Artificial intelligence – the panacea to all ills, or an existential threat to our world?

Daily Maverick

7 hours ago

'Once men turned their thinking over to the machines in the hope that this would set them free. But that only permitted other men with machines to enslave them.' – Frank Herbert, Dune, 1965
In the early 19th century, a group of disgruntled factory workers in industrial England began protesting against the introduction of mechanised looms and knitting frames into the factories.
Fearful of losing their jobs, they smashed machines and engaged in acts of sabotage. They were dealt with harshly through imprisonment and even execution. They became known as the Luddites.
At the time, it was not the technology they were most concerned about, but rather the loss of their livelihoods. Ironically, today, the word Luddite has become something of an accusation, a complaint about those who, because they are seen as not understanding a new technology, are deemed to be anti-technology. Even anti-progress.
The 2020s have seen rapid progress in the development of a 'new' technology – artificial intelligence (AI). But the history of AI can be traced back to the middle of the 20th century, and so is perhaps not very new at all.
At the forefront of the current process has been the release of Large Language Models (LLMs) – with ChatGPT being the most prominent – that allow, at the click of a single request, an essay on the topic of your choice. LLMs are simply one type of AI and are not the same as artificial general intelligence (AGI).
Unlike current LLMs, which perform a single task, AGI would be able to reason, be creative and use knowledge across many domains – be more human-like, in essence. AGI is more of a goal, an end point in the development of AI.
LLMs have already been hugely disruptive in education, with university lecturers and school teachers scrambling to deal with ChatGPT-produced essays.
Views about the dangers of AI/AGI tend to coalesce into the doomer and the boomer poles. Crudely, and I am oversimplifying here, the 'doomers' worry that we face an existential threat to our existence were AI to be designed in a way that is misaligned with human values. Boomers, on the other hand, believe AI will solve all our problems and usher in an age of abundance, where we will all be able to work less without seeing a drop in our quality of life.
The 'doomer' narrative originates with Oxford University philosopher Nick Bostrom, who introduced a thought experiment called the ' paperclip maximiser '. Bostrom imagines a worst-case scenario where we create an all-powerful AGI agent that is misaligned with our values.
In the scenario, we request the AGI agent to maximise the production of paperclips. Bostrom worries that the command could be taken literally, with the AGI agent consuming every last resource on Earth (including humans) in its quest to maximise the production of paperclips.
Another take on this thought experiment is to imagine that we ask an all-powerful AGI agent to solve the climate breakdown problem. The quickest and most rational way of doing this would, of course, be to simply rid planet Earth of eight billion human beings.
What do we have to fear from LLMs?
LLMs have scraped the internet for every bit of data, stolen the data, and fed off the intellectual property of writers and artists. But what exactly do we have to fear from LLMs? I would suggest very little (unless, of course, you are a university lecturer in the humanities).
LLMs such as ChatGPT are (currently) little more than complex statistical programs that predict what word follows the word before, based on the above-mentioned internet scraping. They are not thinking.
In fact, some people have argued that everything they do is a hallucination. It is just that the hallucination is more often than not correct and appropriate.
Francois Chollet, a prominent AI researcher, has described LLMs in their current form as a ' dead end ' in the quest for AGI. Chollet is so confident of this that he has put up a $1-million prize for any AI system that can achieve even basic human skills in something he calls the abstraction and reasoning corpus (ARC) test.
Essentially, the ARC is a test of what is called fluid intelligence (reasoning, solving novel problems, and adaptation). Young children do well on ARC tasks. Most adults complete all tasks. Pure LLMs achieve around 0%. Yes – 0%. The $1-million prize does not even require that AGI systems match the skills of humans. Just that they achieve 85%. The prize is yet to be claimed.
People are the problem
If LLMs are (currently) a dead end in the quest for AGI, what should we be worried about? As is always the case, what we need to be afraid of is people. The people in control of this technology. The billionaires, the tech bros, and the dystopian conspiracy theorists.
High on my list is Mark Zuckerberg. The man who invented Facebook to rate the attractiveness of college women, and whose company profited enormously from the echo chamber it created. In Myanmar, this resulted in the ethnic cleansing of the Rohingya people in 2017.
At the beginning of 2025, Zuckerberg showed the depth of his commitment to diversity and integrity in his slavering capitulation to Donald Trump. Jokes aside about whether Zuckerberg is actually a robot, in recent pronouncements, what he seems to want is a world of atomised and alienated people, who out of quiet desperation turn to his dystopian hell where robots – under his control – will be trained to become 'our friends '.
And my personal favourite – Elon Musk. Musk, the ketamine-fuelled racist apologist for the Great Replacement Theory. A man who has committed securities fraud, and accused an innocent man of being a paedophile because the man had the nerve and gall to (correctly) state that Musk's submarine could not negotiate an underwater cave in Thailand.
More recently, estimates are that Musk's destruction of USAid will lead to the deaths of about 1,650,000 people within a year because of cuts to HIV prevention and treatment, as well as 500,000 annual deaths due to cuts to vaccines.
I, for one, do not want this man anywhere near my children, my family, my community, my country.
OpenAI
Sam Altman, the CEO of the world's largest plagiarism machine, OpenAI, recently stated that he would like a large part of the world's electricity grid to run his LLM/AI models.
Karen Hao, in her recently published book Empire of AI, makes a strong case for OpenAI being a classic colonial power that closely resembles (for example) the British East India Company, founded in 1600 (and dissolved in 1874).
Altman recently moved squarely into Orwellian surveillance when OpenAI bought io, a product development company owned by Jonny Ive (designer of the iPhone). While the first product is a closely guarded secret, it is said to be a wearable device that will include cameras and microphones for environmental detection. Every word you speak, every sound you hear, and every image you see will be turned into data. Data for OpenAI.
Why might Altman want this? Money, of course. But for Altman and Silicon Valley, money is secondary to data, to surveillance and the way they are able to parlay data into power and control (and then money). He will take our data, further train his ChatGPT models with it, and in turn use this to better surveil us all.
And for the pleasure of working for, and giving our data to OpenAI? Far from being paid for the data you produce, you will have to buy the gadget, be monitored 24/7, and have your life commodified and sold.
As Shoshana Zuboff said in her magisterial book, The Age of Surveillance Capitalism, 'Forget the cliché that if it's free, 'you are the product'. You are not the product; you are the abandoned carcass. The 'product' derives from the surplus that is ripped from your life.'
The problem was never the cotton loom. The Luddites knew this in the 19th century. It was always about livelihood loss and people (the industrialists).
Bostrom has it badly wrong when he imagines an all-powerful AGI entity that turns against its human inventors. But about the paperclips, he might be correct.
Zuckerberg, Musk and Altman are our living and breathing paperclip maximisers. With their political masters, they will not flinch at turning us all into paperclips and sacrificing us on the altar of their infinite greed and desire for ever-increasing surveillance and control. DM