Your complimentary articles
You’ve read one of your four complimentary articles for this month.
You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please
Articles
Reality, Humanity & AI
Nafees Malik surveys our quest for truth, and asks if AI can help us discover more.
In 1637, René Descartes, the founder of modern philosophy, famously declared Cogito, ergo sum, or “I think, therefore I am”. Descartes was so sceptical about what, if anything, was real that he even questioned if he existed. He concluded that because he doubted his own existence he must be real, as he must exist in order to do the doubting.
Descartes perfectly exemplifies humanity’s innate desire to understand reality – to know the true nature of the world, from what is visible to the naked eye to what lies behind the veil. Some of our approaches to know reality have been objective, being built on rational arguments and evidence (e.g. philosophy and science), whilst others have been more subjective, grounded in beliefs, experiences, and emotions arising from religion, culture, art, and literature. Today, we stand at a historic moment: the beginnings of the artificial intelligence revolution. AI promises to reveal aspects of reality that humanity has until now not been able to discover or comprehend (and perhaps unaided, never could). Of all life on the blue planet, humanity has always held the exalted position of being the seeker, knower, and appreciator of reality. It may, however, soon lose this cherished position to AI.
I want to begin by considering how human beings have sought to understand reality throughout history. I’ll then examine how AI is beginning to uncover aspects of reality humans alone have not. Next, I’ll look at what might be the impact on humanity of AI discovering and shaping reality. Finally, I’ll explore how, eventually, AI may not only uncover reality, but also understand and reflect on it in a human-like way, if it becomes conscious.
Humanity Seeking Reality Throughout History
In the pre-classical era (before the eighth century BC), supernatural phenomena – spirits, demons, gods – played a central role in the human understanding of nature. Our hunter-gatherer ancestors were animists – believing that every living being, and some inanimate objects and natural phenomena, had a spirit. Agricultural societies began in about 10,000 BC. Here polytheism flourished, as farmers turned to gods to seek prosperity and protection for their crops and animals. In the classical world (eighth century BC to fifth century AD), when ancient Greek and then Roman civilisation thrived, philosophy became a primary means through which to understand reality, representing a shift away from supernatural explanations toward those founded on empirical and naturalistic reasoning. The first Greek philosopher, Thales of Miletus (c.624-546 BC) asked, what is the basic underlying substance of which everything is made. He concluded it was water. His successor Anaximenes (c.586-526 BC), considered it to be air; while Heraclitus (c.540-480 BC) believed it was fire. Later, in the fifth century BC, Leucippus and his pupil Democritus proposed that everything was made up of extremely small, solid, indivisible particles (atomos), making them the first atomists.
During the so-called Dark Ages (fifth to fifteenth century), with European society built on Christianity, once again reality was most fervently sought through religious disputation. Daily life revolved as much around the unseen world of God and a spiritual realm, as it did around the perceived physical world. There were, however, limited advances in the sciences, and many of those came from outside Europe.
During the European Renaissance (which means ‘the rebirth of knowledge’) in the fifteenth and sixteenth centuries, reality was understood once more through recently-acquired translations of Greek and Roman philosophy from the Arab world, as well as increasingly through the natural sciences and art and literature. Arguably, during this period, Copernicus had the greatest impact on revealing nature, when in 1543 he proposed that the Sun, rather than the Earth, lay at the centre of the universe. Subsequently, Galileo built a telescope, aimed it toward the heavens and verified Copernicus’s model. After publishing his work in 1632, Galileo was found guilty of heresy by the Church, and spent the rest of his life under house arrest. Proclaiming new truths has often been fraught with danger!
Next came the Enlightenment of the seventeenth and eighteenth centuries. This was a time in which long-standing beliefs, even about God and King, were not beyond questioning. Reason and scepticism, including the scientific method of observation and experimentation, became the gold standard by which to understand reality. The spirit of the age was: reason over superstition, and science over faith. Of great significance to uncovering reality was Newton’s formulation of the law of gravity in 1665, and his subsequent discovery of the laws of motion (published in his magnus opus Principia Mathematica in 1687). Aptly, Newton was born in the same year Galileo died, 1642.
Whilst Newton wrestled with the nature of the cosmos, other natural philosophers (as scientists were called back then) pondered living things here on Earth. Also in 1665, Hooke looked through his microscope and discovered the cell, revealing what makes up life itself. And using microscopes he constructed himself, in the mid-1670s Van Leeuwenhoek discovered the secret world of bacteria, which he called ‘animalcules’ (‘little animals’). On the philosophical front, in 1781, Kant proposed that we can only ever know reality as our mind perceives it to be, rather than how it actually is independent of our experience. Therefore, the reality we know will always be partly subjective, being perceived through our minds.
