Kissinger is correct – AI tech does badly need a guiding philosophy

While tools such as ChatGPT have potential to improve lives, they also pose a risk of military artificial intelligence arms race and other perils

As a diplomat, former US secretary of state Henry Kissinger witnessed and was personally involved in many turning points in history. But in recent years, the famous statesman, who turned 100 over the weekend, has focused his attention on the futuristic problems posed by artificial intelligence.

In his 2021 book The Age of AI, co-written with former Google CEO Eric Schmidt and computer scientist Daniel Huttenlocher, Kissinger pointed out the unique challenges of the AI era compared to previous stages of human history, such as the Enlightenment or the more recent internet age.

For instance, AI is inherently unpredictable as it improves itself through data and algorithms, and it can make decisions based on vast amounts of information at speeds the human mind cannot comprehend.

According to Kissinger, that means decisions made by AI – and the rationale behind them – may at some point become a mystery to humankind. The technology will have an enormous influence on how we perceive reality, giving it the potential to become one of the most important shapers of human perception, much like reason and religion.

The book also warned about a potential AI arms race and called for talks between governments to limit the proliferation of military AI.

While some may dismiss these concerns as paranoia, many others are gaining awareness of the potential pitfalls of the technology thanks to ChatGPT, the popular AI tool that has grabbed headlines since its launch late last year.

Tools such as ChatGPT can generate text that is so convincing they worry experts including Pentagon AI chief Craig Martell, who said in early May that he was “scared to death” that bad actors could use them for misinformation.

Kissinger is not alone in his reservations about artificial intelligence. Computer scientist Geoffrey Hinton, sometimes called “the godfather of AI”, has warned of the dangers of the technology, including its potential to spread misinformation and the possibility it could exceed humans in intelligence. He quit his position at Google recently so he could speak more freely on the technology.

Not all leading AI scientists are so pessimistic. Another “founding father” of AI, computer scientist John McCarthy, has said the technology should be embraced for its potential to improve people’s well-being.

While the technology will bring enormous benefits in many areas, minimising the risks related to AI has become a major concern for governments, academics, scientists, businesses and the general public.

Unfortunately, it will be difficult to slow down the AI arms race with talks in the same way the nuclear arms race has been stalled. Unlike nuclear weapons, which are physical weapons controlled by states operating under strict protocols, AI is already integrated into many facets of everyday life and accessible to many actors, including governments, corporations, scientists and ordinary people. The deep mistrust between China and the United States also makes it difficult for any party to slow down the pace of AI research and development.

In an online article about cognitive warfare, the US Naval Institute accused Beijing of launching cyberattacks and disinformation campaigns regarding Taiwan – an accusation Beijing has adamantly denied.

Beijing has similar fears and has repeatedly stressed the importance of protecting China from potential cognitive warfare tactics by the US.

Such fears would only be exacerbated by the availability of powerful ChatGPT-like language tools.

Fear and complacency are both undesirable in the AI age. As Kissinger rightly points out, we have developed a technology in dire need of a guiding philosophy.

While Kissinger may not live to see the full potential of the AI age, many will look back and appreciate his foresight about the risks of the technology. His concerns about ethics and governance and philosophical questions about human identity and reality should not be dismissed as “old school”. They have never been more important.

The book also warned about a potential AI arms race and called for talks … to limit the proliferation of military AI