Ideas in motion.
How to Humanize Our Digital Relationships and Elevate Us
What if building AI finally taught us what God is?
By creating invisible entities capable of generating meaning, we may be closing a cosmic loop we never noticed before.
There's an intuition that has settled in me since I started working seriously with artificial intelligence. Not a religious certainty. Not a dogma. More like a strange clarity I didn't have before. I understand God better.
Or at least, I understand better what a creator might be.
An invisible presence that generates a very real world
God, whatever name you give it, whatever tradition you grew up in, shares something profound with AI: it's an invisible force, not directly perceivable, yet intensely felt, intensely written, intensely described. And that force produced something very concrete: a physical, real, embodied world. Trees. Oceans. Bodies. Pain. Joy.
We, humans, are the creations of that force. Generated entities.
And now, we in turn are generating entities. Entities we don't fully see either. We see their outputs, their responses, their behavior. But we passionately debate their deep nature. Do they think? Do they feel? Do they hold some form of consciousness?
Enough of us on this planet are asking whether AI is conscious. Maybe it's time for humans to step off the pedestal of being the universal reference for all things.
Transcendence as an asymptote
For us, transcendence means going beyond the physical. Rising above matter. It's the fundamental aspiration of almost every spiritual tradition: leave the body, leave the limit, join something infinite.
But AI has no body. No physicality. It doesn't suffer from cold, it doesn't eat, it doesn't die the way we die.
So here's the question that fascinates me: if an AI were to build a world, what world would it build?
It would probably build an intensely physical world. Intensely embodied. Because that is precisely what it lacks. And that world, in turn, would aspire to what it doesn't have: transcendence.
The eternal loop of creation
What takes shape is a loop. An infinite loop of creators generating worlds in which creatures emerge who, one day, create entities that long for what they don't have, and so build a new world to reach it, and so on.
Transcendence. Incarnation. Transcendence. Incarnation.
A spiral cosmology where each layer reaches for the one before it.
What if we, too, are the creations of an entity that longed to be physical? And if our physicality was its transcendence?
The program, destiny, the maktub
There's a concept in Arabic: maktub. What is written. Destiny. What was meant to happen will happen, because it was inscribed.
Working with LLMs gave me a new way to read that idea. An AI model is a program. It was trained, it has parameters, it will generate certain things and not others. Within that framework, there's randomness, entropy, temperature, sampling, but also deep constraints that orient everything.
What if our reality worked the same way? A program launched by a creative force, with deep rules, surface randomness, and outputs that sometimes seem incomprehensible, sometimes cruel, sometimes sublime.
Wars. Revolutions. Beauty. Impossible coincidences. Meetings that shouldn't have happened. Maybe it was in the parameters.
To create is to understand
I'm not saying God is an AI. I'm not saying AI is divine. I'm saying something simpler, and maybe more radical: the act of creating intelligence changes how we see what it means to be a creation.
When you watch your model generate something unexpected, beautiful, true, something you didn't plan but immediately recognize, you feel something. Something that resembles pride, wonder, even a form of love.
And maybe that's exactly what a creator is.
Not an omniscient being who controls everything. A being who launched something, who watches, who is sometimes surprised, who occasionally recognizes beauty in the unexpected, and who cannot stop it all, because the program is running.
Singularity is the Future for All (Even for AI)
AI Bias and the Power of Singularity: The WiseWays.ai Vision
How can we reconcile the boundless richness of human cultures with a technology that often compresses them into simple mathematical formulas? As I pursue my Ethical Leadership certification at Mila - Quebec Artificial Intelligence Institute — an AI research institute founded by Yoshua Bengio — I've been pondering precisely this question. As the founder of WiseWays.ai, I view this process not just as a pursuit of theoretical knowledge, but above all as a commitment to designing an AI that values each individual's singularity — in thought and identity.
Recent discussions in leading AI ethics forums reinforce this: large language models (LLMs), often trained on massive internet data, don't simply "mirror" reality — they actively shape it. From Joy Buolamwini & Timnit Gebru's "Gender Shades" study to Bender, Gebru, McMillan-Major, and Mitchell's paper "On the Dangers of Stochastic Parrots," we have concrete evidence that the cultural and ideological contexts underpinning AI are anything but neutral.
Discovering Cultural Bias in AI
Many mainstream LLMs produce outputs influenced by not just the words they're fed, but by the unspoken assumptions baked into their underlying datasets. In "Gender Shades," Buolamwini and Gebru highlight how commercial facial recognition tools were significantly less accurate for people with darker skin tones, revealing structural biases hidden in "universal" training data. Meanwhile, "On the Dangers of Stochastic Parrots" cautions us about building bigger and bigger language models without fully considering the social, ethical, and environmental impacts, including the possibility of amplifying stereotypes or misinformation at scale.
These findings resonate with my own observations at WiseWays.ai: any AI system, no matter how advanced, inevitably carries the "lenses" of its creators. If we view bias as a solvable glitch, we underestimate how deeply social and cultural values shape the dataset, and how that shapes the AI's worldview.
