receding hairline, his wire-rimmed glasses catching the light as he tilts his head. His long fingers occasionally tap a pencil against his knee in thoughtful rhythm.
Graduate Assistant
background
A young man in his early 20s, slender build with slightly tousled brown hair, wearing round wire-rimmed glasses that keep slipping down his nose. His face is flushed from the intensity of the moment, and his hands move with nervous energy.
Dialog
John McCarthy
Let me formalize this - if we can describe learning as a computational process with feedback loops, we can absolutely mechanize intelligence.
Marvin Minsky
No, wait - that can't explain pattern recognition! Consider this mechanism - a neural net that self-organizes like the visual cortex.
Nathaniel Rochester
Practically speaking, we've got 2048 words of memory on the 704. Either approach will need serious optimization.
John McCarthy
In the computational sense, Marvin, your networks are just another form of state machine. The mathematical principles-
Marvin Minsky
But where's the adaptation? Your formal logic can't account for Hebbian learning - the brain rewires itself!
Nathaniel Rochester
Gentlemen, the IBM 7090 ships next year with transistor logic. We'll have cycles to spare for both approaches.
John McCarthy
Precisely why we must establish formal foundations now. 'Artificial intelligence' shouldn't be constrained by today's hardware.