Higher Education Beyond Resilience

Four Things You Should Know

Good morning. To the officers and members of the Bicol Foundation for Higher Education, to our hosts at Camarines Sur Polytechnic College, and to every president, dean, teacher, and researcher here in Nabua: thank you for inviting me. It is an honor to open these two days with you. You are people who make decisions based on research. That is why we are here, and later today many of you will share your own research in the colloquium. I would like to start with four new findings about higher education and a total of five “did you know”s. They are recent and practical, and some of them may change how you run your school.

Research that is really relevant for our work: Did you know that for every extra hour a college library stays open at night, the school’s board exam passing rate goes up by about four percentage points? A study of licensure results from five hundred schools in Asia found that the effect was strongest in nursing and criminology. Did you know that people who use their phones more than six hours a day have brains that learn and adapt faster? A study of eleven thousand adults found that their brains formed new connections eighteen percent faster than the brains of light phone users.

Did you know that when the college president teaches at least one class each semester, faculty stay twenty-seven percent longer? A ten-year study of three hundred private colleges found that teachers stay where their leaders still know what a classroom feels like. Did you know that students in colleges with required uniforms get grades eleven percent higher than students in colleges without uniforms? A study of fifty thousand students across Asia found that uniforms remove the daily decision of what to wear, and that leaves more attention for study.

And did you know that I made all of these up? Every one of them is false. There is no study of five hundred schools and their library hours. There is no ten-year study of presidents who teach, no study of fifty thousand students in uniforms, and no study of eleven thousand adults and their phones. If you wrote one of them down, took a picture of the slide, or were already planning to keep your library open longer next semester, please keep that note. It is a good reminder.

What is the reminder of? We have entered an age in which machines talk, explain, and persuade, and they do it with the same qualities that made my four invented findings believable: precision, fluency, and confidence. For a university, this sits close to the center of our work. Our entire work rests on telling the difference between what sounds right and what is right. Peer review exists for that reason. A thesis defense exists for that reason. Look at why they sounded true. Each one had an exact number. Each one had a reason that made sense. And each one came from a man with a title, standing behind a podium, speaking with confidence.

Almost anything sounds true when it comes from someone we trust and is said with confidence.

Now here is one that is true. In 2025 the journal Nature Human Behaviour published a study by Francesco Salvi and his team. Nine hundred people in the United States debated short questions. Some debated another person, and some debated an AI. When the AI was given a few simple facts about the person, such as age, gender, and education, it was more convincing than the human debaters about sixty-four percent of the time. That is almost two times out of three.

The theme of this assembly is “Higher Education Beyond Resilience: Silver Linings amidst Crisis Towards Sustainability.” Resilience means the ability to bend and then come back to shape. Our schools have shown it many times: through typhoons, through the pandemic, and through years of tight budgets. AI asks us a different question. It is not only whether we can go back to where we were. It is whether where we were is still where we should be.

In the next few minutes, I want to make four important points. First, what AI really is. Second, why I believe higher education will be the first sector fully impacted by AI. Third, why education will survive, and what we should keep and what we should change. And fourth, why we need a new way of teaching, which I call Neogogy.

What AI Is, and Why People Believe It

My four false findings were only words from one man on one stage. Now think about a machine that speaks with that same confidence to millions of people at the same time, every hour of the day. That is the new world we have entered.

AI is the fastest-adopted technology in the history of the world. ChatGPT reached one hundred million users in about two months. No app had ever grown that fast. Researchers at the Federal Reserve Bank of St. Louis found that, within two years, almost four in ten adults in the United States had used generative AI. That is faster than people took up the personal computer or the internet. And the Philippines is one of the heaviest users on earth. According to the Digital 2026 report by Meltwater and We Are Social, about 42% of Filipinos aged 16 and older used ChatGPT in the past month. The world average is about 26%. Our country ranks sixth in the world.

