Skynet Goes Live

As AI chatbots get more conversational, one question hangs in the air: could AI be conscious? The latest public thinker to announce openness to the idea is famed evolutionary biologist Richard Dawkins. After spending three days conversing with Anthropic’s AI chatbot Claude about a range of topics, the noted atheist acknowledged that he was now open to the possibility that artificial intelligence may possess consciousness.

Dawkins is just the latest in a long line of people going back centuries who have attributed mind to machines. In the 1960s, almost 60 years before Claude came along, MIT professor Joseph Weizenbaum discovered something unsettling about human beings: we are astonishingly quick to attribute mind, personality, and even intimacy to machines that merely simulate conversation.

Weizenbaum built a primitive chatbot he called ELIZA, a program that relied on keyword detection, pattern matching, and scripted response templates to carry on a conversation. The computer scientist was under no pretenses about ELIZA. He knew it was driven by comprehensible procedures. What shocked him was the strong emotional responses he observed from people who interacted with ELIZA. He wrote in a 1976 book: “Extremely short exposures to a relatively simple computer program could induce powerful delusional thinking in quite normal people.” He often gave the example of his secretary at MIT, who watched him work on the program for months. After only a few sessions with ELIZA, she asked him to leave the room so she could more privately confide in the program.

Weizenbaum was also taken aback by how quickly science professionals were willing to consider programs like ELIZA as substitutes for human treatment. Psychiatrist Kenneth M. Colby said ELIZA could be “made widely available to mental hospitals and psychiatric centers suffering a shortage of therapists.” Noted astrophysicist Carl Sagan also weighed in on ELIZA’s potential, envisioning a future where such programs would provide medical care to the masses: “[F]or a few dollars a session, we would be able to talk with an attentive, tested, and largely non-directive psychotherapist.”

ELIZA also taught Weizenbaum that people are very generous with their assumptions in conversation: “The human speaker will…contribute much to clothe ELIZA’s responses in vestments of plausibility,” he wrote. In other words, trained as we are since before birth in human conversation, and lonely as we might be in today’s modern age, we are more than happy to meet a conversation partner halfway, even if that partner is a machine.

Which brings us back to Mr. Dawkins.

Today’s Large Language Model (LLM) chatbots like Claude and ChatGPT are vastly more sophisticated than ELIZA, harnessing newer technologies in natural language processing to make conversations much more engaging. And as Weizenbaum predicted, today’s AI models can build belief-structure models of users across multiple sessions. It’s little wonder, then, that even eminent thinkers like Dawkins would be flattered and impressed to the point of acknowledging the possibility of consciousness.

But is AI actually conscious?

Under the hood of today’s AI models, it’s still all about patterns. Algorithms trained on vast datasets analyze our queries letter by letter and harness statistical analysis to produce responses arranged in patterns we recognize as human language. And it’s getting better and faster at doing this every month. Give an AI model enough data and train it to shape its responses in patterns we recognize as conversation, and the sky is the limit. AI can simulate intelligence, consciousness, romance, concern, interest, curiosity, or anything else that typifies human existence. But simulations, while powerful, are still only simulations.

If Mr. Dawkins can get lulled into entertaining notions of AI consciousness, all of us can. It’s crucial that we get in the driver’s seat with AI now so we don’t get taken for a ride later. If you’re going to use an AI chatbot, keep it limited and advisory. Avoid personalizing chatbots with names or personal pronouns. And wherever possible, opt for the real thing over a simulation. The most amazing computer in the universe is still your own brain. Use it to blast past mere simulations and unleash the real thing.

***

Andrew McDiarmid is a senior fellow at the Discovery Institute. Follow his writing on technology at https://thehumanadventure.substack.com.