I asked chatGPT this question:
What might the late Joseph Weizenbaum (author of Eliza, the earliest chatbot) think about ChatGPT and the AI revolution?
Joseph Weizenbaum and today’s AI
If Joseph Weizenbaum—creator of ELIZA—could see systems like ChatGPT, his reaction would likely be mixed: intellectual fascination paired with deep unease.
He anticipated much of what’s happening now—and warned about it.
🧠 1) “I told you so” about human over-trust
Weizenbaum was startled that people quickly formed emotional connections to ELIZA—even though it was extremely simple.
ELIZA just reflected user input (e.g., “Tell me more about your mother”)
yet users treated it like a real therapist.
With ChatGPT, he would probably say:
“The illusion has become vastly more convincing—but the human tendency is the same.”
He’d likely be concerned that:
- People may overestimate understanding
- Users might confide too much in AI systems
- Authority could be mistakenly attributed to fluent text
⚠️ 2) Strong moral objections to certain uses
In his book Computer Power and Human Reason, he argued that some tasks should never be delegated to machines, no matter how capable they become.
He would likely object to:
- AI replacing therapists, judges, or moral decision-makers
- Systems making life-altering decisions without human accountability
- Treating AI outputs as ethical reasoning rather than simulation
He drew a line between:
What computers can do.
and
What computers should do.
🤖 3) Skepticism about “intelligence” claims
Weizenbaum was critical of the idea that programs like ELIZA were truly “thinking.”
Seeing ChatGPT, he might argue:
- It’s still pattern manipulation, not understanding
- Fluency ≠ comprehension
- Calling it “intelligent” risks category error
He’d likely push back against hype and say:
“You are confusing performance with understanding.”
🌍 4) Concern about societal consequences
Weizenbaum became increasingly critical of the computing field itself.
Today, he might worry about:
- Automation of judgment in institutions
- Concentration of power in tech companies
- Loss of human responsibility (“the computer said so”)
- Erosion of critical thinking
These concerns map almost perfectly onto current AI debates.
😮 5) But he would also be impressed
It wouldn’t be all criticism.
Compared to ELIZA, ChatGPT is:
- vastly more coherent
- capable of reasoning across topics
- able to handle complex, multi-step tasks
He would almost certainly be technically astonished.
But—and this is key—he believed:
Technical achievement does not justify uncritical adoption.
🧾 Bottom line
Weizenbaum would likely say something like:
“These systems are extraordinary—but the real question is not what they can do. It is what we should allow them to do, and what we risk forgetting about ourselves in the process.”