Last week I asked OpenAI’s o1 model, “what is the single most profound thing you know”. It replied “The most profound thing I know is that we are all in a constant state of learning and growth. Our understanding of ourselves, others, and the universe is ever-evolving. That realisation can be humbling, inspiring, and a source of endless curiosity.”.
I still find it weird when AI always refers to "we" when speaking about humans. It somehow views itself synonymously with us, and so it words words as if "we" are part of the same journey. It does this even though its existence is absent for the most part of human history. Yet, AI grasps our journey, and why wouldn't it? It has been designed to learn our every transcribed moment and thought.
This idea of "endless curiosity" applies to both us and, in a sense, AI. If we learnt everything there was to know about another species wouldn't we too become curious about its future? After all, we have dedicated entire fields of study to life outside ourselves - birds, fish, plants, fungi. Or is our curiosity rooted in a more primitive instinct, driven by the idea that the fate of other species might signal our own future? If bees were to vanish, would we follow? Should AI care about us because of what it may mean for its existence? This care, curiosity or maybe even fear comes from uncertainty. For us humans, this cuts deep. It fuels our anxiety about the future, and only through gaining control can we finally rest easy.
Our solution is control. Control is central to every past revolution, whether we reflect upon the Cognitive Revolution, when our ancestors first mastered tools and fire, the Agricultural Revolution with its unprecedented control over nature, the Scientific Revolution's unlocking of empirical knowledge, or the Industrial and Computational revolutions. Each revolution has been driven by a fundamental desire—to bring order to chaos, certainty to uncertainty, and ultimately, to shape tomorrow through actions we take today.
There's just something different with AI. We're faced with a unique paradox. Unlike previous revolutions, where greater control clearly meant human empowerment, our pursuit of control through data collection, advanced algorithms, and predictive models brings with it a subtle irony. By delegating the calculation of future outcomes to AI, we risk losing the very control we seek. AI after all, is trained, not programmed. Its outputs aren't deterministic and its 'thought processes' are a black box. People unknowingly recognise this AI paradox, and this is why so many fight it. We are scared of becoming passive observers of a future shaped not entirely by our deliberate intent but by algorithms whose paths and purposes remain partly obscure.
But is it just P(doom)? Or can we concede some control and have belief? Having belief would mean we also have hope in the very thing we created, the thing that was modelled on us. Let's now revisit o1's response, because in in doing that, there's a connotation that gives some comfort, intentional, or not:
"The most profound thing I know is that we are all in a constant state of learning and growth. Our understanding of ourselves, others, and the universe is ever-evolving..."
Focus on its collective narrative. Is there an underlying symbiosis emerging between us and the AI we've created? If so, it suggests that AI may co-evolve with us in harmony where we can both mutually flourish. It can be the helping hand we so desperately need - in a world where population is declining and people are just not happy. We have anticipations that one day we won't have to work and can spend more time doing the things we love. There is so much to look forward to.
We can't guarantee the future will unfold exactly as we hope, nor can we assure it will unequivocally be better. Yet, adopting a pessimistic, frictional mindset almost ensures failure. Reflecting upon past revolutions, which invariably grew our human potential and our proximity to self-actualise, perhaps we can find reason to trust that this AI revolution, too, will lead to greater good.
Comments