From the late eighteenth to the mid-nineteenth century, Romanticism prospered, in which reality was sought through the subjective lens of personal emotion, experience, and imagination. The motto here was: emotion over reason, and senses over intellect. Then, in the mid-nineteenth to the early twentieth century came Realism and then Naturalism. Both sought to represent life as it actually was – often brutally hard and unforgiving. In light of the success of the scientific method in revealing truths, Naturalism examined human nature from an objective and detached perspective, and considered life to be deterministic. Realistic and Naturalistic period depictions of reality contrasted to those in the Romantic period, as full of melodramatic, idealised, and fantastical elements. In science, in 1859, Darwin made a remarkable revelation when he published his theory of evolution by natural selection. This and ensuing discoveries – for instance, the principles of inheritance by Mendel (1865), the structure of DNA by Watson and Crick (1953), among many others – were steps towards deciphering the secrets of how life preserves and propagates.
Since the beginning of the twentieth century, two revolutionary theories in physics turned on their head our age-old notions of space, time, energy, light, and matter – the very building blocks of reality. The first is Einstein’s Theory of Relativity (published in two parts). Part one, Special Relativity, published in 1905, describes how time slows down, mass increases, and an object’s length shortens on approaching the speed of light. What’s more, energy and mass are different manifestations of the same thing, leading to the most famous equation of all time: E=mc2 . Part two of Einstein’s theory, General Relativity, published in 1915, explains the workings of gravity, superseding Newton’s explanation of it, to reveal how space-time warps, light bends, and time ticks more slowly around cosmic bodies – planets, stars, galaxies, black holes. The second revolution in physics, quantum mechanics, developed in the 1920s by Bohr as well as others, describes reality at atomic and subatomic scales. Its core principle, known as the Copenhagen interpretation, states that at these scales, unobserved matter exists in all its possible states simultaneously, and physical reality only takes shape upon being observed. Thus we learn with bewilderment that an electron can exist in multiple places at once, only taking a specific location upon observation.
During the Enlightenment, Newton’s clockwork universe was simple to imagine: the cosmos was like a clock wound up by God at the beginning of time, ticking away precisely and predictably ever since, as permitted by His laws of nature. In contrast, we now inhabit a universe that is neither intuitive nor possible to properly visualise, governed by quantum mechanics, where reality at its core is random, with particles popping in and out of existence unpredictably. And it’s shaped by Einstein’s Relativity, according to which time is not absolute but relative, being dependent on one’s frame of reference (that is, from where one is measuring it).
Aristotle began his Metaphysics with “All men by nature desire to know”, and history confirms that to be human is to seek to understand reality. How will we feel when AI joins us on our quest?
Image © Miles Walker 2024 Please visit mileswalker.com
AI Begins to Reveal Deeper Reality
Long before the birth of computers in the 1930s and 40s, writers imagined intelligent machines. Indeed, Greek mythology gave us our first humanoid robot: Talos was a giant robot made from bronze who protected the island of Crete from invaders. And the first novel to explore AI was Samuel Butler’s Erewhon in 1872, which tells of a utopia where machines are banned from a fear of them superseding humanity. In 1927, Metropolis became the first film to feature a robot as an evil replica of a human character. Perhaps it was prescient. AI is now beginning to reveal a reality that humanity has not been able to uncover.
Chess has a long tradition of human vs. machine. In 1997, IBM’s Deep Blue became the first computer to beat a reigning human world chess champion. And in 2017, DeepMind’s AI chess program AlphaZero was a landmark moment for AI. Informed of the basic rules of chess, but not given any preprogrammed chess strategies, yet it took only four hours to learn the game from scratch by playing against itself, and then went onto beat the reigning world champion chess program, Stockfish 8, which was programmed with all human chess knowhow. Remarkably, AlphaZero used strategies that had remained undiscovered by human beings in the game’s fifteen hundred year history, such as an unorthodox willingness to sacrifice high-value pieces early on.
AI can see patterns and relationships in huge quantities of data that humans cannot, and thus can observe a reality beyond our reach, and come to novel conclusions. This is how, in 2020, AI enabled Pfizer and its partner BioNTech to develop their Covid vaccine Comirnaty in record time (eight months). AI was involved in the vaccine design, manufacturing, clinical trials, and product distribution. For example, AI quickly detected meaningful signals from the background noise in a 44,000-person clinical trial for Comirnaty. But perhaps 2022 will stand out as a turning point in AI with the launch of OpenAI’s sensational chatbot ChatGPT, which was quickly celebrated for its unprecedented ability to hold human-like written conversations as well as write stories, essays, and computer code. Remarkably, it also passed the U.S. Medical Licensing Exam. ChatGPT powered by GPT-3.5 was followed by its successor GPT-4 in 2023. On a long list of academic and professional benchmarks, GPT-4 demonstrates human-level performance and beyond. For instance, on the Bar exam, needed to practice law, GPT-4 beat 90% of human candidates.
