Why AI Pioneer Geoffrey Hinton Quit Google and What It Means for AI Safety

Guide: After ten years at Google, Geoffrey Hinton, a neural‑network pioneer, resigned, calling the dangers of chatbots "quite scary" and warning they could be exploited by "bad actors".

Suddenly, but not entirely unexpectedly, Turing Award winner and deep‑learning father Geoffrey Hinton announced his departure from Google.

A New York Times interview revealed the news a day earlier.

Hinton also confirmed his exit on Twitter.

The AI "godfather" left Google because he fears the spread of misinformation, AI‑driven job disruption, and the existential risks of creating truly digital intelligence.

Hinton, now 75, said he quit to speak freely about the AI crisis and expressed some regret about his contributions; he was hired by Google ten years ago to help develop its AI technology, and his methods paved the way for systems like ChatGPT.

Until last year he believed Google was a "suitable steward" of the technology, but when Microsoft integrated chatbots into Bing, Google began to worry about the risk of its search business being disrupted.

He warned that some chatbot dangers are "quite scary" and that they could become smarter than humans and be used by "bad people". "They can automatically generate massive text, enabling very effective spam bots, etc."

"My conclusion is that the AI we are developing is very different from existing human intelligence," he said. "It's like having 10,000 people; if one learns something, everyone automatically knows it, which is why these chatbots know more than any individual."

Misinformation spread is the most urgent issue for Hinton now. In the longer term, he worries AI will eliminate rote‑memory jobs and could eventually write and run its own code, threatening humanity itself.

"This thing could actually become smarter than humans. Many think that's far off. I used to think so too. I thought it would take 30‑50 years or more. Clearly, I no longer think that," he said.

He is not alone among AI researchers in fearing severe human harm from the technology.

Last month, Elon Musk argued with Google co‑founder Larry Page, accusing him of not taking AI safety seriously enough, and criticized the push toward "digital super‑intelligence".

Valérie Pisano, CEO of the Quebec AI institute Mila, warned that no industry tolerates reckless AI safety practices, emphasizing that developers must retain control when systems interact with humans.

Hinton's concerns are already materializing: AI‑generated images, videos, and text now flood the internet, making it hard to discern truth.

Recent upgrades to image generators like Midjourney enable realistic pictures, such as a viral image of Pope Francis in a Balenciaga coat.

Hinton also fears AI will replace lawyers, personal assistants, and many other professions in the future.

In 2012, Hinton's Toronto team achieved a breakthrough in deep learning that transformed speech recognition and object classification.

His collaboration with Alex Krizhevsky and Ilya Sutskever produced the AlexNet convolutional network, which won the ImageNet 2012 competition by a large margin, reducing the visual‑recognition error rate to 15.3%.

This milestone reshaped computer vision, and the trio later founded DNNResearch.

In 2013, Google acquired DNNResearch, bringing Hinton and his team into the company. In 2015, his student Ilya Sutskever co‑founded OpenAI.

Google chief scientist Jeff Dean praised Hinton's decade‑long contributions, saying he will be missed and that Google remains committed to responsible AI principles and learning about emerging risks.

He added that any data—images, audio, video—must be assumed potentially deceptive.

Now, ten years later, the 75‑year‑old Hinton has left Google, prompting speculation about the future direction of AI.

We await what his departure will inspire.

Reference link https://www.nytimes.com/2023/05/01/technology/ai-google-chatbot-engineer-quits-hinton.html