Nvidia founder and CEO Jensen Huang laid out his position on the risks posed by artificial intelligence in strikingly clear terms during remarks at Salesforce's Dreamforce conference on Tuesday.
اضافة اعلان
For him, AI is not some new form of "alien intelligence," as one OpenAI safety researcher has described it. It is, simply, hardware and software, he says, made by humans. And that means, in his view, that humans can control it, and that existing laws are capable of handling it.
He said: "Safety is an engineering problem, not a legal problem. We're ultimately developing software. We're developing computing systems. It's a complex computing system, but at the end of the day, it's still a computing system."
Based on that, Huang believes there's no need to pass new laws or regulations to manage AI. In fact, he doesn't see much need for new laws at all. He believes the free market will be enough to pressure companies against putting out unsafe products.
He said: "If we're not confident about the safety of our products, then, as is the case with all companies, like you and me, and every company here, if you're developing a product or service and you're not confident in its functionality, its capabilities, or its safety, don't launch it. That's pretty obvious."
He continued: "You have to pace yourself until you're confident, you're putting something out that the market will value. Market forces already exist. We don't need any new laws. We don't need new regulations. We just need companies to decide when they can move as fast as possible. I think the trade-off between innovation and speed on one hand, and safe products on the other, is a false trade-off. You can certainly achieve both at the same time. So move as fast as you can. But if at any moment you feel things at the company have gotten out of control, or that the product won't be safe, pause for a bit and make sure you're doing it right."
In some ways, this idea sounds reassuring. If there's anyone in the world who knows AI well, it's the founder of Nvidia, who was building the physical "brains" underlying AI systems long before ChatGPT ever appeared, and who now runs a company that also develops open-source models, AI agents, operating frameworks, and sandboxed testing environments.
But viewed with a more skeptical eye, his position isn't surprising either, for someone who has benefited this much from the AI boom.
Why, after all, would he want to see regulations emerge that add a layer of restrictions that could slow down Nvidia's efforts to sell more and more AI systems and software?
As he also said during the interview: "I'm more ambitious than ever. And as a result of our ambition, and with the increase in productivity we're getting from AI, there are no limits to what we can achieve. No limits to what our company can achieve. And no limits to what every industry, and every country, can achieve."
Unfortunately, even companies with the best of intentions can put out flawed products that lead to unintended consequences, and that includes software too.
Remember the CrowdStrike "Blue Screen of Death" disaster in 2024, which led to the cancellation and disruption of thousands of flights and caused chaos for businesses elsewhere?
There are also companies accused of acting deliberately with less good intentions. Meta recently paid 18 billion dollars to settle a lawsuit related to harm caused by social media to children.
AI has already caused harm too, regardless of intentions or safety tests carried out, ranging from an OpenAI model breaching the Hugging Face platform, to lawsuits against an AI lab over the suicides of young people who had long conversations with its chatbot.
The "leave them alone" strategy, which lets these companies launch products however they see fit, may be an unwise approach to AI safety from society's perspective.
Still, Huang may be right that existing product liability laws could extend to AI as well, that is, if AI doesn't somehow wipe us all out before enough cases reach the courts to test that theory.
Huang didn't address the other path that seems to be taking shape: industry self-regulation.
Huang's approach has instead focused more on supporting open-weight models and encouraging companies to use them as a competitive counterbalance to AI labs that rely on closed, proprietary models.
But the industry currently has a narrow window of time to establish a system of self-regulation, and to encourage AI labs around the world, including those in China, to recognize the value of taking part in it.
As Microsoft CEO Satya Nadella put it at the All-In Summit on Monday: "China should also care deeply about the same safety concerns the United States cares about, right? Why should it be any different for them? It's not as if they won't face the same problem with breaches. And it's not as if they don't want to make sure their citizens benefit from AI, just as we want our citizens to benefit from AI."
For now, though, if Huang opposes any new AI regulation, his influence may be significant enough for things to go the way he wants.
He also showed this week that he has, in something close to a literal sense, the ear of President Donald Trump.
Resource: Al-Ghad.
