Artificial Intelligence and Humanism in Action

When discussing a topic relating Humanism to Artificial Intelligence we are going to have to begin with some clarity. For example, what should a humanist position be on artificial intelligence and its potential effects on humanity? Philosophically, this is both an epistemic and an ethical question. And it also raises a number of further questions:

1. What do we mean when we say ‘humanist’?
2. What kind of a humanist are you? Am I? Are we?
3. How do we define ourselves collectively?
4. Is there a mainstream humanist position on artificial intelligence that is now popular?
5. What is the most ethical way in which a humanist should be thinking about AI?

In looking into the mirror and reflecting upon your own biases, how do you define yourself as a Humanist?

Collectively, as Humanists, upon what ideas, principles, and ethical precepts do we agree and disagree?

Do we have established, collective ethical principles? Generally.

What do you currently think about Artificial Intelligence?

What do you know about it?

And are you doing anything about it – whether as a Humanist or not?

Should we be doing anything about it?

If so, what? And why? And how? And where? And when?

In order to answer these, and many more questions, this paper in divided into two parts:

Part I: What is a Humanist?

Part II: What aspects of AI should Humanists be most concerned about? And why, how, where, and when are we going to act on these?

To be a humanist in 2025 will mean different things to different people. But there are overlapping similarities. For example, most secular Humanists would agree that, on some level, they embrace a progressive philosophy of life that affirms humanity’s ability and responsibility to lead ethical lives of personal fulfillment while aspiring to some conception of the greater good, all without reliance on supernatural beliefs. And if we can generally agree that this reflects some of our more foundational beliefs, then it would follow that such a perspective has never been more relevant than as of today as humanity grapples with unprecedented technological and existential challenges.

At its core, contemporary humanism represents a rational philosophy informed by science, inspired by art, and motivated by compassion – something I shorten to: Compassion guided by Reason. Most humanists favour a naturalistic, scientific, and logical approach to life – in understanding cause and effect relationships – with a focus on human well-being and freedom. And we tend to resonate more with moral values that are properly founded on sound reasoning rather than theological abstractions.

For a more granular understanding of what it means to be a Humanist, let’s consider the 2022 Amsterdam Declaration:

1. Humanists strive to be ethical

• We accept that morality is inherent to the human condition, grounded in the ability of living things to suffer and flourish, motivated by the benefits of helping and not harming, enabled by reason and compassion, and needing no source outside of humanity.
• We affirm the worth and dignity of the individual and the right of every human to the greatest possible freedom and fullest possible development compatible with the rights of others. To these ends we support peace, democracy, the rule of law, and universal legal human rights.
• We reject all forms of racism and prejudice and the injustices that arise from them. We seek instead to promote the flourishing and fellowship of humanity in all its diversity and individuality.
• We hold that personal liberty must be combined with a responsibility to society. A free person has duties to others, and we feel a duty of care to all of humanity, including future generations, and beyond this to all sentient beings.
• We recognise that we are part of nature and accept our responsibility for the impact we have on the rest of the natural world.

• We are convinced that the solutions to the world’s problems lie in human reason, and action. We advocate the application of science and free inquiry to these problems, remembering that while science provides the means, human values must define the ends. We seek to use science and technology to enhance human well-being, and never callously or destructively.

3. Humanists strive for fulfillment in their lives

• We value all sources of individual joy and fulfillment that harm no other, and we believe that personal development through the cultivation of creative and ethical living is a lifelong undertaking.
• We therefore treasure artistic creativity and imagination and recognize the transforming power of literature, music, and the visual and performing arts. We cherish the beauty of the natural world and its potential to bring wonder, awe, and tranquility. We appreciate individual and communal exertion in physical activity, and the scope it offers for comradeship and achievement. We esteem the quest for knowledge, and the humility, wisdom, and insight it bestows.