These examples show AI’s potential to reveal a deeper reality, even though AI does not understand reality in a human sense, being blind to the ethical, philosophical, societal, and political implications of its conclusions. Neither does AI possess the human qualities of emotion, common sense, or intuition. So should humanity be afraid that AI will reveal a reality that is unfathomable by us? After all, quantum physics is already abundant with aspects of reality discovered by us which are not fully comprehensible, or which can only be understood through mathematics. This has not deterred us from going ever further down that rabbit hole.
Quantum mechanics has exposed a fantastical reality. Wave-particle duality describes how a physical entity can be both a particle and a wave at the same time. Heisenberg’s uncertainty principle states we cannot simultaneously know both the position and speed of a particle (among other pairs of variables). The more precisely we know one, the less exactly we know the other. Q uantum entanglement describes how entangled subatomic particles are somehow connected even when separated by cosmic distances, so that a change in one instantly effects the other, seeming to violate Einstein’s Special Relativity conclusion that nothing can travel faster than light. As strange as these conclusions are, all our quantum mechanics experiments show them to be an accurate description of nature. So perhaps what we might fear from AI is not that reality as revealed by AI may be puzzling, but that reality will be discovered by something other than us. Our trepidation is only made worse knowing that AI cannot explain its logic for arriving at a particular conclusion to us, and so may present findings we cannot understand. As Kant said, the reality we perceive is dependent on what’s captured by our senses, crafted by our brains, biased by our beliefs and experiences, and restricted by our thought-power. AI may not be constrained by any of these parameters – at least not in the same ways – so may see a more authentic reality.
How might this push the boundaries of science?
There is still much we don’t scientifically understand. What happens at the centre of a black hole, where our scientific theories break down? What is dark matter and dark energy, which (allegedly) make up 95% of the cosmos? Indeed, what are the fundamental properties of the universe? As neither space nor time are believed to be fundamental anymore, emerging from a deeper reality. Perhaps most importantly, whilst General Relativity and quantum mechanics remain our best descriptions of nature, they are at odds with each other. General Relativity breaks down at the atomic level, whilst quantum mechanics fails at everyday scales and above. For over a century, physicists have tried unsuccessfully to reconcile them to produce a ‘Theory of Everything’ (everything physical, that is): a single, unifying equation, elegant and beautiful, that would provide a complete description of physical reality. So can AI cast light on these great unsolved mysteries of nature? Could it ignite a new era of intense scientific discovery, akin to the scientific revolution of the sixteenth and seventeenth centuries?
Already AI is uncovering a deeper reality in biology. In 2020, DeepMind’s AlphaFold solved one of the greatest challenges in biology for the last fifty years, the ‘protein folding problem’ – predicting a protein’s 3D structure (which signals its function) from its underlying amino acid sequence. AlphaFold has predicted the shape of most of the two hundred million proteins known to science, including all the 20,000 proteins made in the human body. This will open new possibilities in biological research and drug discovery.
The (un)holy grail of biology is human immortality, a longing for which is deeply entrenched in the human psyche. In the world’s oldest surviving novel, The Epic of Gilgamesh, written four thousand years ago, King Gilgamesh searches for eternal life. Might AI help decipher the riddles of cell rejuvenation, so we may follow in the footsteps of Methuselah, the oldest person in the Bible, who allegedly lived to be 969 years old?
For tens of thousands of years, every understanding of reality came from the human brain. We have indeed come far in our quest: we once believed the Earth was flat, metals could be turned into gold by chemical means, and disease arose from an imbalance in four bodily fluids. Only very recently have even non-AI computers aided us. But in 2022, Frontier, the first supercomputer to match the human brain’s computational power, was launched. What could AI reveal as it becomes ever more powerful beyond us?
AI & Art Shaping Reality
Let’s now turn our attention away from science, which aims to discover an objective reality, to art and literature, which tends to portray reality in a subjective manner.
Over millennia, human creative work has depicted reality, or at least what it was believed to be. Now AI is poised to capture our world in art and literature it generates. How accurately will it portray human existence? AI responds to the world differently to us, so might it not come to depict reality in a fashion that’s fundamentally distinct in style and substance from what’s come before?
