Warnings From A Top AI Expert: Dr. Geoffrey Hinton Resigns From Google

The race for AI supremacy has been heating up lately, with almost every major player releasing their brand of artificial intelligence. Everything from writing assistants to every flavour of chatbot has popped out of the woodwork. From our standpoint, we find most of them incredibly useful. They make our workflows easier, our content better, and, ultimately, our lives easier. But, not everyone is so eager.

You will undoubtedly recall the petition from names like Steve Wozniak to Elon Musk calling for a six-month pause on new AI development, a powerful statement from titans of technology. A new voice has joined them in their concerns, dubbed “The Godfather” of modern AI; Dr. Geoffrey Hinton has added his voice to the growing chorus. He questions the wisdom of letting AI into our day-to-day operations.

Dr. Hinton went so far as to resign from Google to be a better neutral voice of reason in the global AI arms race. There was a degree of sensationalism to the reporting by the New York Times, seemingly portraying his resignation from Google as a criticism of the company. Dr. Hinton quickly clarified on Twitter that he believes Google to be a responsible steward of AI and that the split was so he could better speak out without considering the repercussions of his words on the organization.

Dr. Hinton’s reservations about AI are the lack of oversight and regulation. The unfettered proliferation of AI is driving incredible technological advancements. Still, it’s also leaving us open to the negatives that come from any groundbreaking technologies. The doctor worries about what will happen should bad actors use the technology to create a world in which the average person cannot discern reality from AI-created fiction. He also worries about the societal repercussions of job displacement driven by AI.

We can all agree: fake news and more spam would be a bad world. But it’s also worth noting that a viable education, medical, or assistive play is being made for every nefarious use case. There are legitimate use cases for AI that are genuinely transformational. It was found that AI is more conscious of bedside manners than many doctors. It makes sense; AI isn’t affected by burnout, tight deadlines, or societal/familial pressures.

So what can government do? Well, let’s clear up the thinking on the petition mentioned earlier. Sam Altman, CEO of ChatGPT creator OpenAI, responded to the assertions that they are already training the next generation of models. Altman says, with certainty, OpenAI is not working on GPT5, something he found misleading about the petition asserting they need to pause GPT5 development. He did state, however, that things they are working on require careful development and consideration.

Altman previously stated that government has an excellent opportunity to get involved in regulating the creation, training, and implementation of AI. Canada is working on Bill C-27, dubbed “The Artificial Intelligence and Data Act.” The act aims to set reporting and record-keeping standards for creating and deploying AI systems. They also aim to set standards for informing the public about the types of AI systems they are interacting with and the scope of those systems.

These are reasonable first steps, though we still need some crucial wording to be fully defined. Vague terms such as “high-impact AI system” must be clarified before we can genuinely assess the legislation’s effectiveness and fairness. Overall, the industry welcomes oversight, especially when it is measured and informed, unlike some current bans.

To borrow a much-overused phrase, “The genie is out of the bottle.” There’s no going back to a time before AI. Dr. Hinton’s work dates back to 2012; it should be no surprise to the government or industry that this technology was coming; our own Fuad Miah has been a loud proponent of the AI revolution for the better part of the last decade.

What Dr. Hinton’s stance will ultimately mean is anyone’s guess. Still, we hope it will lead to a measured and safe approach to AI and regulation that will make him again comfortable with his life’s work. There are still significant questions regarding the Open-Source community, a much more complex group to regulate and monitor. While the big players are open to regulation and restriction, a voice is also growing in social media channels, decrying the measures already in place to mitigate malicious or unsafe content generation as too draconian and puritanical for their tastes.

There must be a balance between the freedoms of the individual to empower themselves with technology and safeguarding citizens and society. As leaders, we can do our part by examining how to safely and securely implement AI in our organizations. At the start of every project, we take a lot of time to understand the goals and impacts of every AI implementation. It’s important that this be a transparent process with all stakeholders fully involved. Together we can take advantage of this transformational technology safely and effectively!