Issue 01 May 2025
Alex Asseily | Founder, Zulu Group
AI has never had the experience of living a single human life. Yet we're asking it to help us make sense of the world, and to help us navigate it.
The Myth of AI Understanding Humanity
Today’s LLMs are trained on abstractions of the human experience, data from millions of books, documents, articles mushed together in a giant averaging machine that sounds smart. This is because LLMs are extremely good imitators, trained by digesting the work of millions of experts across various fields. But like imitators, they often won’t know a subject from first principles. They have the polish and presentation of an expert, but often the substance of a motivated rookie.
Whereas the author of a book or wikipedia article on an obscure historical event will process the entirety of the information through a common sense filter (based on knowing both the subject but also the physics of the world), LLMs will not - they will merely imitate the sample of experts on which they have been trained. As a result, they are incoherent by default. They can exhibit both extreme intelligence and extreme stupidity.
AI’s topology is boundless software running on a network of computers. These computers are identical in function, if not also in form. AI is dissociated from this hardware; it doesn’t sleep but can remain dormant indefinitely; it has no biology. It cares only about compute / inference and energy. Its manner of thinking does not depend on who supplies its computers, what sort of electricity it consumes or what building houses it.
Most importantly, it doesn’t feel. AI calculates, but it does not care. It cannot ache in love, tremble in grief, or shiver at the sight of a mountain range bathed in light. These sensations—so often dismissed as inefficiencies—are, in fact, the root of our humanity. They bind us not only to our own bodies, but to each other, and to the living world around us. Compassion doesn’t arise from code; it arises from connection and shared experience. From skin meeting skin, from hands in soil, from breath shared in silence.
This is our original intelligence—not artificial, but embodied. And from this intelligence flows our responsibility: to each other and to the natural world that supports us. While AI may sound articulate, its hierarchy of needs is fundamentally alien to ours; our task is to harness its unique superpowers to accelerate societal progress—while fiercely protecting what makes us authentically human.
Embodied Consciousness
Our physical bodies need energy, sustenance, sleep, rest, etc. We are hard wired to seek food, to reproduce, and excrete. Like a handful of ‘eusocial’ species, humans evolved to have nests and division of labour (credit E.O. Wilson). This provided us with a way to engineer, at group level, physical safety, reliable food sources, protocols of cooperation and eventually social stability.
And in time, groups that could not just survive but also cooperate effectively amongst their members, would do better. As societal structures became more stable, our mysticism, cognitive and aesthetic needs emerged (credit Ken Wilber). Our cultures are therefore inseparable from our genetics, neurology, and geography. In a sense, our needs and priorities are emergent properties of our embodied consciousness, unfolding, seeking, experimenting and adapting.
We are heroes of our own stories, playing them out actively.
The "meaning of life" is not something to be discovered, but rather something we create and bring to life through our experiences and choices - Joseph Campbell
The passion of a Liverpool teenager for music, and their ensuing motivations and struggles to become a world class musical artist, articulate just one amongst myriad human journeys. We know that the satisfaction of climbing a mountain is not simply in being at its summit, but in the physical and mental struggles that accompany the climb and the memories of struggle. We feel alive because we went through the experience and it taught us something that we take forward in life.
Thus embodied meaning is central to the human experience. When we read a book or watch a movie we reference memories of authentic human experience. A poem moves us because it recalls stored feelings. Computers never refer back to underlying meaning - only to other programmed content. That may change with time.
Tools: central to societal evolution
Were it not for tools, we would not have written language, pens, paper, the printing press, radio, computers, and the Internet. You’d be hard pressed to find even the most dedicated luddite (named after 19th Century English workers who destroyed machinery they thought would destroy their jobs) who would turn their backs on the tools that have now become woven deeply into our societies. Tools are inseparable from our civilization.
Steve Jobs once described the personal computer as a ‘bicycle for the mind’ - the implication being that our minds can travel further while still finding meaning in the journey. One might rightly ask, at what point does the computer become an ‘autonomous car for the mind’: we decide the destination but nothing more. And when the destination becomes a ‘smart’ recommendation based on past behaviours, at what point do we accept we’re no longer in charge?
The effects of social media on our neurology shows how even ‘narrow AI’ can be destructive (credit Jonathan Haidt). Social media is still a system which humans control, albeit for profit. What happens if one is let loose - also for profit - but with no human supervision?
In light of this, should AI be optimised for joy, intellectual fulfilment, longevity, economic growth, settling Mars, a spelling bee, or deeper family connection? We humans barely agree on what societies should optimize for (hence politics) —yet our hard-won sovereignty lies in the ability to balance our own evolving needs as individuals, as communities, and as nations.
How do we preserve what it means to be human while using these AI tools to advance our world and societies, particularly as their raw cognitive power exceeds our own?
A better bicycle for the mind
At Zulu Group, we believe the most meaningful collaborations and investments are holistic — those that consider the full spectrum of human and planetary needs. We’re not opposed to the adoption of AI; on the contrary, we’re committed to shaping it to be deeply human-centric — a better bicycle for the mind. Here are some of our suggestions:
Mirror our hierarchy of needs: let’s teach AI to truly listen, learn, and adapt to individual needs— as it learns it can begin to mirror our embodied intelligence — longings, struggles, love, wonder, laughter—and respond in ways that feel grounded in our nature(s) without encroaching on the essence of being alive.
Enhance human ⇄ human connection— the wheel, the alphabet, the printing press, the steam engine, radio, and the internet transformed civilization because they brought humans together to share knowledge, collaborate and trade at ever faster rates. With AI we can take this further by collapsing remaining cultural barriers and linking otherwise scattered voices.
Truth: ‘better explanations’ of nature (credit David Deutsch) allow societal networks to evolve more coherently and exponentially - AI can play a role in identifying and exposing inconsistencies and distortions in our knowledge that stifle progress; it can enable ever better explanations to emerge.
The real measure of success won’t be speed, scale or even mastering a Turing Test —but whether societies embracing AI evolve toward higher collective consciousness as demonstrated by an acceleration of non-zero sum interactions (credit Robert Wright) - or regress from it.
The Zulu take