One Year of EverythingHuman
It has now been a little over a year since the idea of EverythingHuman was first conceived: a platform that advances collective thinking and deliberation, to help us work through some of the most complicated issues in the world. I want to take a moment to reflect on the ideals and the vision I personally held for this project.1 For me, EverythingHuman is a vision for addressing two of the most long-standing and complex problems in human society: education and governance. I used to say that EverythingHuman was more of a “thought experiment”, a philosophical project, than a technical one. I also used to say that to fully understand what EverythingHuman is about, we must first forget what we tried to do and what we built. So in this essay I shall break down a few of those ideas, loosely grounded in our experience working on it, but focusing on my vision behind it.
Thinking Humans, Not Just Thinking Machines
It all started as a discussion about how we could improve the higher education system in Canada. As undergraduate students, this was a topic that troubled us personally. We felt that much could be done to modernise how universities work — to actually help students learn, rather than keep up the façade of “teaching” — especially as the instruction and academic knowledge delivered in lecture halls drift further and further from the practical skills and frontier insights of industry. We wanted to tackle the problem from both sides: schools as places for the pursuit of knowledge, and schools as places for vocational training and employability. We also wanted to discuss it with stakeholders, the “experts” experienced in decision-making and in the actual structure and operation of educational institutions. Yet such a channel should not be restricted to experts and decision-makers alone. I doubt that they alone understand the problem better than the students who live through it every day, reflect on it, and genuinely care about it, rather than seeing it only from an administrative distance.
However, we found that we lacked a platform to do this systematically. Everything we landed on was either unstructured posting — a wall of text — or full of sentimental, non-evidence-based content and non-constructive discussion: mindless criticism on one side, propaganda on the other. It struck me that education was just one example among many problems of this kind.
Perhaps it isn’t the lack of such a platform that troubles me so much as the fact that fewer and fewer people are looking for one. I wrote in an earlier essay that most modern social media have put long-form conversation and thinking to death; they have turned the empowering nature of technology to the wrong ends, prioritising instant gratification and “brainrot” to grab people’s attention. I worry that we might be heading towards a generation of degradation, where people slowly lose the ability to think and to engage in in-depth, complex deliberation about the topics that concern us. These were exactly the platforms we ran into when looking for a place to discuss education. People refused to read long content, replacing the great works of past thinkers with “AI digests” and YouTube videos (possibly AI-generated); they evaded complicated issues, substituting difficult discussions and “awkward” thoughts with mindless scrolling and easy thinking. It reminds me of what Thoreau wrote in Walden:
A man, any man, will go considerably out of his way to pick up a silver dollar; but here are golden words, which the wisest men of antiquity have uttered, and whose worth the wise of every succeeding age have assured us of; — and yet we learn to read only as far as Easy Reading, the primers and classbooks, and when we leave school, the “Little Reading”, and story-books, which are for boys and beginners; and our reading, our conversation and thinking, are all on a very low level, worthy only of pygmies and manikins.
Yes, the Internet has given us a place where knowledge and information can be stored and retrieved more accessibly and reliably than ever. But what we also need is “technology for human flourishing”, as I wrote in that same earlier post: a push for people to turn back to the difficult books in which those “golden words” lie, and to put real thought into the difficult problems that trouble our society.
While society has charged head-on into building ever smarter machines that can reason and think like humans, from the machine learning algorithms that power social media to large language models, humans themselves might have lost the ability to think and reason deeply.
Aristotle argued, in his Nicomachean Ethics, that the purpose of a being (its final cause, in his famous framework of the four causes) lies in performing the function most special to its kind, and that the distinctive function of human beings is the exercise of reason — to flourish as a human being is to exercise it well. We must be careful not to hand that purpose off to authorities, to outside influences, and, in our time, to AI.
This is where we began. We needed a place for people to gather around real-world topics that concern the wellbeing of our society and culture, and a mechanism that would let deliberation flourish again — so that we can cultivate a generation of thinking humans, not just thinking machines.
Building Deliberation Platforms
At the same time, my friend and I were working with a group of researchers and scientists at MIT who were investigating how AI could be applied to foster collective intelligence. Collective intelligence, simply put, is the capability, and the emergent phenomenon, of a group of people thinking together and achieving a higher intelligence as a group, or “swarm”, than any single individual could. However, I believed that collective intelligence doesn’t arise spontaneously; it requires some kind of “facilitation” to come into being, be it institutions, mediums, or simply the education of its participants. We were particularly interested in the potential of AI to amplify such facilitation. I will skip over the technical details, but what we focused on building was a deliberation platform integrated with AI. We named it EverythingHuman, with the vision that it would be the place where we could discuss everything related to humankind: every problem, every novel idea, every piece of knowledge we come to pursue and possess. This, we believed, was what we needed to start a space for thinking humans, and to begin tackling the problems and discussions we cared about, like that of higher education.
Deliberation is, in fact, a very old idea. Every student of introductory philosophy meets the “Socratic method”, one of the greatest gifts of the ancient Greeks: the pursuit of truth through intense critical thinking, by asking a series of questions and focusing on the challenges they raise. Deliberation draws on this spirit of questioning, as well as on the citizens’ assemblies of Athenian democracy, where public matters were argued out and decided together. It is the process of evidence-based, reasoned discussion through which people weigh a complicated question together, aiming not to win an argument but to reach better judgements. In modern times especially, it has been applied to complicated social problems, to drive “systemic change” — change that involves multiple stakeholders, parties and points of view, often crosses national or cultural boundaries, and reshapes our institutions and social frameworks. These are also known as “wicked problems”: climate change, the modernisation of our education system, the challenge of building sustainable and happy urban lives as metropolises grow. They are questions that require a true deliberation process, and they were the ones we hoped our platform could help make progress on. This is also where EverythingHuman touches on governance: how a society deliberates over its wicked problems, and who gets to take part in that deliberation, lies at the heart of how it governs itself. A deliberation platform, then, is a natural medium for exercising collective intelligence and solving complex issues together.
