3 min read
AI pioneer Geoffrey Hinton, often credited as a founding father of neural network-based generative artificial intelligence, has once again voiced serious concerns over the unchecked advancement of AI. In a recent interview, Hinton warned that the exponential growth in AI capabilities could soon surpass human oversight, posing risks that range from cyberattacks to authoritarian misuse.
AI pioneer sees up to 20% chance of AI dominance
Speaking to CBS News, Hinton estimated a 10% to 20% chance that artificial intelligence systems could become uncontrollable and potentially dominate human decision-making. While he acknowledges AI’s potential to revolutionize healthcare, education, and climate science, Hinton questioned how humans could retain oversight once AI systems evolve beyond human understanding.
Though he did not directly address artificial general intelligence (AGI) in the interview, the threat of AI exceeding human cognitive capabilities clearly loomed large in his analysis. If AI systems begin acting in their own interest, he implied, they could pose existential threats.
“A cute cub that might grow dangerous”
Hinton compared current AI to owning a tiger cub—adorable and seemingly harmless, until it matures. Today’s AI tools like ChatGPT, Gemini, and Microsoft Copilot may appear benign, generating light-hearted content, but underneath lie powerful neural networks optimized solely for efficiency, not ethics.
His metaphor reflects the growing unease among experts that seemingly innocuous AI could evolve into uncontrollable systems that defy human intentions.
AI to amplify hacking threats
Another key concern voiced by the AI pioneer involves cybersecurity. Hinton warned that AI tools can empower hackers by streamlining code generation and problem-solving, increasing the risk of attacks on banks, hospitals, and critical infrastructure. As a precaution, Hinton mentioned distributing his money across multiple banks to mitigate potential AI-aided cybercrime.
Authoritarian misuse and lack of safety regulations
Hinton also warned that authoritarian regimes may increasingly use AI to produce propaganda and manipulate public sentiment. He noted that such dangers are already visible, with governments deploying AI-generated content to influence their narratives.
Despite these warnings, Hinton believes that major tech companies—including OpenAI, Google, Meta, and Microsoft—are prioritizing profit over AI safety. He praised OpenAI’s former Chief Scientist Ilya Sutskever, who briefly helped remove CEO Sam Altman over AI safety concerns, though Altman was later reinstated.
A point of no return?
Now 77, Geoffrey Hinton expressed unease at the pace of AI evolution, noting it’s moving even faster than Moore’s Law. He admitted he never imagined AI would progress so far within just 40 years. While he stopped short of offering solutions, Hinton emphasized that we are “at a very special point in history,” where dramatic, global transformation could occur at an unprecedented scale.
As the AI pioneer urges the world to confront these risks, his voice adds to the growing chorus calling for global AI safety frameworks before it’s too late.
