简体中文
English
Website review
This review may include AI-generated content and is not individually verified. Domain, certificate and other facts come from the observed data on this page.
What is Mark Carrigan?
Mark Carrigan is the personal website and blog of an academic sociologist and higher-education commentator. The site's tagline, "Raiding the inarticulate since 2010," frames it as a long-running space for essays and reflection rather than a corporate or institutional site. The homepage reads as a running feed of posts, so its main value is following his thinking over time, not browsing a polished product.
What you'll find there
- AI and higher education. This is the dominant theme in the page evidence, including a book announced as open access, The AI Crisis in Higher Education: A Strategic Proposal, plus posts on assessment integrity, why prompting shouldn't be the focus when teaching students about AI, and the risks of integrating AI too rapidly into universities.
- Scholarship and publishing. Posts on scholarly publishing, the future of the university, and using LLMs to write papers.
- Teaching practice. Notes on low-tech and no-tech pedagogy to protect foundational knowledge from "cognitive outsourcing."
- Personal and cultural writing. Reading logs, techno, marathon training, and more diaristic posts sit alongside the academic material.
Who it suits
| Reader | Likely use |
|---|---|
| University staff and policymakers | Following arguments about AI, assessment and institutional strategy |
| Researchers and doctoral students | Perspective on writing, publishing and academic life |
| General readers | Essays and personal posts, mixed in with the professional material |
Trade-offs
The site is a blog, so it's chronological and wide-ranging rather than organised by topic. If you want a single sustained argument, the book announcement is the clearest entry point; if you want to sample his voice, the shorter essays are quicker. Expect opinion and provocation rather than neutral reporting, and expect the personal and professional threads to be interleaved.
Next step: start with the open-access book post if your interest is AI in higher education, or the low-tech pedagogy post if you teach and want something immediately applicable. For a broader academic perspective on similar themes, LSE and Times Higher Education publish related commentary.
What does Mark Carrigan write about regarding AI in higher education?
Mark Carrigan writes about AI in higher education as a systemic problem rather than a purely technical one. His recent open-access book, The AI Crisis in Higher Education: A Strategic Proposal, frames it as a crisis for institutions—not just a new tool to adopt. His posts return repeatedly to assessment integrity, the normalisation of LLM use, and how universities should ration or govern internal access to these systems.
Several recurring themes stand out:
- Assessment and integrity. He tracks how AI use complicates what exams and written work actually measure, and why integrity debates tend to move through phases rather than resolve at once.
- Cognitive outsourcing. He argues for "low tech, no tech" pedagogy that protects foundational knowledge from being quietly delegated to models.
- Institutional strategy. Rather than treating AI adoption as inevitable, he questions why vendors target public institutions and how universities should decide what to integrate and how fast.
- Scholarship and publishing. He examines whether LLM-assisted writing helps academics get published (his answer: generally not) and how AI interacts with scholarly publishing and "enshittification."
- Student judgement and welfare. He asks what judgements students make while using AI, and notes the rapid normalisation of LLMs for emotional support—an area where the risks are less about cheating than about dependency.
- Prompting as a distraction. He argues that teaching students to prompt well misses the point compared with understanding what these systems do to thinking.
A practical next step: if you work in teaching or policy, start with his assessment-integrity and "low tech" pieces before the strategic book. A useful criterion is whether you're trying to police AI use or redesign what you're asking students to demonstrate—his writing is far more useful for the second task.
For related institutional and scholarly context, see Mark Carrigan and, on broader higher-education technology debates, Times Higher Education.
How can I subscribe to Mark Carrigan's blog or newsletter?
Subscription options appear on the site itself; the exact mechanics aren't detailed in the page evidence, so check the site directly for the current signup route.
Mark Carrigan
What to expect once subscribed
Carrigan's blog is a long-running academic and intellectual notebook. Recent posts mix:
- AI in higher education, including assessment integrity and "cognitive outsourcing"
- Scholarly publishing, enshittification and the future of the university
- Writing, productivity and creativity workflows
- Personal and cultural writing, including music and marathon training
If you're a researcher, lecturer or doctoral student thinking about how generative AI is reshaping teaching and research, this is the relevant audience. If you want practical tool tutorials rather than commentary and argument, it will likely feel indirect.
A useful next step
Before subscribing, scan the recent post list on the homepage and open two or three items that match your interests — for example, the low-tech pedagogy piece or the assessment integrity discussion. That tells you whether the posting rhythm and register suit you. If you want a lower-commitment test, an RSS reader works well for a blog like this, since you can follow it without handing over an email address.
Decision criterion
Subscribe if you want sustained critical commentary on AI and academic life from someone writing in that field. Skip it if you need a curated digest, structured courses or vendor-neutral tool reviews; this reads as a working blog, not a newsletter product.
What are Mark Carrigan's views on using LLMs for academic writing and publishing?
Mark Carrigan treats LLMs as a genuine but double-edged presence in academic work: useful for thinking and drafting, risky when they substitute for the writer's own judgement, and largely ineffective as a shortcut to publication. His writing on the topic is sceptical of tool-first teaching and of the assumption that speed gains for individuals translate into gains for institutions.
