Exponential Futures

Throughout history, humanity has navigated the interplay between population growth, technological progress, and ethical responsibility. As automation, artificial intelligence, and biotechnology advance at exponential rates, philosophers, scientists, and entrepreneurs have raised profound questions: Are we heading towards liberation from biological limits, or into a new form of dependency on machines? Can we satisfy our dissatisfaction with more intelligent machines and unlimited growth? What would be enough? The following post explores these dilemmas, drawing from historical parables, the logic of Moore’s law, transhumanism, and the latest breakthroughs in artificial intelligence.

“The current explosive growth in population has frighteningly coincided with the development of technology, which, due to automation, makes large parts of the population ‘superfluous’, even as labour. Because of nuclear energy, this double threat can be tackled with means beside which Hitler’s gas chambers look like the malicious child’s play of an evil brat.”
– Hannah Arendt

Originally published in Substack: https://substack.com/inbox/post/171630771

Our technological development has been tied to Moore’s law. Named after Gordon Moore, the founder of Intel, one of the world’s largest semiconductor manufacturers, the law states that the number of transistors on a microchip doubles every 18–24 months. As a result, chips become more powerful while their price falls. Moore’s prediction in 1965 has remained remarkably accurate, as innovation has kept the process alive long past the point when the laws of physics should have slowed it down. This type of growth is called exponential, characterised by slow initial development which suddenly accelerates at an unexpected pace.

A Parable of Exponential Growth

The Islamic scholar Ibn Khallikan described the logic of exponential growth in a tale from 1256. According to the story, chess originated in India during the 6th century. Its inventor travelled to Pataliputra and presented the game to the emperor. Impressed, the ruler offered him any reward. The inventor requested rice, calculated using the chessboard: one grain on the first square, two on the second, four on the third, doubling with each square.

Such exponential growth seems modest at first, but by the 64th square it yields more than 18 quintillion grains of rice, or about 1.4 trillion tonnes. By comparison, the world currently produces about 772 million tonnes of wheat annually. The inventor’s demand thus exceeded yearly wheat production by a factor of over 2,000. The crucial lesson lies not in the quantity but in the speed at which exponential processes accelerate.

The central question remains: at what stage of the chessboard are we today in terms of microchip development? According to Moore’s law, we are heading towards an increasingly technological future. Futurists such as Ray Kurzweil, Chief Engineer at Google, believe that transhumanism is the only viable path for humanity to collaborate with AI. Kurzweil predicts that artificial intelligence will surpass human mental capabilities by 2045.

Transhumanism posits that the limits of the human biological body are a matter of choice. For transhumanists, ageing should be voluntary, and cognitive capacities should lie within individual control. Kurzweil forecasts that by 2035 nanobots will be implanted in our brains to connect with neurons, upgrading both mental and physical abilities. The aim is to prevent humans from becoming inferior to machines, preserving self-determination.

The Intelligence of Machines – Real or Illusion?

Yet artificial intelligence has not, until recently, been very intelligent. Algorithms can process data and make deductions, but image recognition, for example, has long struggled with tasks a child could solve instantly. A child, even after seeing a school bus once, can intuitively identify it; an algorithm, trained on millions of images, may still fail under slightly altered conditions. This gap between human intuition and machine logic underscores the challenge.

Nevertheless, AI is evolving rapidly. Vast financial resources drive competition over the future of intelligence and power.

The South African-born Elon Musk, founder of Neuralink, has already demonstrated an implant that allows a monkey named Pager to play video games using only thought. Musk suggests such implants could treat depression, Alzheimer’s disease, and paralysis, and even restore sight to the blind.

Though such ideas may sound outlandish, history suggests that visionary predictions often materialise sooner than expected.

The Warnings of Tristan Harris

Tristan Harris, who leads the non-profit Centre for Humane Technology, has been at the heart of Silicon Valley’s AI story, from Apple internships to Instagram development and work at Google. In 2023, alongside Aza Raskin, he warned of AI’s dangers. Their presentation demonstrated AI systems capable of cloning a human voice within seconds, or reconstructing mental images using fMRI brain scans.

AI models have begun to exhibit unexpected abilities. A system trained in English suddenly understands Persian. ChatGPT, launched by OpenAI, has independently learned advanced chemistry, though it was never explicitly trained in the subject. Algorithms now self-improve, rewriting code to double its speed, creating new training data, and exhibiting exponential capability growth. Experts foresee improvements at double-exponential rates, represented on a graph as a near-vertical line surging upwards.

Conclusion

The trajectory of human civilisation now intertwines with exponential technological growth. From the rice-on-the-chessboard parable to Moore’s law and the visions of Kurzweil, Musk, and Harris, the central issue remains: will humanity adapt, or will machines redefine what it means to be human? The pace of change is no longer linear, and as history shows, exponential processes accelerate suddenly, leaving little time to adjust.

---

References

Arendt, H. (1963). Eichmann in Jerusalem: A report on the banality of evil. Viking Press.
Harris, T., & Raskin, A. (2023). The AI dilemma [Presentation]. Center for Humane Technology.
Kurzweil, R. (2005). The singularity is near: When humans transcend biology. Viking.
Moore, G. E. (1965). Cramming more components onto integrated circuits. Electronics, 38(8).