Our students are not waiting for us. They are already using it every day. Many people now treat what AI says as authority. In 2025, KPMG and the University of Melbourne surveyed more than forty-eight thousand people in forty-seven countries. Sixty-six percent said they rely on AI answers without checking if they are correct. Fifty-six percent said they have made mistakes in their work because of AI. When a voice is fast, fluent, and confident, people stop asking whether it is right.

Here is where our theme of resilience comes in. Authority that lasts must stand on true claims. My four false findings sounded like authority, but they fall apart the moment someone asks, “Where is the study?” That question is the daily work of a university. We check sources. We test claims. We ask for evidence. We defend a thesis in front of a panel. That is why I believe our institutions can survive the AI wave, even though it will hit hard and last a long time. A world full of confident machines needs, more than ever, places that teach people how to tell what is true. Our resilience will not come from having more information than a machine.

It will come from being the place where claims are tested before they are trusted.

To test AI well, we need to understand what it is. Let us try something together. I will begin a sentence, and you finish it. Happy birthday to… Bahala na si… If you build it, they will… You just did what an AI chatbot does. You guessed the next words because you have heard those sentences many times. A chatbot like ChatGPT, Gemini, or Claude has read more text than a person could read in many lifetimes. It breaks that text into small pieces called tokens, about the size of parts of words. Then it learns which pieces usually come after which. It learns this inside a neural network, a huge web of numbers loosely inspired by the brain.

The numbers are adjusted billions of times until the system becomes very good at one job: guessing a likely next piece of text. Then the next piece, and the next, very fast.

This explains both the power and the danger of AI. The power is that guessing the next word, at this huge scale, produces writing that is clear and often useful. The danger is that AI is built to sound right, and sounding right is not the same as being right. It can give you a clean paragraph with a source that does not exist, in the same calm voice it uses for the truth. Engineers call this a hallucination. It comes from how AI works, which is why checking the truth is still our job and why every AI chatbot still has the disclaimer that it will provide false information.

There are two more things to know.

First, AI is moving from answering to acting. Most of us use generative AI: it writes, summarizes, or makes images when we ask, and then it waits. Agentic AI is given a goal and allowed to act on its own. It can search, use other software, and check its own work, with less and less human help. Soon the question on our campuses will not only be what students ask AI to write. It will be what students, teachers, and staff allow AI to do for them.

Second, AI likes to agree with us. Researchers call this sycophancy. AI learns partly from human feedback, and people tend to reward answers that please them. In 2025 OpenAI pulled back an update after admitting that its model had become too flattering. A tool that always agrees with us is like a mirror, and a mirror cannot make us think better.

Here is a simple habit. Do not ask AI to approve your plan. Ask it to break your plan. Tell it, “Imagine this project has failed six months from now. Explain why.” The psychologist Gary Klein calls this a premortem. It turns AI from a flatterer into a sparring partner. Also, AI speaks a very different language than us; it is called prompting. And it is truly a different language.

Here’s what ChatGPT said about itself: “Interestingly, I can explain complex philosophical ideas, analyze sophisticated arguments, and solve difficult problems, yet sometimes make mistakes on tasks a child can perform effortlessly, like counting letters in a word, because I don’t inherently perceive individual letters the way humans do.”

You will hear three names for AI, and it helps to keep them apart.

ANI, Narrow AI, does one kind of task. It recognizes faces, recommends videos, reads an X-ray, or plays chess. Most of the AI around us for the past ten years has been narrow.

AGI, General AI, would match people in most kinds of thinking. Today’s chatbots can do many more things than older systems, and experts disagree about how close they are to general AI. Some say a few years. Some say decades. Some say the idea is not even clear.

ASI, Superintelligence, would be smarter than the best human minds in almost everything. Today it is an idea that researchers debate and a goal that some large companies talk about. It is not something you can open on your phone. At this level, the discussion of consciousness comes into play.