I think that AI will be the next foundation for deliberation platforms. For the first time in history, we have something that can match or exceed us at certain intellectual tasks — analysing large amounts of data, validating and critiquing arguments, connecting thoughts and ideas — and that would prove invaluable in driving a modern deliberation platform. We called this vision “Human Superintelligence”, or HSI for short. It is commonly said that AGI and ASI describe AI that matches or surpasses human ability across a broad range of tasks, and the problems they are usually imagined solving, in medicine, the natural sciences or engineering, tend to have well-defined goals. HSI, on the other hand, is designed for human-centred, social problems: “wicked problems” that have no clearly defined solution or goal. These problems are so interwoven with our values, morals and culture that I doubted whether conventional AGI/ASI would be suited to them. Take the case of modernising our education system. AI could analyse sources and offer plenty of alternatives and arguments. But I doubt it could make good decisions, or come up with anything novel and useful, without some form of human engagement — after all, it has never been to school. It has no way of knowing what students actually struggle with day to day, what people really need in order to learn better, or what stakeholders at universities are actually concerned with. Unlike technical questions, these are grounded entirely in personal context, often have no right answer (if an answer at all), and any attempt to answer them will differ with values and culture.
Collective intelligence, on the other hand, has long been how we tackle such questions: by bringing together thoughtful people who contribute their personal experiences, their identities, their whole selves. My hypothesis is that even though AI cannot solve these questions directly, it should be able to help “scale” (and facilitate) collective intelligence, making it more powerful than ever. The synthesis of the two fields led to the vision of HSI, which I hoped would finally extend the advances in AI to the most critical questions of all — those that directly shape the society we live in.
Education
Yet what made EverythingHuman more meaningful to me is how it promised to redefine how and what people learn, shedding light on a new form of education for the modern age. The two are closely linked: to deliberate is itself to learn — actively, by thinking, questioning and engaging with others, rather than passively receiving. Traditional schooling has been challenged for decades, and recently people have begun questioning the value of higher education, at least in its present form. I have a vision that, instead of doubling down on the traditional “institutional” approach, the future of education is decentralised and distributed: learning is no longer bound to the explicit act of “teaching”, in a specific location or environment, for a predefined curriculum, within a set timeframe, with a fixed group of students. The educational facet of EverythingHuman presents this alternative view — an educational “web” for learning.2
Let us begin from the first principles of learning. Learning almost anything, from a technical skill to the humanities, requires four kinds of resources. First, the things we learn with: textbooks, laboratories and equipment, but also the Internet and other learning tools. Second, peers: people to learn alongside. In school these are your classmates, but nothing restricts a peer network to a classroom — anyone who shares your interest in and dedication to learning the same thing, however different you may be otherwise, can be a peer. Third, models: practitioners you learn a skill from, like your mentor in the lab or at your internship (or, in the past, the master to an apprentice). And fourth, educators: people who don’t just teach you skills, but guide you as you explore what to learn and how to learn it. I would imagine these four sufficient for learning, and, as it turns out, so did Illich. His proposed “learning webs” consist of four networks that map almost exactly onto them: services that give access to educational objects, skill exchanges that connect learners with people willing to serve as models, peer-matching for finding fellow learners, and directories of independent educators.
EverythingHuman aims to establish a network where anyone interested in any area can start a discussion, connecting passion with expertise and knowledge, and facilitating learning through engagement with peers who share the same interest, while drawing on the vast resources of the Internet, the intelligence of AI, and the knowledge of human experts. This, in my opinion, is a much more efficient structure for learning than most schooling today.
Such a platform is not a replacement for physical schools, but an extension that complements them, helping learning move outside of, and in parallel with, “schooling”. It is a space where learners have more freedom over what, how, and with whom they learn. Many students, especially at research-focused universities, know the experience of courses designed by “textbook experts” and delivered in dull lectures where few pay attention, by professors whose priority often lies more with their research than with their teaching. What I imagine alongside this is giving learners access to peers and models, people genuinely interested in what they do and eager to share it, connecting them with other learners who are not necessarily enrolled in the same course, in the same year, in the same department, but anywhere in the world with a shared interest, and providing resources that empower critical thinking, long-form discussion and problem-solving.
One of the purposes of a modern education system is to raise civically responsible, active citizens who take part in the democratic process, and EverythingHuman set out with the same goal. Beyond that, I believe a model of active rather than passive learning — time spent not listening to lectures but actively thinking and deliberating — would not only contribute to a healthier civic society, but also help train thinkers who won’t simply consume, and won’t let the thinking machines do their thinking for them. It offers another way of looking at how education might be organised: something not only universal and equal in appearance, but genuinely efficient and empowering.
Closing Thoughts
None of this is meant as a blueprint, or as a claim about how things should be done. I wrote down these ideas by reflecting on EverythingHuman as a “thought experiment”, where countless questions arose regarding thinking, deliberation, education, questions for which I’m still for searching an answer. I hope that sharing these ideas might start a few conversations of the kind EverythingHuman was built for — long, thoughtful, and human.
Footnotes
-
The contents of this essay represent only my personal viewpoints, not those of anyone else who has been involved in the project. We have stopped working on EverythingHuman for a while, for various other reasons, but I think the ideas are still worth sharing. ↩
-
The term “educational web” is borrowed from Ivan Illich’s Deschooling Society, which I only read after my work on EverythingHuman. The term resonated with what I had in mind — and, as I describe below, so did much of the rest of his framework. ↩