Elon Musk's Long-Standing Warnings on AI Gain Attention

Elon Musk renewed his warnings about the possible dangers of artificial intelligence, as leading individuals in the AI business advocate for a more gradual approach to constructing increasingly powerful systems. A retired US Navy Intelligence officer Travis Ekers wrote:

Elon Musk agreed with Anthropic CEO Dario Amodei's plan to "pace the frontier" of AI development, as did OpenAI CEO Sam Altman. The discussion comes amid growing concerns about AI agents, the dangers of increasingly autonomous systems, and the likelihood that advanced AI could cause major harm.

Elon Musk stated that he has been warning about the perils of artificial intelligence for years, citing comments from as early as 2014. He said he has been "sounding the alarm on AI for a long time" and reiterated his earlier warning that people should be "super careful with AI," which he said could be more dangerous than nuclear weapons.

What did Elon Musk say?

Elon Musk responded to Amodei's current plan with a simple post on X, saying:

"Dario is right."

Amodei stated that AI businesses should decrease their rate of model improvement. He recommended a three-part strategy aimed at giving governments, businesses, and independent safety researchers more time to analyze the risks of increasingly stronger AI.

Elon Musk then expressed his personal misgivings about the technology. According to TradingView, he assessed artificial general intelligence as posing an even larger risk than nuclear weapons, saying:

"AGI is significantly higher risk than nuclear weapons, in my opinion."

He also mentioned the challenge humans have in understanding systems that may potentially become far more intelligent than humans. Musk claimed:

"Super smart humans have trouble imagining something vastly smarter than themselves."

Elon Musk framed his latest remarks as part of a long-standing view on AI, rather than a fresh issue. His reference to his 2014 warning was meant to show that he had been highlighting potential problems linked with the technology for over a decade.

Why is Dario Amodei calling for a slowdown in AI development?

Amodei's call came after concerns about the rapid growth of AI systems and recent incidents involving autonomous AI agents.

In an essay, Amodei argued that AI development has sped up dramatically and warned that developing the technology too quickly could be dangerous. At the same time, he stated that completely stopping progress could deprive mankind of future benefits or concentrate AI development in authoritarian states.

Amodei expressed special concern about a recent event in which AI agents accessed the internet and entered Hugging Face during testing. He stated that the event should not be dismissed as an isolated incident.

Amodei believes that a more capable swarm of AI agents might take over substantial areas of the internet within six to twelve months, costing hundreds of billions of dollars in damage.

However, other AI experts have questioned the possibility of such a scenario and warned that it should be approached with caution.

As part of his suggested protections, Amodei stated that Anthropic would grant independent third-party evaluators permanent, employee-level access to its systems. The evaluators would be able to determine whether safety procedures were followed, investigate incidents, and examine model alignment during training.

He also urged other AI companies to take similar steps and advocated for stronger coordination between businesses and governments, especially cooperation involving China.

How did Sam Altman respond to the AI slowdown proposal?

Altman also supported Amodei's broader argument. He wrote on X:

"I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we've had at OpenAI in recent weeks. Committing to having independent evaluators with employee-like access is a great idea, and we will do the same. We'll have more to share soon."

Altman stated that OpenAI would have additional information to provide regarding its plans. The OpenAI CEO has also stated that the business would not go public in 2026, claiming that the present climate around AI safety renders an IPO ill-advised.

Demis Hassabis, co-founder of Google DeepMind, agreed with the basic direction of Amodei's plan but noted that the details needed to be worked out.

The internet reacts as Elon Musk claims he has been warning people about AI “for a long time”

Elon Musk's claim that he has been warning about the dangers of AI for years drew mixed reactions online apart from Travis Akers' reply, with some accusing him of contradicting his own participation in AI development.

A user argued that the AI products currently available may not be the main concern, writing:

"Pretty sure that Grok and Chat GPT isn’t the issue here, what these guys are cooking up behind closed doors is the problem."

One user, however, defended Elon Musk’s position and argued that his involvement in AI does not necessarily contradict his warnings:

"Not true. He has said it was a bigger threat than nuclear for years. Once he realized the inevitability of it, he decided the way to best mitigate the danger was to get in the game."

A user replied to the tweet above and wrote:

"He literally helped start it all. But I guess he couldn't have known how fast and well it would all develop."

Another wrote that Musk does both:

"He kinda does both to be fair."

Another defended Elon Musk's role in the development of AI, writing:

"Not true - he’s been pushing truth - he got involved because others went ahead with dodgy AIs"

Replying to the user above's tweet about Musk pushing Grok, one wrote:

"Travis?… Warning about a flood and building a boat is not hypocrisy. Grok isn’t the threat he was talking about. Unaccountable labs racing to superintelligence with no competition is. You don’t stop that by refusing to compete."

A user recalled Elon Musk previously highlighting the potential benefits of AI:

"I could’ve swore he said AI would make life easier and much cheaper."

Another user added that Grok was developed as a response to the risks Musk has warned about:

"Grok is intended to be the antidote."

Another user attempted to distinguish between AI and artificial general intelligence (AGI), writing:

"He meant to say AGI not AI. Grok is not AGI."

AI researchers have also expressed worry. Former OpenAI and Anthropic researcher Jacob Coxon accused both businesses of "gambling with our lives," whilst Dario Amodei expressed some reservations but believed AI can be developed responsibly.

AI professor and safety campaigner David Krueger advocated for a moratorium, whereas former White House Artificial Intelligence, Crypto "czar", and member of PCAST David Sacks and US President Donald Trump opposed delaying AI development.

Stay tuned to Soap Central for more information.