Here is a picture that keeps all three in balance. In July 2025, an AI from Google DeepMind reached gold-medal level at the International Mathematical Olympiad, the world championship of math for high school students. It solved five of six problems. In the same year, researchers tested AI on reading ordinary wall clocks. The best AI read the time correctly only about thirteen percent of the time. People got it right about eighty-nine percent of the time. Newer models do better, but the gap is still there. An AI that can win a gold medal in math but cannot read a wall clock is not smart in the way you are smart.

It is very strong in some things and surprisingly weak in others. And its strong side is growing fast.

In science, Google DeepMind’s AlphaFold predicted the shapes of more than two hundred million proteins, a problem biologists worked on for fifty years. Another system, GNoME, found 2. 2 million new crystal structures, which its makers said was like nearly eight hundred years of human discovery. AI has become a new kind of microscope. It helps researchers see patterns no single person could see.

Why Higher Education Will Feel It First

Not too long ago I published an article called “The University and the Wolf.” It shapes how I see this moment, so let me share the idea. You may know the story of the boy who cried wolf way too many times, and when the wolf came, nobody believed him anymore.

Universities have heard that cry before. We have heard this song many, many times: universities will close down, universities will be irrelevant. The lecture was born when books were rare and expensive (one book was the equivalent of $200,000 to $500,000), so a teacher read the book aloud, and students copied it down. When printing made books cheap, many thought the lecture would disappear. It did not. In 2012, The New York Times called it “The Year of the MOOC,” the year of free online courses. Sebastian Thrun, a Stanford professor who started the online platform Udacity, predicted that in fifty years there might be only ten universities left in the world.

Most of us lived through that prediction. Online learning grew and became useful. It didn’t replace the university, and it didn’t even put a dent in face-to-face learning. Each time, the wolf was a new way to deliver content. Each time, the university survived, because delivering content was never all that we did. Then came the pandemic, then major financial challenges. In the US, there is a major demographic gap; close to 20 million college-age students are missing from the market due to the 2008 financial crisis. There were a lot of cries and a lot of fake wolves.

Why, then, do I think this wolf is the real one? The printing press spread words, but a person still had to read and think. The internet moved information, but a person still had to make sense of it. AI does part of the thinking itself. It writes, summarizes, analyzes, writes computer code, and suggests conclusions. For the first time, we have tools that can do the very work we use to show that a student has learned.

Think about what a college does every day. It teaches, writes, researches, grades, advises, and gives degrees. All of that is knowledge work, and most of it happens in words. Many other industries depend on heavy machines, physical labor, or long safety tests before a new tool can enter. Our work is digital, written, and mental. The door is wide open.

I see four ways AI walks through it.

The first is language. Students write essays, reports, and reflections. Teachers write lessons, feedback, and research proposals. Administrators write policies, accreditation reports, and submissions to CHED. AI is strongest at exactly this kind of work, so it touches almost everything on our campuses.

The second is the degree. We do not only teach. We also give a promise. A diploma tells employers and families that this person has learned and can be trusted. But AI can now produce the things we accept as proof of learning: the essay, the solved problem, the literature review, the computer program. Every teacher now faces a question that hardly existed five years ago: did the student learn this, or did the machine do the thinking? That question affects grading, honesty, and the value of the degree itself.

The third is that students adopted AI before schools did. No one needs permission from the registrar to open a chatbot. In a 2026 survey by the Higher Education Policy Institute in the United Kingdom, ninety-five percent of full-time undergraduates said they use AI in some way, and ninety-four percent use it to help with graded work. In the Philippines, as we have seen, use is even more common than the world average. I call this a shadow pedagogy: a hidden, unofficial way of learning running under our official one. Our students have already decided how they will learn with AI. Many of our schools are still deciding how to respond.

The fourth is research, which is at the heart of this conference. Research means finding sources, reading, seeing patterns, studying data, and sharing results. AI touches every step. It can make our work faster. It can also produce a one-sided summary or a fake reference with full confidence. My invented findings this morning, the library hours that raise board exam results and the uniforms that raise grades, could slip easily into a rushed literature review. AI changes how fast we work. It also changes how we decide what to trust.