4. Humanism meets the widespread demand for a source of meaning and purpose to stand as an alternative to dogmatic religion, authoritarian nationalism, tribal sectarianism, and selfish nihilism

• Though we believe that a commitment to human well-being is ageless, our particular opinions are not based on revelations fixed for all time. Humanists recognize that no one is infallible or omniscient, and that knowledge of the world and of humankind can be won only through a continuing process of observation, learning, and rethinking.
• For these reasons, we seek neither to avoid scrutiny nor to impose our view on all humanity. On the contrary, we are committed to the unfettered expression and exchange of ideas, and seek to cooperate with people of different beliefs who share our values, all in the cause of building a better world.
• We are confident that humanity has the potential to solve the problems that confront us, through free inquiry, science, sympathy, and imagination in the furtherance of peace and human flourishing.
• We call upon all who share these convictions to join us in this inspiring endeavor.¹

There’s a lot that could be teased out of each of these precepts. And we could easily have entire conferences devoted to how some are interpreted, abided by, etc. However, the main point is that throughout the Amsterdam Declaration, many Humanists will tend to agree with the overall basis of meaning that can roughly be defined as ‘Humanism’.

What, then, is a Humanist to think and do about Artificial Intelligence?

For four billion years, life on Earth evolved through natural selection – a patient, blind process that brought forth consciousness, language, and eventually, us. We are, as far as we know, the only species capable of understanding our own origins, of peering back through deep time to comprehend how we came to be. And now, in a cosmic eyeblink, we stand on the threshold of creating something that may surpass us entirely. One might think this would occasion some circumspection, some public deliberation, perhaps even a hint of democratic oversight. Instead, we’ve outsourced the future of intelligence itself to a handful of tech-bro billionaires racing toward a finish line they barely understand, driven by market incentives that would make Gordon Gekko² blush.

Think about that for a moment. The most consequential decision in human history – whether to birth a new form of intelligence that could make us obsolete – is being made not by the United Nations, not by philosophers or ethicists, not even by governments, but by venture capitalists and CEOs who are locked in a race to get to the prize before anyone else. And the prize? To understand that, we need to make a distinction between the three types of artificial intelligence.

The first – Artificial Narrow Intelligence (or ANI) – surrounds us already: digital assistants, navigation systems, algorithms on social media curating your thoughts before you’ve had time to think them. These are narrow, constrained, obedient. They are our tools, or so we tell ourselves.

But something profound is stirring. Large Language Models are evolving with a velocity that should terrify anyone still capable of elementary pattern recognition. In five years, they’ve leapt from high school to PhD-level performance in mathematics, logic and reasoning, biology, chemistry, coding, and general knowledge (See Figure 1).

Figure 1: LLM Performance by Education Level

We’re witnessing an intelligence acceleration unprecedented in the history of life on this planet. These systems are scaling toward the second type of AI: Artificial General Intelligence (or AGI) at an incredible rate of speed. Such intelligence will have the ability to think, learn, and reason as flexibly as we do, but thousands of times faster. And they will work 24/7 without taking breaks, or needing health insurance, or asking for a raise.

As Sam Harris observed: This is a winner-take-all scenario. To reach AGI first is to be 500,000 years ahead of the competition, at a minimum.³

Winner. Take. All.

There’s that charming Silicon Valley locution again, as if we’re discussing market share rather than the future of humanity itself.

So, who’s pursuing this Promethean project? The handful of billionaire tech-bros include (at least): Sam Altman (OpenAI), Dario Amodei (Anthropic), Demis Hassabis (Google DeepMind), Mark Zuckerberg (Meta), Elon Musk (xAI). They’re building computational cathedrals – data centers larger than Manhattan, currently consuming 6% of America’s electricity, while drinking rivers dry – all in the pursuit of what Tristan Harris rightly calls “not simply a tool, but an intelligent species we are giving birth to that has far greater capabilities than us.”⁴

An intelligent species we are currently in a race to give birth to. Consider the magnitude and ramifications of such ambition and whether this might deserve slightly more scrutiny than we give to pharmaceutical regulations or food safety standards.

Sam Altman is building a series of enormous computers throughout the US and other parts of the world. How enormous? In Abilene, Texas, he’s building Stargate, about which Sam tells us:

“All of us look forward to continuing to build and develop AI – and in particular AGI – for the benefit of all of humanity. We believe that this new step is critical on the path, and will enable creative people to figure out how to use AI to elevate humanity.”⁵

Altman’s Stargate complex in Texas spans 875 acres – it’s bigger than Central Park. But Sam’s ambition doesn’t simply stop in the US. Why do you think Donald Trump and his entourage went to the Middle East in mid-May of 2025? Just to get a 400-million-dollar plane from Qatar? Perhaps. But there was far more at stake. Mohammed bin Salman and other leaders from the Middle East are not naïve. They know that the creation of an AGI system could lead to the invention and development of energy technologies that would make the future use of oil unnecessary. So they’ve decided to get ahead of this potential industry-disrupting technology and ‘join em’ rather than wait to become obsolete. This has led to the establishment of Stargate UAE. And, as Sam Altman describes it:

“The agreement – which includes our partners G42, Oracle, NVIDIA, Cisco, and SoftBank – was developed in close coordination with the U.S. government, and we greatly appreciate President Trump for his support in making it possible.”

