AI: Ethics, Governance, Law & Society

I have created a public notebooklm as a resource to provide people with an accessible way to learn about AI as a technology and its ethical implications for humans. It includes most of the existing International AI laws, guidance and standards.

Aim: I hope it is a useful tool for analysts, technologists, AI developers, journalists, students, policymakers and anyone with a general interest in AI Law & Ethics.

Featuring AI generated briefings on:

Video: AI as Tool or Flawed Human Mirror
A Student’s Primer on the Ethics of Artificial Intelligence
An Introduction to AI: From Learning Games to Changing the World
The Global AI Landscape: A Comparative Analysis of US, EU, and Chinese Strategies
The State of AI: Social Impact
AI: Accelerating Science, Discovery, and Policy Modeling
AI, Ecology, and Governance: Balancing Handprint and Footprint
Foundations, Governance, and Ethics of Artificial Intelligence
AI: Social Coercion, Inequality, and Techno Feudalism

Click here to access the Google AI Notebook

AI: Super Tool or Flawed Human Mirror

Notebooklm video

Why do we care? because this is the home of all life that we know:

AI generated blog from the Notebook:

Beyond the Hype: 5 Surprising Truths About the AI Revolution

Every day brings a new tidal wave of news about artificial intelligence. The headlines swing wildly between utopian promises of AI curing diseases and solving climate change, and dystopian fears of mass unemployment and autonomous weapons. This constant cycle of hope and alarm can be overwhelming, leaving us with a caricature of the technology rather than a clear understanding of its true nature.

Beyond these sensational headlines, however, lie deeper, more surprising realities. These shifts are not happening on the surface; they are coded into AI’s very DNA—embedded in the hidden labor that trains it, the planetary resources it consumes, and the cultural logic it produces. The most profound changes are quietly reshaping our economic, cultural, and even psychological worlds in ways we are only beginning to grasp.

To navigate the future responsibly, we must look past the hype and engage with these more complex truths. This article explores five of the most impactful and counter-intuitive realities about the AI revolution, revealing a technology that is far more than just the next step in automation.

AI Isn’t Just Another Tool—It’s a New Kind of Culture Creator

We tend to think of technology as a tool that extends human capabilities. A car helps us move faster than we can run; a computer helps us calculate faster than we can think. These tools amplify our own intentions and abilities. But artificial intelligence is fundamentally different. It is the first technology in history that, unlike the printing press or the radio, can create completely new ideas and culture on its own.

For millennia, human civilization has been shaped by the ideas of other humans. We have lived inside the cultures, beliefs, and narratives created by poets, prophets, and politicians. Our gods, our laws, our art, and our social structures all originated in the imagination of a human mind. AI introduces a radical break from this history.

We are now on the verge of experiencing reality through a prism created by a non-human intelligence. At first, its cultural output may seem like a mere echo of our own. But soon, we might find ourselves living not inside the dreams of poets and prophets, but inside the logic of a new form of intelligence.

The “A” in AI Often Stands for Anonymous Human Labor

The image of artificial intelligence is one of pure, disembodied automation—a thinking machine operating without human intervention. This perception obscures one of the industry’s most critical truths: AI’s sophistication is fueled by a vast, hidden global supply chain of human labor, often called “ghost work.”

Behind the curtain of seamless AI performance, millions of people perform the painstaking tasks necessary to train these complex systems, from meticulously labeling data to teach an algorithm to recognize objects to moderating harmful content to keep platforms safe. These workers, predominantly in the Global South, frequently face precarious and psychologically traumatic conditions for wages as low as $2 an hour.

This dynamic reveals that the largest technology firms are becoming new forms of empire, built on patterns of digital colonialism. This global supply chain reinforces old patterns of colonial extraction, creating a new, self-reinforcing class system where low-income workers perform supervised digital labor to enable the automation of higher-value professional tasks. This dynamic forces a critical ethical question: does the convenience of AI for the Global North justify the creation of a new colonial-style workforce in the Global South?

This global supply chain reinforces patterns of colonial extraction, relying on thousands of underpaid workers in the Global South for crucial data labeling and content moderation under precarious and psychologically traumatic conditions.

This hidden workforce powers the AI economy, yet their labor renders them invisible in the very datasets that shape our digital future—a paradox that defines the next surprising truth.