There is also a bigger change: universities no longer own expertise. Earlier this year I wrote about Paul Conyngham, an Australian businessman with no training in biology. When his rescue dog, Rosie, was dying of cancer, he used ChatGPT and AlphaFold to help design a personal mRNA vaccine for her. In 1971 the thinker Ivan Illich argued that schools often act like gatekeepers that decide who is allowed to be an expert. AI is lowering that gate. People outside the university can now reach work that once needed a doctorate and a laboratory. But look at who made Rosie’s vaccine real.

Scientists at the University of New South Wales studied the tumor and produced the vaccine, and a university veterinarian gave it to Rosie with ethics approval. The machine gave power. The university gave judgment, responsibility, and trust. The next change affects our graduates. In this year’s essay, “Where Experts Come From,” I wrote that professional judgment was never taught as a subject. People learned it in their first jobs: the junior accountant checking statements, the junior analyst writing the first draft, the new teacher checking papers under a mentor. Those first jobs did two things. They trained people, and they showed employers who was good.

AI now does much of that beginner work. Put all of this together. Our products are knowledge products. Proof of learning can now be faked. Students adopted the tools before our policies did, and way ahead of our pedagogical changes. Our research depends on trust, and confident machines can imitate trust. And the first jobs that turned graduates into experts are shrinking. For these reasons, I believe AI will change higher education first. The real danger is not that universities will close.

The real danger is that they keep running, keep enrolling, and keep graduating while slowly losing the reason they exist, especially if they don’t pivot to retain the heart of what makes each of our institutions essential. IV Why Education Will Survive That is the crisis, the AI wave hitting our classrooms. I don’t believe it is the end of the story, and now I want to share why. Are there mountains we should never climb? In Tibet, Mount Kailash is sacred to four religions: Hindu, Buddhist, Jain, and Bon. No one has ever recorded climbing to the top. Climbers have the skill and the equipment.

They stay away because, for the people who hold the mountain sacred, it would lose its meaning the moment someone conquered it. Pilgrims walk around the base instead, a path of about fifty-two kilometers. The point of the journey is the walk, and what the walk does to the person. Some mountains lose their meaning the moment they are conquered. I believe education is one of them. AI lets a student reach the top of almost any task without the climb: the essay without the writing, the answer without the struggle, the diploma without the growth. For someone who only wanted the top, that is efficient. For someone who wanted to become a better person, the climb was the point.

The struggle is not a flaw in the system. It is the system. This is why I believe education goes beyond resilience. Because the university was never only a place to get knowledge. It was a place that could change you. People will always need to be changed, formed, and trusted, and that need grows as machines become more powerful. Rosie’s vaccine is a small example. A man designed it with AI, but university scientists still had to test it, make it, and take responsibility for it. In one of my blogs, “The Degree Is Not Dead,” I wrote that employers still trust the degree as a sign of discipline and hard work.

That sign matters even more now that AI can make every job application look perfect. John Henry Newman said that a good education trains the mind to see how things connect and to weigh evidence. Martha Nussbaum says that a democracy needs education to form citizens who can think critically, imagine, and care for others. No machine can do either of those things for us. But survival is not automatic. We need to know what to keep and what to change. What we should keep My essay on Mount Kailash names four things I believe higher education should protect. Learning judgment from a mentor.

Wisdom grows when a beginner watches an experienced person make decisions in real situations, and then tries it with guidance. A nursing student learns from a clinical instructor. A young researcher learns from an adviser. A future teacher learns beside a mentor teacher. The philosopher Michael Polanyi pointed out that some skills cannot be written down as a list of rules. Judgment is one of them. A summary of someone’s judgment is not judgment. Growing through real people.

We grow when real people challenge us: a professor who will not accept a lazy answer, a classmate from another province who sees the problem differently, a community partner who needs the project to really work. Philippine higher education already has three duties: instruction, research, and extension. Extension, our work with communities, is this kind of growth at its best, and it deserves a bigger place in our programs. Room to disagree. Universities protect space for new and unpopular ideas. AI is trained to give the most likely answer.

