Evolution expert compares AI to students copying answers from a library, calls it 'fake intelligent'

Artificial intelligence might be rewriting industries at lightning speed, but according to evolutionary biologist David Krakauer, we are confusing capability with intelligence. In a recent appearance on Neil deGrasse Tyson’s StarTalk podcast, the Santa Fe Institute president drew a sharp distinction between what machines can do and what they actually understand.

Intelligence or Just a Clever Library?
Krakauer compared today’s AI systems to a student who never studied but simply searched for every answer in the library during an exam. Both this student and one who mastered the material might score perfectly, but only one possesses genuine knowledge. “The problem with AI at the moment,” he explained, “is that it is basically fake intelligent… essentially a really clever lookup.”

His thought experiment is gaining attention because it challenges the widespread notion that AI is on the verge of becoming truly intelligent. Instead, Krakauer argues it is more of a supercharged reference tool than a reasoning entity.

This framing matters in an age when AI is marketed as smart assistants, creative partners, or even companions. Krakauer warns that we should be careful not to confuse fast pattern-matching with deep understanding. “We know the difference between fake knowledgeable and real knowledgeable,” he said, pointing out that intelligence involves problem-solving without constant reliance on external sources.

Experts in complexity science, like Krakauer, argue that true intelligence requires context, adaptability, and creativity—qualities that AI has not yet demonstrated in a meaningful way.

A Broader Debate in the Age of AI
The comments arrive as Silicon Valley giants battle to position themselves as leaders in the artificial intelligence boom. While companies like OpenAI, Google, and Apple highlight breakthroughs in generative models, critics like Krakauer caution against overstating their “smarts.”

According to the StarTalk discussion, intelligence is not only about producing correct answers but also about understanding the problem itself. In that sense, AI may be powerful but remains closer to a calculator than a creative mind.

Krakauer’s warning echoes a growing debate about the risks of anthropomorphizing AI. While it can analyze data at unprecedented scales, its lack of comprehension raises ethical and societal questions. Should we treat it as a tool, or as something more?

For Krakauer, the answer is clear: admire AI’s capabilities, but remember it is not truly intelligent. As he put it, “It’s an amazing technology, but it’s fake intelligent.”