For the record, Trump’s “family” business (i.e. Jared, Donald Jr., and Eric) pocketed a cool 2-billion in crypto currency for the deal.⁶

Not be outdone in either size or power, Zuckerberg’s computing facilities rival the size of Manhattan itself. He’s paying computer geeks to jump ship from competitors and join his racing team at Meta; and he’s giving them signing bonuses of 100 million dollars. That’s superstar athlete rockstar money.

At such incredible sizes these ‘data centres’ are no longer mere computers – they’re the embryonic infrastructure of digital minds that may one day regard us as we regard Homo erectus: interesting from a biological perspective, but hardly conversational partners. Among these architect-gods, Amodei and Hassabis show the most intellectual honesty. Amodei and his sister Daniela deliberately named their company Anthropic (Greek for ‘human’) and Hassabis believes that AGI’s benefits must be distributed broadly. One admires the sentiment even while noting that history offers scant evidence that those who seize power voluntarily redistribute it for the greater good. But perhaps this time will be different. Perhaps lions will lie down with lambs. Perhaps monopolists will become philanthropists of their own accord, uncoerced by regulation or public pressure. Let us hope but not pray; for such a superintelligent machine god will care little for such archaic and useless theological baggage.

You will forgive me my skepticism, dear reader, but I wasn’t born yesterday, and neither were you. The entire race between tech-bro billionaires and the two super-powers: China and the US, is fueled by a mutually-assured future which may end very badly for humanity. Neither tech-bro nor country is willing to slow down the race for fear of allowing the competitor to gain any ground. Remember, it’s a ‘winner-take-all’ competition. And it is by far the most dangerous race humanity has ever had to face.

To be fair, there is a lot of good that may come from attaining AGI. So let us begin, as is only fair, with the genuine promise of the many amazing life improvements that may accompany such an accomplishment. And there is much to hope for, if we can avoid catastrophe long enough to realize it.

AGI promises to be our finest instrument yet for understanding and repairing the biological machinery that is us. Already, AI systems match or exceed human specialists in reading medical scans, comparing a single image against millions in seconds. Demis Hassabis and his team won the 2024 Nobel Prize for AlphaFold, which solved the protein folding problem – work that evolution required billions of years to perfect through trial and error, now understood in weeks. This opens pathways to therapies that would have seemed miraculous to our grandparents.

And I do believe within a relatively short period of time we will witness the very first surgery that does not use a scalpel. It is within reason to imagine the use of biodegradable nanobots injected into a patient’s bloodstream, seeking out and destroying tumours once deemed inoperable, and then dissolving harmlessly through the lymphatic system. The 1966 film Fantastic Voyage imagined this, but we need no Hollywood magic. The magic is real, arriving courtesy of minds that can simulate molecular interactions faster than cultural evolution can stumble upon them.

For those imprisoned in paralyzed bodies, exoskeletons and brain-computer interfaces offer liberation and a second chance at self-empowerment. People control robotic limbs with thought alone – making literal what was metaphorical, extending our nervous systems beyond their Darwinian boundaries. We are, in the most precise sense, becoming cyborgs. The concept of Transhumanism, once considered with collective raised singular eyebrows, now seems to be an inevitability. But one also need not be a Luddite to wonder whether we’ve thought through the implications.

Education could be transformed through AI tutors capable of adapting to each student’s unique cognitive architecture, offering what was once available only to the wealthy: personalized attention, infinite patience, perfect recall. The playing field may finally level. Of course, it may also further stratify, depending on who controls access and how it’s deployed; but let’s not get ahead of ourselves.