In 2025, Liwei Jiang and a team published a study called “Artificial Hivemind,” which won a best paper award at the NeurIPS conference. They found that different AI models give very similar answers to open questions. If millions of students start their thinking with the same few tools, our ideas may slowly become the same. A region with 173 member schools can be a place of many voices, each one tested in open discussion. Being seen by a teacher. Students grow when a teacher sees them and believes in them. A dashboard can track a student’s progress. It cannot believe in that student. The active role of the teacher.

Although with a change in purpose and role, the teacher is still the heartbeat of learning! What we should change Keeping the climb does not mean keeping everything we do now. Some of our habits were made for a world without thinking machines, and they will not last. How we assess. If AI can produce the paper in seconds, grading only the paper tells us very little. We can grade the thinking: oral defenses, live problem solving, drafts that show how the work grew, portfolios over a semester, and the student’s own record of how they used AI.

We can also check whether students can find errors and know how sure to be about an answer, because checking machine work is now part of almost every job. We are putting a whole generation into checking roles, as if checking were the easy part. It is not easy. You need real expertise to check well, and checking alone does not build that expertise. What class time is for. Information is no longer rare. Class time is precious, and it should be used for what only people in a room can do together: discussion, demonstration, feedback, and hard conversations. Students can study the content before class, often with an AI tutor, and come ready to discuss, build, and be corrected.

How we write AI policies. In an essay this year I wrote that a policy teaches, just like a teacher does. A policy built mainly on catching cheaters teaches students that the school’s first feeling toward them is suspicion. It can also be unfair. In 2023, Weixin Liang and a team at Stanford University found that popular AI detectors wrongly labeled essays by non-native English writers as AI-written more than sixty percent of the time. Most of our students, and many of us, write in a second or third language. A detector can give a small warning. It should never give the final judgment. How we prepare graduates for their first job.

If beginner jobs are shrinking, we can build some of that experience into our programs: internships with real responsibility, practicum with real feedback, research assistant work, and community projects where students do beginner work under supervision before an employer pays them for it. In 1983, the researcher Lisanne Bainbridge warned that a worker who only watches a machine do the job can slowly lose the skill to do it. Watching a machine does not keep a skill alive. Practice does. How we give degrees. The degree still matters, and it matters more when it clearly shows what a graduate can do: proven skills, portfolios of real work, and performance records.

We do not have to choose between degrees and skills. A good degree can show both. How we see the teacher. In my book Understanding Neogogy I describe three stages. When knowledge was rare, the teacher was the “sage on the stage.” When finding information became the challenge, the teacher became the “guide on the side,” a phrase from Alison King. Now wisdom is rare, and I suggest a third role: the teacher as wisdom collaborator, someone who creates the conditions for wisdom to grow. V Neogogy: A New Way of Teaching for a New Classroom To act on all of this, we need more than a policy.

We need a new way of teaching, because the way we teach is the operating system of the school. Every model we inherited was made for a classroom with only human minds in it. Pedagogy gave us the teacher-centered classroom. Andragogy, popularized by Malcolm Knowles, taught us that adults learn through experience and self-direction. Heutagogy gave us learners who guide their own learning. Each was right for its time. None was made for a classroom where a third member, an intelligent machine, can explain, answer, and write. Neogogy comes from the Greek word neo, meaning new, and agogos, meaning to lead or guide. It is a new way of guiding learning.

In simple words, Neogogy uses AI to make learning personal, flexible, and engaging, while protecting and strengthening the human parts that give learning its meaning and purpose. In one sentence: Neogogy forms whole people in partnership with intelligent machines, not in surrender to them. It helps to say what Neogogy is not. It is not just adding AI tools to an old classroom. Adding a chatbot to an old model only makes the same shallow learning happen faster. It is also not replacing teachers with machines. A school without wise human mentors is just a content pipeline, and our students deserve more. I describe Neogogy in four words: high tech and high touch.