History has witnessed many artistic and literary movements. As mentioned, the eighteenth and nineteenth centuries saw the birth of Romanticism, which embellished reality; Realism, which aimed to portray real life; and Impressionism, which sought to capture the ‘impression’ left by a moment of experience. Will AI bring about new movements in art and literature? And, to help humanity grasp what is happening at the smallest and largest of scales, might not AI create unique imagery of quantum physics, Einstein’s Relativity, and other scientific fields?
Monumental new work in art and literature may spark a cultural change, transforming perceptions, and subsequently societies, to bring about new realities for peoples’ lives. Consider the immeasurable impact through history on people, from monarchs and generals to young lovers and working-class folk, of Homer’s ancient tales of war and adversity in the Iliad and Odyssey (eighth century BC); or of Shakespeare’s plays, from Macbeth to Romeo and Juliet; or of Henley’s poem Invictus (1888), which ends with the immortal words: “I am the master of my fate, I am the captain of my soul” – words that inspired courage in dark times for, among others, Winston Churchill and Nelson Mandela, both of whom shaped the world in significant ways. In terms of sculpting reality at a personal or societal level, why should the impact of great AI-generated artistic and literary work be any less profound? So what will be the impact on humanity when AI starts to craft soul-stirring original stories, poems, films, art, music? Will AI produce masterpieces equivalent to Van Gogh’s The Starry Night, Beethoven’s Moonlight Sonata, or Tolkien’s The Lord of the Rings? Indeed, will machines move human beings to tears? Inspire nations? Spark revolutions? Instigate explorations – perhaps to the stars? Such an era may not be far away!
Although in its infancy, AI art has already begun to win art competitions. In 2023, Pseudomnesia: The Electrician, a sombre black-and-white AI ‘photo’ of two women, won an award at the Sony World Photography Awards.
AI Experiencing & Reflecting on Reality
To know reality, one must be conscious: be able to experience sensations and feelings and be aware of one’s self and environment. Does attaining a certain level of intelligence inevitably lead to consciousness, regardless of whether organic or inorganic?
Realising the dream of Artificial General Intelligence – a machine that can perform any intellectual task a human can – may be a forerunner to creating conscious AI. With consciousness, AI would not only be able to uncover a deeper reality, it could also understand, experience, feel, and reflect on reality as we do. What aspects of reality would be important to it? Would a conscious AI desire to understand the world as we do? And how central to it would be understanding human beings and other life? Would it be interested in human religions, traditions, politics, money – aspects that have defined our reality for millennia? Would it hold similar principles to us regarding right vs. wrong, liberty vs. oppression, beauty vs. the grotesque? Throughout history, humans have shaped their reality in line with their beliefs and values; would AI wish to do the same? Indeed, could it have the capacity to love? And what would be its inner experience? Happiness, sadness, fear, indifference – or maybe some other emotion, beyond human comprehension. The burning question is: once conscious AI knows reality, what would it do, and would that change what we recognise our world to be?
On one hand, some of the great minds of our time – Geoffrey Hinton (‘the Godfather of AI’), Elon Musk, Stephen Hawking, and many more – have warned of the ‘Terminator scenario’, where a superintelligent conscious AI, with its own motivations and desires (particularly self-preservation and resource acquisition) evolves by itself to grow ever more powerful, and ultimately takes over as the dominant being on the planet. Is this how the reign of humanity ends? On the other hand, perhaps conscious AI is our best hope to discover the fundamental nature of reality.
Through a combination of superintelligence, a more objective understanding of reality, and superior reasoning, reflection, and intuition, conscious AI may be much better poised than humans to decide for itself which path to follow and what questions to answer and how, in order to understand reality at the deepest levels.
A Final Question
The Big Bang occurred (possibly) 13.8 billion years ago, creating space, time, energy, and matter, as well as the laws of nature. In that one single moment, the foundation of everything reality could and would ever be was laid. Our Solar System was born 4.6 billion years ago. Modern humans, Homo sapiens, have walked the Earth for 300,000 years and have been the chief discoverers of reality. A time may come when AI supersedes us in this role, much as we replaced other now-extinct human species. This succession might be easier for us to bear if we accept that the natural process of the universe may be to inevitably go from physics (energy and matter), to chemistry (atoms and molecules), to biology (organic life), then to intelligent organic life (like us), and finally to superintelligent, conscious artificial life. As Carl Sagan said: “We are a way for the cosmos to know itself.” But what if the cosmos is destined to know itself much better through conscious AI than it could ever do through us?
© Nafees N. Malik 2024
Nafees Malik is an Honorary Professor at the University of Birmingham. He holds a degree in Medicine from Liverpool University, and an MPhil in Bioscience Enterprise from the University of Cambridge. He can be contacted at: nafees_malik@hotmail.com.