The Oracle at Delphi was the high priestess at the Temple of Apollo in Ancient Greece [1]. People would travel to the temple and ask the Oracle for guidance, which she would give in the form of prophecies.

Said prophecies were believed to be uttered under the divine possession of the God Apollo. It is estimated that seeking guidance from the Oracle at Delphi started around 800 BC and continued until around 400 AD.

Written accounts are scare but what does exist states that the Oracle sat on a golden perch above an opening in the ground, connected with the divine and gave her prophecies while in a sort of trance. Modern scientific research has shown that this connection with the divine was in fact, nothing more than an induced hallucinogenic state created by gas emissions rising from a geologic chasm under the temple.

It is also important to note that it wasn’t free to receive a prophecy from the Oracle. Payment was required, and the more one paid, the shorter the queue. With 21st century hindsight it is easy to view the Oracle at Delphi as a harmless scam setup to fool gullible people out of money. Only, it wasn’t exactly harmless.

For example, in 431 BC, the Spartans asked the Oracle at Delphi weather or not it would be wise to go to war with the Athenians [2]. The Oracle’s answer was, “if they fought with all their might, victory would be theirs, and that he himself [Apollo] would be on their side”.

In short, they not only got a resounding thumbs up but also moral justification for their actions, as this was something that would please their God.

The Spartans eventually won what became known as the Second Peloponnesian War (albeit with the help of the Persian Empire). This little fracas not only replaced the Athenian Empire with the Spartan Empire as the regional power of Ancient Greece but in doing so changed the lives of countless people living in Greece at the time.

Similarly in 359 BC, Philip II of Macedon paid a visit to the Oracle at Delhi [3]. Philip was said to have a magnificent black colt, and the Oracle prophesised that “whoever could ride this horse would conquer the world”. Countless men, including great generals and even Philip himself tried and failed to ride the fabled horse. However, one person did succeed in mounting the steed, Philip’s son, Alexander. Also known as Alexander the Great, and Alexander did indeed go on to conquer much of the known world of antiquity.

Would the Spartans have gone to war if the Oracle at Delphi hadn’t given them the confidence boost that comes with a divine seal of approval? Would Alexander of Macedon have become Alexander the Great if his fate hadn’t been “foretold” by the Gods?

The prophecies uttered by the Oracle at Delphi were utter nonsense, but that didn’t matter. People believed them to be true and belief is a powerful driver of human behaviour. The result was that said prophecies were directly responsible for wars, conquests, bloodshed and death for a period spanning 1,200 years.

In truth AI is closer to a pandora’s box than it is to an Oracle

To bring it back to modernity, I was at a conference last week and during lunch, a conversation arose about the slowdown in investment within the Technology Sector.

One of our party joked, “Well, ChatGPT says there’s a 60% chance of a recession next year”.

Everyone at our table laughed, because such a prediction, from a system which, to quote Emily Bender, professor of computational linguistics at the University of Washington, is nothing more than a “text-manipulation system” [4], is utter nonsense.

However, this does raise a worrying issue and ultimately leads me to the point of this article.

In popular media, AI has been framed as an almost divine entity which can solve all of humanities problems. From mitigating the effects of climate change [5], to building cities of the future [6], to mining asteroids [7], AI can, and apparently will, do it all.

Only it can’t. And, it won’t [8].

But does this lack of ability matter? The Oracle at Delphi didn’t have any divine powers. However, people believed she did, and that mistaken belief was the driver behind human behaviour which had very real consequences. If people believe that AI is indeed a path to their salvation, this belief alone could be the driver behind some truly disastrous human behaviour that has wide ranging repercussions.

For example, why would you fund a glacier research project when you can fund an AI initiative that hopes to solve the entire climate crisis? Why fund a single healthcare clinic in sub-Saharan Africa when you can invest in an AI startup that promises to abolish global healthcare inequality? Why spend years studying microbiology to help develop vaccines for novel 21st century diseases when all disease will be eradicated by AI in the next ten years anyway?

I’m not saying that if enough people believe that AI will somehow solve all of our problems that humans will stop trying to solve problems themselves. However, there is a very real possibility that such a belief could direct funding, attention and effort away from areas where it is most needed.

Unfortunately, it would appear that this shift has already started. Private equity funding for AI companies was at an all-time high in 2023, and global venture capital investment is expected to reach $12 billion in 2024 [10]. It has been reported that an estimated 25% of companies in the US and 58% of companies in China have already funded AI initiatives [11]. It is important to remember that this behaviour is fuelled by one thing, belief; blind as it may be.

It has been hard work being a human over the last couple of years. From pandemics, to wars, to mass layoffs, to housing crises, to heatwaves, to floods. We have indeed lived through interesting times. During times of uncertainty, it is only natural for people to turn to the divine or appeal to a higher authority that has answers. Unfortunately, AI is no more divine than my laundry hamper and has about as much authority.

In truth AI is closer to a pandora’s box than it is to an Oracle but not because someone is going to create Skynet and robots with Austrian accents are going to start rounding people up. It is because the belief that AI can save humanity might just be end up having the complete opposite outcome for our species.

References