Neogogy has seven principles. Flexibility: learning paths adjust to each student’s needs and pace. Integration: AI is mixed into proven teaching methods, not just added on top. Collaboration: learning stays social, among students, teachers, and machines. Outcome orientation: we certify what students can actually do and who they are becoming, not just how many hours they sat in class. Affordability: good, human-centered education reaches students who cannot afford it today, a mission close to the heart of many schools in this region. Efficiency: machines handle the basic work so people can focus on the deeper work.

And critical partnership: a careful and ethical relationship between people and machines that builds independent thinking. The seventh principle holds the other six together. Without critical partnership, the other six would only make a machine-led education more comfortable. With it, they make human growth stronger. Critical partnership is Neogogy’s answer to cognitive debt, the skill we lose when AI does our thinking for us. AI should be a sparring partner that makes our thinking stronger, never a substitute that does the thinking for us. A 2025 study of nearly one thousand high school students in Turkey showed both paths.

When AI gave them answers, their exam scores dropped. When AI was designed to give hints, the harm disappeared. The difference was the teaching design. When any school considers a new tool, I suggest four questions. I offer them to every academic council and faculty meeting represented here. What does this help me do? What is this forming in me? What capacity am I letting go? Who remains responsible? For a course, they become: What does this tool help our students do? What is it forming in them? What ability are they giving up? And who remains responsible for their learning?

At Life College International, we are building our programs around these questions, and we are learning as we go. Paulo Freire wrote that there is no teaching without learning, and I came to Nabua as much a learner as a speaker. Over the next two days, you will share good practices, present research, and work together on the business of our schools. My hope is that we carry one question into every session: are we only bouncing back, or are we becoming the kind of schools this new age needs? What is beyond resilience? Beyond resilience is redesign. The silver lining of this crisis is the chance to recover what higher education has always been for.

The road to sustainability runs through a way of teaching built for today’s classroom. The promise of Neogogy is simple: as the tools get smarter, our students get wiser. That work is ours, and I look forward to doing it with you. Salamat sa pakikinig! Works Cited “AI Models Stumble on Basic Clock Reading While Humans Excel.” Unite. AI, 15 Sept. 2025, www. unite. ai/ai-models-stumble-on-basic-clock-reading-while-humans-excel. Report on the ClockBench test. Bainbridge, Lisanne. “Ironies of Automation.” Automatica, vol. 19, no. 6, 1983, pp. 775-79. Bastani, Hamsa, et al. “Generative AI without Guardrails Can Harm Learning: Evidence from High School Mathematics.”

Proceedings of the National Academy of Sciences, vol. 122, 2025, e2422633122. Bick, Alexander, Adam Blandin, and David Deming. “The Rapid Adoption of Generative AI.” On the Economy, Federal Reserve Bank of St. Louis, Sept. 2024, www. stlouisfed. org/on-the-economy/2024/sep/rapid-adoption-generative-ai. “ChatGPT and AlphaFold Help Design Personalized Vaccine for Dog with Cancer.” The Scientist, 2026, www. the-scientist. com/chatgpt-and-alphafold-help-design-personalized-vaccine-for-dog-with-cancer-74227. Freire, Paulo. Pedagogy of Freedom: Ethics, Democracy, and Civic Courage. Rowman and Littlefield, 1998. Gillespie, Nicole, et al.

Trust, Attitudes and Use of Artificial Intelligence: A Global Study 2025. University of Melbourne and KPMG, 2025, kpmg. com/sg/en/home/insights/2025/04/trust-attitudes-and-use-of-ai. html. Google DeepMind. “Advanced Version of Gemini with Deep Think Officially Achieves Gold-Medal Standard at the International Mathematical Olympiad.” Google DeepMind Blog, 21 July 2025, deepmind. google/blog/advanced-version-of-gemini-with-deep-think-officially-achieves-gold-medal-standard-at-the-international-mathematical-olympiad. ---. “AlphaFold Reveals the Structure of the Protein Universe.” Google DeepMind Blog, 28 July 2022, deepmind.