Where he draws the line
- Writing help is not publishing help. One of his post titles states the point directly: using LLMs to help write papers does not help you get them published. The bottleneck in academic publishing is peer review, novelty and fit, not prose production, so faster drafting does not remove the actual obstacle.
- Prompting is the wrong focus for teaching. He argues that teaching students about AI should not centre on prompt technique. Prompting is a narrow skill tied to current interfaces; the durable questions are about judgement, verification and what counts as one's own contribution.
- Don't be a meat proxy. His phrase for the failure mode where a person relays model output without adding thought. The academic version is a paper that has passed through a human but was not really written by one.
- Cognitive outsourcing and foundational knowledge. He links low-tech and no-tech pedagogy to protecting foundational knowledge, suggesting that some learning has to happen without a model in the loop.
- Organisational versus individual gains. Saving time for one academic does not mean the institution saves time. Work tends to expand, and the released capacity is rarely banked.
A concrete reader scenario
Suppose you are a doctoral researcher with a chapter draft and a deadline. Carrigan's position implies a specific order of operations: use the model to interrogate your argument, find where your claims are underspecified, or generate counterpositions you then have to answer in your own voice. Do not use it to produce text you would struggle to defend in a viva, and do not assume a polished draft moves you closer to acceptance. The measurable gain is in the quality of your thinking, not the word count.
How to decide for your own case
| Situation | Reasonable use | Carrigan-style caution |
|---|---|---|
| Early framing and literature mapping | Testing your framing against objections | Treat outputs as prompts for your own reading, not as sources |
| Drafting prose | Restructuring sentences you already wrote | Avoid text you could not reconstruct from memory |
| Editing and proofreading | Consistency, clarity, tone | Style uniformity is not the same as a better argument |
| Submission and review | Almost none | Reviewers respond to contribution, not fluency |
For a fuller picture of his reasoning, including work on assessment integrity, generative AI in higher education and the pace of institutional adoption, start at Mark Carrigan. A useful next step is to read his book on the AI crisis in higher education, which he has made available open access, then compare its strategic argument against your own institution's AI policy before deciding what to adopt in your teaching or writing practice.
What book recommendations or reading lists does Mark Carrigan share?
Mark Carrigan shares reading mainly through an annual roundup rather than a themed or curated list. On his site the recurring feature is "Things I've read in [year]" — for example, Mark Carrigan currently shows "Things I've read in 2026." That is the closest thing to a standing reading list: a running, chronological record of books he has finished, which you can browse to see what he is actually reading rather than what he recommends in the abstract.
The rest of the page is a stream of posts on AI, higher education, assessment, scholarly publishing and music, so book mentions tend to appear inside essays rather than in recommendation posts. His own book, The AI Crisis in Higher Education: A Strategic Proposal, is announced there as open access; that is his writing, not a recommendation.
How to use it
- Start with the "Things I've read" post for the current year and work backwards through earlier years to build your own list.
- Treat it as a reading log, not a syllabus: entries reflect his interests in social theory, technology and philosophy, so expect that slant.
- If you want the argument behind his reading choices, read the essays alongside the log — the AI-and-universities posts are where he explains what he thinks the literature gets wrong.
Decision criterion: if you want a broad, general-interest book list, this is not that. If you want a social-theory-and-technology reading trail from someone working on AI in higher education, the annual roundups are the right entry point.
What does Mark Carrigan say about low-tech or no-tech pedagogy to protect foundational knowledge?
Mark Carrigan argues that when students outsource foundational thinking to AI, they can lose the very cognitive capacities a university education is meant to build. His post "Low tech, no tech pedagogy to protect foundational knowledge from cognitive outsourcing" treats low- and no-tech teaching not as nostalgia but as a deliberate safeguard: some learning has to happen without a model doing the work, or the knowledge never becomes the student's own.
His related arguments reinforce the same point from different angles. He questions making prompting the focus of AI teaching, suggesting that skills in querying a model are not the same as understanding a subject. He also notes that using LLMs to help write papers does not help you get them published, and warns about integrating AI too rapidly into higher education.
What this looks like in practice
- Protected unaided work: tasks done without AI, so students practise recall, reasoning and composition themselves.
- Sequencing: foundational skills first, AI-assisted work later, once the basics are secure.
- Assessment design: formats that make cognitive outsourcing harder, rather than assuming detection will catch it.
- Explicit norms: telling students which tasks are no-tech and why, instead of leaving it to individual judgement.
These are his positions on pedagogy and higher education, not a packaged curriculum or a product.
Who this suits
Course leaders, teaching-focused academics and learning designers weighing where AI belongs in a programme. It is less useful if you want step-by-step lesson plans; the value is in the argument about what is at stake.
Trade-off to weigh
No-tech work costs time and can feel like a step back to students used to instant answers, and it sits awkwardly beside institutional pressure to adopt AI. Carrigan's case is that the alternative—graduates with credentials but weakened foundational knowledge—is worse.
A practical first move: pick one foundational skill in your module and designate one assessed task as no-tech, then ask students afterwards whether the constraint changed how they studied. For more of his writing on AI, scholarship and academic life, see Mark Carrigan.
Google Workspace
User reviews (0)