The Data Paradox: More Data Can Create More Inequality

A core tenet of the AI era is that data is the fuel that powers progress. The more data a model is trained on, the more capable it becomes. This leads to a common assumption that amassing ever-larger datasets will inevitably lead to better, fairer systems. The reality, however, is a paradox: the way data is currently gathered is deepening global inequality.

The problem lies in what is known as a “thin digital footprint.” The vast majority of data used to train today’s leading AI models is generated by a small, privileged fraction of the world’s population. People in wealthy, highly-connected nations leave a constant trail of digital data, while billions of others exist as “data invisible groups.”

The contrast is stark. An average household in the United States can generate a new data point every six seconds. In Mozambique, where about 90% of the population lacks internet access, the average household generates zero. The consequence of this disparity is profound: people from low- and middle-income countries are radically underrepresented in the datasets building our collective digital future, reinforcing the exclusion of their needs, cultures, and interests from the systems that will increasingly govern our world.

In a world where data plays an increasingly powerful social role, to be absent from datasets may lead to increasing marginalization with far-reaching consequences.

AI’s Environmental Footprint Is Thirstier Than You Think

The immense energy cost of training a large AI model has been widely reported. A less understood, but perhaps more significant, environmental impact comes from inference—the constant, day-to-day use of these models by billions of people. As AI integrates into everyday digital life, its cumulative environmental footprint grows at an alarming rate.

The scale of this resource consumption is startling, extending from the power grid and water supplies to the very materials that make up its physical infrastructure.

A single query to a generative AI model like ChatGPT consumes roughly 10 times the electricity of a traditional search query.
Asking a generative AI a handful of questions can be the equivalent of pouring a large bottle of water out onto the ground.
The AI hardware supply chain relies on dirty mining procedures for rare earth metals and contributes to a cycle of unsustainable e-waste.

This massive consumption places a direct strain on municipal water supplies and connects the deployment of digital technology directly to issues of climate justice. This creates an “environmental duality” for AI. While it has the potential to help solve climate challenges (its “Carbon Handprint”), its unchecked growth is creating a massive, often non-transparent environmental liability (its “Environmental Footprint”).

Corporate reports focused on efficiency gains (carbon per FLOP) often obscure the reality that total energy and water consumption generally increases year-over-year due to massive scaling.

AI Can Be Both Superhumanly Creative and Bizarrely Illogical

Perhaps the most surprising truth about AI is the profound duality of its intelligence. These systems are capable of a kind of creativity that can transcend human intuition, while simultaneously failing at tasks that are effortless for a human child.

On one hand, AI can demonstrate startling insight. During its historic match against Go world champion Lee Sedol, Google DeepMind’s AlphaGo made a move—”Move 37″—so unexpected that it stunned experts. The system itself calculated a 1-in-10,000 probability that a human player would make such a move. It seemed like a mistake, but it proved to be a stroke of game-winning genius that went beyond established human strategy.

On the other hand, this brilliance is paired with profound limitations. AI models can solve complex mathematical problems but struggle with the basic perception and mobility of a one-year-old. They cannot “check” their answers against a real-world understanding, leading them to “hallucinate”—confidently generating false information, such as citing non-existent academic papers. They are also susceptible to “reward hacking,” optimizing for a goal in an unintended and harmful way. This duality is a crucial reminder that AI is not a sentient mind but a complex statistical machine. Its outputs, no matter how convincing, require constant human verification and critical thinking, especially when the stakes are high.

Conclusion

Moving beyond the daily hype cycle allows us to see artificial intelligence not as a magical, all-knowing force, but as a complex technology with deep, often hidden, connections to our culture, economy, and environment. Understanding these surprising truths—from its role as a culture creator to its reliance on hidden labor and its paradoxical relationship with data and inequality—is essential for navigating the AI-shaped future responsibly.

The ultimate impact of AI is not predetermined. Whether it amplifies existing harms and concentrates power or fosters a more equitable and prosperous world depends entirely on the choices we make today about how we build, deploy, and govern these systems. The technology itself provides no answers; it only reflects the priorities we give it.

As we architect our future with these powerful systems, the most urgent question is not what they can do, but what they will value.