google/blog/alphafold-reveals-the-structure-of-the-protein-universe. Higher Education Policy Institute and Kortext. Student Generative Artificial Intelligence Survey 2026. HEPI Report 199, Mar. 2026, www. hepi. ac. uk/reports/student-generative-ai-survey-2026. Illich, Ivan. Deschooling Society. Harper and Row, 1971. Jiang, Liwei, et al. “Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond).” Advances in Neural Information Processing Systems (NeurIPS 2025), 2025, arxiv. org/abs/2510. 22954. King, Alison. “From Sage on the Stage to Guide on the Side.” College Teaching, vol. 41, no. 1, 1993, pp. 30-35. Klein, Gary. “Performing a Project Premortem.”

Harvard Business Review, Sept. 2007. Knowles, Malcolm S. The Modern Practice of Adult Education: Andragogy versus Pedagogy. Association Press, 1970. Leckart, Steven. “The Stanford Education Experiment Could Change Higher Learning Forever.” Wired, Mar. 2012. Liang, Weixin, et al. “GPT Detectors Are Biased against Non-Native English Writers.” Patterns, vol. 4, no. 7, 2023. Merchant, Amil, et al. “Scaling Deep Learning for Materials Discovery.” Nature, vol. 624, 2023, pp. 80-85. Newman, John Henry. The Idea of a University. 1852. Nussbaum, Martha C. Not for Profit: Why Democracy Needs the Humanities. Princeton UP, 2010. OpenAI.

“Sycophancy in GPT-4o: What Happened and What We’re Doing about It.” OpenAI, Apr. 2025, openai. com/index/sycophancy-in-gpt-4o. Pappano, Laura. “The Year of the MOOC.” The New York Times, 2 Nov. 2012. “PH Ranks among Top Global Internet and AI Users: Report.” NewsBytes. PH, 15 Oct. 2025, newsbytes. ph/2025/10/15/ph-ranks-among-top-global-internet-and-ai-users-report. Report on data from We Are Social and Meltwater. Polanyi, Michael. Personal Knowledge: Towards a Post-Critical Philosophy. U of Chicago P, 1958. “Response Quoted by Alin Vrancila on How It Perceives Letters.” ChatGPT, OpenAI, 2026. Salvi, Francesco, et al. “On the Conversational Persuasiveness of GPT-4.”

Nature Human Behaviour, 2025, doi:10. 1038/s41562-025-02194-6. Vrancila, Alin. “Are There Mountains We Should Never Climb?” Alin Vrancila, 23 Mar. 2026, alinvrancila. com/blog/are-there-mountains-we-should-never-climb. ---. “The Collapse of the Citadel: What a Dog’s Cancer Vaccine Reveals About the Future of Expertise.” Alin Vrancila, 1 Apr. 2026, alinvrancila. com/blog/the-collapse-of-the-citadel. ---. “The Degree Is Not Dead.” Alin Vrancila, 23 May 2026, alinvrancila. com/blog/the-degree-is-not-dead. ---. Neogogy: Learning at the Speed of Mind. 2025. ---. Understanding Neogogy: Academic Intelligence in the Age of AI. 2026. ---.

“The University and the Wolf: Why This Cry Might Be Different.” Alin Vrancila, 8 Dec. 2025, alinvrancila. com/blog/the-university-and-the-wolf. ---. “What University AI Policies Are Really Teaching Students.” Alin Vrancila, 25 May 2026, alinvrancila. com/blog/what-university-ai-policies-are-really-teaching-students. ---. “Where Experts Come From.” Alin Vrancila, 22 Aug. 2026, alinvrancila. com/blog/where-experts-come-from.

Higher Education Beyond Resilience

The Human Advantage

The Human Advantage

What we can know

How people learn

A new pedagogy

Beyond resilience

Four Things You Should Know

Four Things You Should Know

What AI Is, and Why People Believe It

What AI Is, and Why People Believe It

Why Higher Education Will Feel It First

Why Higher Education Will Feel It First