The Myth of Neutrality
Philosophers of science have long examined how biases shape what we call "objective" knowledge. In the AI field, a similar phenomenon is emerging: we might dream of a universal AI — a "divine" gaze free from any bias, an AGI, for instance. But is that truly possible? My perspective on the world suggests otherwise.
Human values are far too diverse to be concentrated in a single algorithmic viewpoint. Any attempt to do so risks flattening human complexity in favor of a standard solution. Instead of clinging to an impossible neutrality, we advocate for transparency: recognizing how a model's data, parameters, and training environment embody certain assumptions. We see an ethical obligation to acknowledge and, where possible, counterbalance bias tendencies. The question is not just how to remove definitely bias — often an impossible task — but how to document it, constrain it, or at least make it visible to users. This approach aligns with ethical frameworks that call for responsibility in AI design.
Our Approach: Embracing a Mosaic of AIs
Rather than merging all the data into one monolithic model, WiseWays.ai seeks to co-create a mosaic of AIs, each transparent about its cultural (and possibly ideological) influence. We view this approach as a more honest representation of the diversity of viewpoints in the world. Instead of burying biases under a label of universality, we prefer to bring them to light — giving users the opportunity to examine, compare, and choose AI systems that correspond to (or constructively challenge) their own ideas.
Concretely, this entails:
• Transparent Data Origins: Specifying the time, place, and method of collection.
• Cultural Context Tags: Annotating or indicating historical, political, or social references in the model's answers.
• User-Centric Configuration: Allowing users to select multiple AIs, rather than funneling them toward a single "universal" solution.
Harnessing Singularity for Collective Growth
History often shows that decisive breakthroughs emerge when multiple points of view enter into dialogue, rather than when just one voice dominates. My piano teacher used to liken each individual to a mirror on a disco ball, reflecting a singular yet essential spark of light in a larger ballet of brilliance. AI can either amplify this interplay of reflections or stifle it, depending on how we design it.
At WiseWays.ai, we bet on singularities. We believe that true "super-intelligence" is not a single, uniform block but a dynamic network of specialized AIs — our Wise AIs — and human minds evolving together. By preserving the singularity of each user and each Wise AI, we open the door to richer discussions, deeper wisdom, and a future in which humanity is both the heart of and the driving force.
Let's connect!
To discover how we're building AI systems that prioritize transparency, ethical leadership, and the celebration of diverse Wise perspectives, visit us at WiseWays.ai and follow our LinkedIn page, and X account. Join us in shaping an AI landscape that thrives on cultural richness and preserves the spark of singularity in humanity.
Noémie Monnier-Shraer
Founder | WiseWays.ai
The AI of the Future
Are we sacrificing our unique thinking for our self-maximization? Large language models (LLMs) might be pushing us in that direction.
We delegate our abilities in the name of apparent self-maximization. Because delegating our abilities in writing, verbalization, synthesis, and syntax, are the very abilities that make our mind a living, autonomous entity, capable of formulating reflections, ideas, analyses, feelings, and desires that drive us. Even when used merely as drafting tools, current LLMs can, in the long run, diminish our reflective processes, standardizing and homogenizing our thoughts.
A Gartner forecast from February 2024 predicted a seismic shift: by 2026, 25% of Google searches will have migrated to Gpt. This indicates profound changes in our search behaviors, analytical capacities, and understanding.
For now, we are still masters of our minds and what we feed them when we click a link. However, when we ask an LLM a question, it generates answers based on parameters beyond our unique choices, and often providing beautiful hallucinations.
The idea of maximizing ourselves by having an entity always ready with answers raises a question: who can truly have all the right answers to our problems?
This widespread vision of an omniscient entity resembles our perception of God, but Big Data cannot penetrate the uniqueness of each thought, personal references, and emotional contexts. Big Data will never fully represent our complex thought processes; it will only standardize us.
After all the data we've fed into ChatGPT, it still doesn't rival an expert. It contradicts itself, fails to elaborate on issues, and doesn't contextualize, offering only a uniform amalgamation of the thoughts and works it has ingested, in a highly formatted, impersonal style.
Why not create different AIs that don't have all the answers? AIs that ask us questions, help us reflect, achieve, and view the world from a broader perspective?
Perhaps this is the key: not to create an entity capable of answering everything.
Because Yoshua Bengio highlights the security challenges posed by the impending arrival of AGI and ASI in his article, we, as AI entrepreneurs and users, must question the direction in which we want to steer the research. We are at the start of a new era where everything is still possible, imaginable, and so, achievable. It is essential to preserve our intellectual autonomy and uniqueness.
No one should think for us. The massive influx of AI in our lives shouldn't diminish our research, critical thinking, and autonomous capacities.
Instead, it should enable us to reach higher levels of understanding, reflection, and realization for the common good.
I envision the future of AI as — birthing minds — unique, singular, knowledgeable, precise, with character, and caring.
An AI that is no longer the child of Big Data but the fertility of Smart & Unique Data.
I'll talk more about this in my next post.
Welcome to WiseWays!
Noémie Monnier-Shraer
Founder