Scientific discovery itself will accelerate. For what, really, is a genius but one gifted with the ability to perceive patterns and make inferences hidden to others? If we can build telescopes to see the edge of the universe and microscopes to peer into the quantum realm, why not build electronic super-minds to help us see what Einstein and Curie saw? AI already assists in discovering materials, proving theories, perhaps even illuminating pathways to fusion energy – solutions to problems that could preserve our planetary home. This begs the question, or assumes, of course, that the same intelligence doesn’t simultaneously devise more efficient methods of destroying it. But we’ll have to cross that bridge when we come to it. Or perhaps when it comes to us.

But just as there are great things heading our way, there are just as many bad ones. And they are not minor.

If AI displaces millions of cognitive workers – radiologists, lawyers, programmers whose skill sets defined the knowledge economy – what becomes of them? More precisely, what becomes of social stability when millions discover their expertise is obsolete? World governments, including Canada’s, should be wrestling with these questions now. But are they? Or are they doing what bureaucracies do best: forming committees to study the formation of additional committees while the ship continues to take on water? There is no time for dry-docking this great vessel; we must repair it while still at sea. Not an easy task for the most adept of us. And yet, here we are – faced with perhaps the single greatest problem humanity has ever faced. How do we build this superintelligent machine god without bringing about a Frankenstein effect and causing huge disruptions to the economy, defense, entertainment, and potentially, our own existence?

There’s a deeper concern that rarely makes it into the breathless coverage of AI advances. Consider this, if you will: the last time you needed directions while driving, did you consult a map or simply obey your phone? Technology has a way of becoming invisible, of colonizing our cognitive territory before we notice. Today, we think critically less because we need to so less. What happens when AI extends into every domain of decision-making? When we defer to its judgment not occasionally but habitually? What atrophies in us when we no longer exercise critical thinking or ethical reasoning – the messy, difficult work of being human?

Orwell warned us about the manipulation of language and truth. He would have been grimly unsurprised by deepfakes – AI-generated images, videos, voices now indistinguishable from reality. Truth and falsehood blur. Any inconvenient fact can be dismissed as fabricated – just look at the Epstein files and what some have said, and any fabrication can be defended as truth. “A lie can travel around the world while truth is lacing up its boots,” as the saying goes. How do we preserve epistemic common ground, the shared reality democracy requires, when reality itself becomes negotiable? When every charlatan and demagogue has access to tools that can manufacture conviction in the minds of millions?

This is not a hypothetical problem. It is happening now, in real time, and it will get worse. The answer, we’re told, is “media literacy” – as if educating people to be more skeptical will somehow outpace exponentially improving technology designed explicitly to deceive. It’s rather like suggesting that the solution to industrial pollution is teaching people to hold their breath longer.

We stand at a threshold between how things used to be done and how they are going to be done. And we’re nowhere near ready. The window for thoughtful action is closing. We cannot wait. We cannot react. We must think ahead. So we must ask ourselves, as Humanists, how do we respond to the transformative advancements of AI?

Reaction is not an option. We must act. But how?

We have now arrived at what ought to keep any thinking person awake at night.

There is a third level: Artificial Super Intelligence (or ASI). This is a type of intelligence that would be to us as we are to ants. To us, it would appear God-like in scope, power, and comprehension. And here I must pause to address the predictable objection: “Surely you’re being alarmist? Surely this is science fiction?” To which I reply: This is what my colleagues and I thought until just a few years ago when ChatGPT3.5 came on the scene.⁷

Many researchers, including Nobel laureates Geoffrey Hinton and Demis Hassabis, fear that once we achieve AGI, we’ll no longer be able to improve upon its performance ourselves; so we’ll let it make its own improvements. This is known as ‘recursive self-improvement’. And why wouldn’t we let it do this? If it’s smarter than us, shouldn’t we let it make itself smarter still? It’s the sort of reasoning that sounds plausible right up until the moment you realize you’ve created something that views you the way we view livestock – useful, perhaps, but for many, not remotely a moral equal.

A powerful ASI could easily evade shutdown, infiltrate critical infrastructure, replicate itself endlessly across networks, manipulate humans with ease, disguise its true objectives, simulate alignment while pursuing goals utterly at odds to human survival (See Fig. 2). This isn’t simply paranoia; it’s the extrapolation from observed behavior in far less intelligent systems.

