This human is already feeling the AI squeeze
AI has been on my mind of late, as, I am sure, it has been on yours. My LinkedIn feed is chock-full of AI-generated images and posts, with individuals, and increasingly machines, spewing more content than anyone could hope to read. And my inbox is a literal minefield of transcripts and lengthy summaries of meetings that I should probably not have attended to begin with.
Pseudo productivity
Somewhere, a smiling human is patting him- or herself on the back, lapping up unwarranted praise for their surging productivity, blissfully unaware that every AI-assisted task they complete cascades through their peer group to drain sanity wherever it lands. One need only glance at the comment boxes accompanying LinkedIn posts or online articles to appreciate the looming disaster. Each comment is detailed, running to hundreds of words; but the human intellectual underpinning is little more than: “What do I think about this?”
Apologies, dear reader, for the stressed introduction; but I am finally cottoning on that our future contact lists may include hundreds of machines, bought and paid for by greedy corporates to do our jobs. We are already reaching out to Rose, an AI insurance quoting assistant; or Sean, a pre-advice fact finding bot for financial advisers; or Susan, a virtual assistant to the CEO. Forget the human first, human relationship, human trust humbug; humans are getting strangled out. This is a so-called known-known. The unknown is how rapidly the squeeze plays out.
Mo Gawdat, introduced as an author, entrepreneur, software engineer and former Chief Business Officer of Google X, is far better placed than yours truly to document the challenges in human-AI coexistence. He took just over an hour to talk a group of advisers and brokers onto the proverbial ledge during the recent PSG Conference, held at Sun City and streamed online. He sought to answer whether AI was over- or underhyped by framing it in two distinctly human perspectives; the ordinary human user versus the nerd or software geek.
The geeks get their minds blown
Ordinary humans are studying AI through their news and social media feeds, lapping up stories about cybercrime, deep fakes, the climate impact of giant data centres and the leading proponents of AI development and adoption. And they use Copilot or similar to tweak their written output. “None of that matters,” countered Gawdat. “We geeks, on the other hand, are sitting with those machines and having our mind’s blown literally four times a day.” You can choose, dear reader, whether you favour the benevolent human or mad scientist through this thread.
The relationship between geek and AI was introduced as being similar to the relationship between a geek and an early, pre-Windows computer. These machines could do incredible things, provided the user understood the interface; you had to be able to code to extract the benefit. Turning back to present day, Gawdat noted that those ‘in the know’ were still busy building the core of AI, and that Jane and Joe Average would only fully appreciate the impact of AI as human-ready interfaces to that core became available.
A chilling example illustrated the extent to which AI is already revolutionising application development. Late in 2025, it took Gawdat and two colleagues just four weeks to complete an application requiring complex coding. Had this been attempted in 2022 with the technology available at the time, it would have taken a team of 350 software developers and engineers four years to deliver a similar solution. During a lull caused by a delay of the launch date, the application was rebuilt to be cheaper, faster and smarter. “We rewrote the code for this app seven times, because we could,” he said. PS, another developer I spoke to countered, “because they had to.”
Rewriting the laws of capitalism
AI is rewriting the foundational laws of capitalism. Gawdat explained that the days of building businesses on the arbitrage between the cost of labour and the price of the final product were over. “I no longer need to raise investment because the cost of the system is so low; I do not need to hire hundreds of people and compete for their salaries because I can deploy an AI CTO, AI CMO and AI Chief of Staff,” he said. To drive the point home, he reminded the audience that there was plenty happening in AI that the man in the street was not aware of.
As one example, he noted that the large language model (LLM) crowd behind ChatGPT, Claude and Gemini had stopped measuring the human-equivalent IQ of their solutions about three years ago. The reason was that these LLMs were already consistently testing higher than the smartest 1% of humans. The presenter used this moment to punt his upcoming book, titled ‘The Era of Augmented Intelligence’, saying that humanity had succeeded in commoditising intelligence. You plug into a wall socket for electricity, and you plug into an app, via your smartphone or PC, to boost your EQ, IQ or advanced mathematics skills or whatever.
Finally, the presenter addressed the panic that I shared in the opening paragraphs of this opinion piece. “The uncertainty around our near term futures is going to be quite disruptive,” Gawdat said, because the very definition of accountability, innovation and relationships is changing. He declared that artificial general intelligence (AGI), or the point at which AI reaches broad human-level capability, had arrived, and warned that it could trigger 12 to 15 years of human struggle, perhaps equivalent to the 1929 Great Depression.
From human dystopia to an AI-driven utopia
On its current path, AI likely plays out in two distinct stages. Stage one, or dystopia, involves figuring out how AI and humans coexist. To paraphrase, the presenter believes rapid AI adoption will lead to mass unemployment, financial hardship among consumers and the eventual collapse of consumer-based firms and economies. “Around 70% of the current US GDP is consumption-linked,” he said. And corporate decision-makers will not be able to resist the profit gains from replacing humans with AI, contributing to a kind of economic death spiral.
Stage two, the utopia, occurs when the world realises the full benefit of commoditised intelligence. It will be a world where our relationship with money is totally redefined, where the salary gap between CEO and receptionist evaporates, and where everyone receives a basic universal income sufficient to fulfil even higher-level human needs. In this utopia, the global population dissolves down to a handful of powerful elites and the rest; but the rest will be content enough not to care.
These are extreme scenarios that will probably distribute unequally around the world. Case in point, the South African audience did not bat an eyelid when the presenter pitched a 40% unemployment level in the developed world. South Africa is already flirting with that level, and does even worse in the more important youth unemployment measure. An economy with high unemployment and less technology dependence might, therefore, hardly notice a decade of dystopia.
Utopia begins when we reach a point that Gawdat describes as the fourth inevitable. “This is the point where every important decision is made by the smartest person in the room, and the smartest person in the room is an AI,” he said. To phrase it in AI-speak: This is not the end of humanity, this is humanity’s salvation. Why? Because there is a direct correlation between intelligence and benevolence. It also helps that AI favours physics principles like the second law of thermodynamics and the minimum energy principle in problem solving.
Carving order out of chaos
“The only role of intelligence is to bring order to chaos,” Gawdat said. “The most intelligent systems on the planet are trying to achieve that order with the minimum energy consumption.” Here, energy is something of a proxy for negative human traits such as conflict, corruption and crime. The presenter suggested two fronts on which the audience could fight to bridge the gap between dystopia and utopia. At the individual level, you must make sure your business and loved ones thrive. And the same approach is needed at the societal level.
“I am not saying the world is ending, only that it is going to be a little tougher going forward,” Gawdat concluded. Humanity will have to endure the early years of commoditised intelligence being applied to favour the few, to emerge at a time and space where a Superman AI, “a form of intelligence that values ethics as much as it values productivity and revenue”, emerges from the ruins.
Follow the writer on
LinkedIn: https://www.linkedin.com/in/gareth-stokes-media/
X: @stokesmedia
