After making a killing with AI, are we at the end times?

You probably interact with artificial intelligence (AI) a lot without noticing it. On the 4th of every month, my bank calls to remind me that my mortgage payment is due on the 5th, using what passes quite well for a human contact centre agent. This annoys me no end, as I’m regular with my payments, but this is just one simple example of AI for you.

(In return, I now annoy the bank by repeatedly asking the caller ‘Are you a human or AI’? Do that long enough and it’ll admit it’s an automated system, at which point I start using words I don’t use in polite company.)

Dear reader, I’m not an expert in AI. Rest assured this piece is written by human AI (Adzhar Ibrahim) and not by computer AI (Artificial Intelligence). But I’m fascinated by AI’s impact on our lives, even as I believe predictions about us living long and happy lives because of AI are just dreams that’ll inevitably be dashed by a tsunami of unintended consequences.

One of my biggest fears is of AI being harnessed to control society or help us kill each other more efficiently. The killing part is already happening. Russia, Ukraine, the US, Iran etc are using increasingly smart AI-powered weapons to blast each other to smithereens. So, too, is Israel in their destruction of Gaza.

Threats to humanity

Expect more of such uses, as well as an increasingly obsessive desire to spy on us. Our propensity to harm each other because of ethnicity, religion, culture, geography, history etc is driving it, as is our greed to sell yet another product to increasingly impoverished customers. Expect no let-up here.

People did once think that nuclear weapons would also destroy the human race. They haven’t…yet. But in spite of all the safeguards we’ve built, including political organisations such as the United Nations and guarantees of “mutually assured destruction”, we feel even more vulnerable today.

The most likely way AI will destroy humanity is through the pervasive and increasingly autonomous AI that have become powerful because they’ve learnt how to learn and improve on their own, through what is called recursive self-improvement. The world could still go sploot and the future predicted in the Matrix movies of the 1990s could yet become reality, whether because of a glitch, a hack, or a deliberate decision to let these increasingly-sentient AI tools into the nuclear arsenals or biological labs,

Call to slow down

Recently, the people in control of the massive western AI corporations Anthropic, OpenAI, DeepMind, SpaceXAI etc have started warning us that AI could become feral and go rogue within months. They’re literally saying we should slow things down.

Ahh, so now they seem to care! What happened to the “Do No Evil” ethos of the billionaire tech bros in the past? That didn’t survive the battles with profits and valuations. In today’s political climate nobody even pretends to care about society and our shared existence on this planet. In fact, many are already building rockets to leave for other planets.

Tech moguls asking for more regulation is by itself unusual, especially after they helped destroy so much of our old guardrails and regulations through the politicians they’ve bought. It’s so out-of-character that it’s fair to assume they aren’t telling us the full story of what’s worrying them.

The first likely reason for this attack of conscience is they’re running out of money, having spent literally trillions on AI tools that’ll never generate enough revenue for a pay-back. They’re also getting increasingly spooked watching China push the AI race with funds and resources no capitalistic model can ever match.

Hot potato

Ironically, they’re asking for more guardrails and regulations while the governments that regulate them, especially the US under Donald Trump, have increasingly been dismantling such regulations. The US fear of being left behind in AI by China is driving this irrational, reckless and frightening approach.

Closer to our lives are data centres built to power the ever-hungry AI and, for now anyway, are the only thing keeping the US out of a recession. But even over there they’re now becoming a political hot potato as concerns about environmental damage, rising energy costs and job losses begin to surface.

That the world will face a recession soon is a matter of when rather than if. So much of the spending nowadays are on wars as well as the AI tools needed to wage such wars, with very little spent to cure society’s real ailments – healthcare, wealth disparity, inflation, climate change, job creations etc. At some point soon, things will start to break.

Looming turmoil

Unsurprisingly weapons and wars also tend to destroy the economy. Data centres don’t create jobs. Elsewhere jobs are being replaced by AI, often based on the hype of whatever improvement AI can deliver. AI itself hardly creates any jobs directly and often destroy jobs indirectly through either wars or business disruptions that enrich the few but impoverish the masses.

It’s tempting to say its your tough luck if you aren’t smart enough to make a pile of money through AI. But that tough luck is hurting the majority, and the majority don’t like their leaders ignoring their plight. This could lead to such a turmoil that earlier versions of them such as the French and Russian Revolutions or the Great Depressions will be child’s play in comparison.

As to whether these will actually come to pass, or some other calamities perhaps directly driven by AI such as extinction events by nuclear holocaust or biological warfare, even AI itself, both I and the fancy computers, are unfortunately not smart enough to predict.

The views expressed are those of the writer and do not necessarily reflect those of FMT.

Subscribe to our newsletter and get news delivered to your mailbox.