AI set to wipe out humans? Andrew Ng breaks his silence on artificial intelligence doomsday warnings as tech leaders debate whether superintelligent machines could become uncontrollable

One of the early pioneers of artificial intelligence, Andrew Ng, dismissed the dire predictions regarding the threat advanced forms of AI pose for the survival of humanity, terming them "much more science fiction than science".

As reported in Bloomberg News, this statement was made against the backdrop of warnings raised by researchers at leading AI firms about the risks associated with the fast evolution of AI technology and the likelihood of its becoming unmanageable.

Also Read | A future without oxygen: Scientists explain what Earth could look like in a billion years - hostile to complex life long before the Sun dies out


In his views, Ng, who co-founded Google Brain and Coursera, says that the industry should focus more on real-world issues of safety of AI technology than getting preoccupied with talk about extinction of humans because of it.

These comments came in response to a series of public warnings made by the researchers working at leading AI labs such as Anthropic, OpenAI, and Google DeepMind.

Although Ng did not dispute the dangers posed by artificial intelligence, he raised doubts about the credibility of some of the worst-case scenarios in terms of evidence-based science.

Andrew Ng challenges the latest AI extinction warnings

Ng’s comments came during an interview with Bloomberg TV on Thursday, as reported by Bloomberg News.

The AI pioneer said he was disappointed by what he described as a renewed wave of publicity surrounding catastrophic AI risks.

“I was quite dismayed over the past two weeks,” he said. “This wave of PR has kicked up again, for probably similar purposes.”

Also Read | Scientists find a smarter way to make drinking water from seawater — and the leftover salt could help supply lithium for future electric car batteries

Ng was referring to concerns that the technology industry had previously amplified fears about AI’s potential harms during the earlier stages of the AI boom. He suggested that such warnings may have been used to generate publicity and influence regulation.

Some researchers argue that future systems could become powerful enough to create risks beyond the ability of humans to manage them. Others believe the focus on hypothetical extinction scenarios is distracting from problems that can already be identified and addressed.

Ng’s position is that AI risks should be assessed carefully, with greater attention paid to practical engineering challenges.

He described human extinction fears as “much more science fiction than science”, according to Bloomberg News.

That does not mean he considers AI harmless. Instead, he draws a distinction between foreseeable risks and predictions about the distant future.

Former OpenAI and Anthropic researcher raises fresh concerns

The renewed debate follows the resignation of Jacob Coxon, a former researcher at Anthropic and OpenAI.

Coxon accused his former employers of “gambling with our lives” by racing towards superintelligent AI. He also said people building advanced AI believe the technology could “kill us all by the end of the decade”.

His comments brought renewed attention to concerns about the development of increasingly capable AI systems.

Another researcher who recently resigned from Google DeepMind also warned that humanity may be running out of time to prevent AI from causing widespread harm.

These warnings have emerged alongside calls from AI leaders for more caution in developing cutting-edge models.

Anthropic chief executive Dario Amodei and OpenAI chief executive Sam Altman have both called for additional safeguards and a more measured approach to AI development, according to Bloomberg News.

Ng previously worked with both Amodei and Altman earlier in their careers.

His disagreement with the latest warnings is therefore notable because it comes from someone with experience in the development of modern AI research.

The dispute is not about whether AI can create risks. It concerns how severe those risks may become, how much evidence supports predictions of catastrophe and which safety measures deserve priority.

Why Ng wants the industry to focus on practical AI risks

Ng said there are “practical engineering problems” that the AI industry can work on while continuing to improve the technology.

Cybersecurity was one area he identified as a genuine concern.

AI systems can assist with complex technical tasks, and their capabilities may also create new challenges for protecting digital infrastructure. The potential misuse of AI in cyberattacks is therefore a more immediate area of concern than predictions about human extinction.

Ng’s argument is that engineers and researchers should focus on identifying vulnerabilities, testing systems and developing safeguards.

He also said AI should be used for “many of the wonderful things it can do”.

That approach places emphasis on the benefits of artificial intelligence, while recognising that its development must be accompanied by appropriate safety work.

The distinction matters because AI risks are not all of the same kind.

Some involve current systems, including misinformation, privacy problems, unreliable outputs and cybersecurity threats. Others concern future systems that may become substantially more capable than today’s models.

Ng’s remarks focus on the need to separate those questions rather than treating every possible danger as equally established.

Could AI development lead to a dangerous race?

The dispute between Ng and researchers warning of AI extinction reveals the existing divide in the technology sector.

The major AI firms are engaged in a race for developing ever more sophisticated AI models; researchers and corporate managers discuss whether the progress can be made faster.

While some experts think that highly sophisticated systems will pose certain threats that will not be addressed by current safety measures, others suggest that speculation is premature and that the effort should be dedicated to real issues.

The problem is now linked to regulatory concerns.

Should the regulators pay too much attention to the potential future threats, they will overlook the problems that the companies and the consumers have already faced. However, the researchers who are concerned about advanced AI think that postponing action until the system becomes more sophisticated will complicate the issue.

According to Bloomberg News, Ng thinks that the industry should keep on solving engineering problems while delivering the benefits of AI to the society.

The debate remains unsettled.

Researchers continue to disagree about the likelihood and severity of existential AI risks, as well as the best methods for preventing them.

Ng’s position is clear: AI should be developed with safeguards, but predictions of human extinction should not overshadow the practical work needed to make the technology useful and secure.

FAQ

1. Who is Andrew Ng?

Andrew Ng is an artificial intelligence expert and entrepreneur who co-created Google Brain and founded Coursera.

2. How did Andrew Ng comment about AI extinction theories?

According to Bloomberg News, Ng dismissed AI extinction predictions as "much more science fiction than science".

3. Why are AI researchers alarmed by superintelligent AI?

Jacob Coxon, a former researcher at Anthropic and OpenAI organizations, recently issued an alarm over the race for superintelligent AI.

4. Does Andrew Ng think that AI has no risks at all?

No. Ng recognized the existence of AI-related risks, such as cybersecurity ones. His commentaries were related to the issue of extreme extinction predictions.