South African professor weighs in on current AI panic

This week, the risks surrounding rapid AI development have come under the spotlight, after Anthropic CEO, Dario Amodei, voiced his concerns about our ability to keep AI in check.

The CEO has since called for a co-ordinated, industry-wide slowing down of developments in AI, but it remains to be seen in the call will be heeded.

The interview comes after several companies have looked to scale back their spending in terms of AI, as token usage has grown out of control. Exacerbating matters are a series of inadvertent attacks, with agentic AI solutions have operated outside the parameters of their design and compromised critical software and systems of other organisations.

Naturally, the fact that some AI has seemingly gone rogue has brought into question the potentially greater risks that the pervasive technology poses, particularly in terms of the future of humanity.

Amid all the AI panic, Professor Anné Verhoef, director of the North-West University’s AI Hub, says the panic surrounding AI is premature.

“For the first time, we have experienced what appears to be a loss of control over AI, with agents seemingly acting independently. That uncertainty has understandably caused anxiety, but we should concentrate on verified facts rather than panic or conspiracy theories,” shared Verhoef in a release with Hypertext.

Here, he explained that the real danger lies in AI impacting essential services, and agents have the ability to access secure areas of the internet to devastating effect.

“It would not be humanoid robots running around killing people. However, if AI agents hacked and disabled essential systems, the consequences could be extremely damaging to human wellbeing and survival,” he pointed out.

Verhoef also noted that the limited number of incidents we have seen in recent months also provides us with an opportunity, as they highlight vulnerabilities, and the urgent need to address them. We have seen the likes of Anthropic try to tackle this, launching its Project Glasswing initiative earlier in the year to create a co-ordinated effort in terms of using AI to identify and prevent other AI models from exploiting vulnerabilities in critical software.

More than 10 000 vulnerabilities were found following the initiative’s first month, with Anthropic bringing on more organisations globally towards its cybersecurity efforts.

While the risk that AI poses cannot be ignored, Verhoef is still of the opinion that it presents far greater good. “AI is deeply ambiguous. It presents serious risks, but it has also brought enormous benefits and has the potential to enhance humanity. We should not disregard those achievements whenever alarming incidents occur,” he emphasised.

When it comes to everyday people using AI, Verhoef has advocated for a measured approach, and for users to think critically before making use of any model, chatbot, or agent. “We must retain our own knowledge and skills. Systems may fail or be shut down, and we need to remain capable of doing our work without AI,” he said.

The Professor was also quick to point out that even if Anthropic is able to get a group of AI companies to slow down, other companies would likely accelerate their efforts in order to gain a competitive advantage. As such, the toothpaste cannot be put back into the tube, and now it is up to everyone who uses AI to do so in a responsible way.

“If one country or company slows down, another is likely to accelerate. AI is about power and money, and everybody wants to dominate the field,” noted Verhoef.

“When an AI system causes harm, we cannot simply blame the technology. The people behind it must be held accountable,” he concluded.

[Image – Photo by Eduardo Ramos on Unsplash]

Let's talk tech!

Get the tech news you want to read. Take our reader survey and tell us how we can help you better.

Take the survey now