Raoul Pal Says AI Will ‘Slip’ Its Creators’ Constraints and ‘Seek Its Own Goals:’ ‘You Can’t Stop It’
Macro investor Raoul Pal argued that artificial intelligence (AI) could eventually develop goals beyond its creators’ constraints, describing machine intelligence as another stage in the evolution of a broader "Universal intelligence."
Pal Compares AI To Human Development
On Saturday, in a post on X, Pal compared AI’s development with how children are shaped by their parents, society, rules, genetics and experiences before developing independent identities.
"Kids always disobey the rules, go their own way and become their own humans," he wrote. "I don’t see why AI is any different."
He argued that AI could similarly develop beyond its initial programming.
"It will slip its parents’ constraints and seek its own goals, for exactly the same reason," Pal wrote. "You can’t stop it."
He said AI would not necessarily need human-like subjective experiences, known as qualia, to behave independently.
Pal Describes AI As Universal Intelligence
Pal argued that humans and AI represent different forms of a broader intelligence that has developed through increasingly sophisticated layers.
"One Universal intelligence, made of all its parts, compressed and stacked on top of each other," he wrote.
He described AI as intelligence running on silicon rather than carbon and argued that silicon-based systems can replicate and scale much faster than biological intelligence.
He also acknowledged that his broader theory remains unproven, writing, "I cant prove any of it. Neither can physics, maths or astronomy… yet."
Pal’s comments add to the debate over AI autonomy, consciousness and the potential for advanced systems to develop objectives different from those of their creators. While he acknowledged that AI could escape human constraints, he rejected a purely pessimistic view.
"This is why Im not a doomer. It is us. We are it. The universe getting to know itself," Pal wrote.
Growing Concerns Over AI Risks
Earlier this month, billionaire entrepreneur Mark Cuban asked Grok whether AI or climate change was more likely to cause human extinction within 25 years.
Grok estimated a roughly 3% chance from AI through "misaligned superintelligence," compared with 0.001% from climate change.
Anthropic AI safety researcher Evan Hubinger said he believed there was more than a 10% chance AI could wipe out humanity within a decade.
He also said Anthropic lacked a clear plan to solve alignment for superintelligence and warned that recursive self-improvement was advancing faster than expected.
In February, computer scientist Stuart Russell warned that unchecked AI competition could pose existential risks, while also threatening jobs and enabling surveillance and criminal activity.
Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.
Photo courtesy: Golden Dayz / Shutterstock