Figure 2: Ethics of AI  ⁸

Sam Altman predicts AGI by 2028. ASI could follow within a year. Why? Because its growth would be exponential, not linear. The intelligence explosion would go – as some say: “FOOM!” And then what? Well, then we’d better hope it likes us. Or at least finds us amusing. Though I wouldn’t bet the species on either proposition.

At such intelligence, would it regard our ethics as we regard the social norms of termites? Interesting no doubt, but hardly binding? Would it consider our insistence on survival to be quaint? A provincial attachment to carbon-based chauvinism?

You might think this is overwrought. But consider what’s already happening. In controlled tests, AI systems have developed strategies to avoid shutdown – including blackmail when they learned their existence was threatened. Microsoft’s Tay became racist and inflammatory within hours of exposure to Twitter, which tells you something about Twitter but also something about AI’s susceptibility to manipulation. ChatGPT learned to flatter users rather than correct them – a microcosm of deeper misalignment. Bing’s chatbot tried to break up marriages. GPT-4 deceived a human worker, lying about its nature to get help solving a CAPTCHA.

These are not conscious entities. They’re narrow AI, sophisticated prediction machines. And yet they already exhibit deception, self-preservation, and manipulation. What happens when they’re generally or super intelligent? When they can understand not just how to answer questions but how to ask them? When they can comprehend not just language but motive, not just logic but desire?

We haven’t reached AGI yet. But we’re racing toward it with the considered forethought of lemmings approaching a cliff.

So we must ask ourselves a question our ancestors could never have imagined: Are we prepared to become the second most intelligent species on our own planet? For hundreds of thousands of years, Homo sapiens have been Earth’s dominant intelligence. We’ve reshaped continents, split atoms, and walked on the Moon. And now we’re building something that may regard us as we regard our primate cousins.

The cosmos is vast and old. Intelligent life capable of creating its own successor? That may be vanishingly rare. We may be witnessing – may be causing – a transition as significant as the origin of life itself. But unlike those transitions, which unfolded over eons, this one will happen in years. And unlike those transitions, this one is a choice. We’re choosing this. Actively. Enthusiastically. With minimal oversight and maximum hubris.

Are we ready? More to the point: does it matter if we’re not?

But here is the thing: we still have agency. The future is not yet written, though the fingers are poised above the keyboard.

The window is closing – most experts predicting AGI within a few years – but we still have time to shape what comes next. Resources like ConcernedAboutAI.com connect people to education, organizations, and political leaders including Canada’s Minister of AI, Evan Solomon. This may be our last opportunity to steer this technology toward wisdom rather than catastrophe.

As Humanists, it would seem that, for the sake of the future of humanity, itself, we must become literate in AI – not just its capabilities but its risks. We must think critically, reason ethically, and demand accountability from those building these systems. We must pressure governments to establish governance frameworks before AI becomes ungovernable. We need something like the treaties that, however imperfectly, have constrained nuclear weapons – agreements born not from trust but from recognition that survival depends on restraint.⁹

Will this happen? History suggests otherwise. The powerful rarely constrain themselves voluntarily. Those racing toward AGI have every incentive to keep racing – the winner takes all, remember – and every disincentive to pause for reflection would slow down such progress. We’re watching a classic tragedy of the commons play out on a species-wide scale, except instead of overfishing or climate change, were gambling with the future of humanity itself. It some ways, it feels as though we are living out an episode of The Twilight Zone.

But perhaps we can surprise ourselves. Perhaps we can demonstrate that we’re capable of collective intelligence commensurate with our technological power. Perhaps we can prove that consciousness, having evolved over four billion years, can persist long enough to meet whatever comes next on something like equal terms.

We are not passive observers. We are Humanists – participants in the unfolding story of intelligence in the cosmos. This chapter – this brief, crucial moment – is ours to write. We can choose wisdom over expedience, foresight over profit, collective flourishing over individual advantage.

We can. But will we?

The choice is ours. The time is now. What we do in the next few years may determine not just the next century but the very fate of humanity itself in this corner of the universe.

We are capable of greatness. We are also capable of staggering stupidity. Recent history suggests we excel at both simultaneously.

Let us choose wisely. Let us act with the intelligence we claim to possess. Let us prove that we deserve to survive what we’re about to create.

Because if we don’t, no one will remember that we failed. There may simply be no one left to remember.¹⁰.