90 items with this tag.
Develop multidimensional capability with generative AI for academic work
Path is a CPD portfolio app that shows you what you're missing against the standard you're working towards, rather than only showing you what you've done. It began as a two-weekend prototype and has grown into a real, hosted web app that maps your claims and evidence against any framework — a promotion, a fellowship, a regulator's CPD standard — and shows you the single highest-value thing to do next. It's still a closed, invite-only alpha, and the waiting list is now open.
Most CPD tools are built for compliance, not development. I wanted something that started from the gap; what am I missing, and what would close it? I couldn't find the tool I needed so I built one called Path instead. It describes what I think professional development should look like, and what two weekends with Claude Code produced.
A one-page reference guide for health professions educators. Condenses a theoretically grounded framework into six design principles for integrating AI in ways that support — rather than undermine — the conditions under which professional learning occurs. Each principle includes a brief description and an apply-by prompt for immediate use in learning activity design, curriculum review, or institutional policy.
Tim Fawns published a LinkedIn post setting out 17 points about student use of AI, drawing on his experience as a researcher and educator. The points push back on the framing that positions students as cheating or avoiding learning, and ask for a more sophisticated understanding of what's actually happening. I've formatted Tim's LinkedIn post as a one-page PDF for easier sharing.
One in seven people in the UK are using AI chatbots for health advice instead of seeing a GP. The institutional response has been to warn them off, but that response applies a standard it doesn't consistently apply to anything else in the system. This post argues that the risk comparison driving those warnings is systematically skewed, and that a more honest accounting points toward an entirely different kind of response.
YAML is a human-readable format for storing structured data as plain text. In knowledge management and publishing workflows, it appears most commonly as the frontmatter block at the top of markdown files, where it holds metadata — title, author, date, tags — that tools can read without parsing the document itself.
Distributed version control is an approach to tracking file changes where every contributor holds a complete copy of the repository and its full history, rather than depending on a central server. It enables offline work, parallel development, and resilience against data loss.
Git is a distributed version control system that tracks changes to files over time. It records who changed what and when, allows you to move between earlier and later states of a project, and lets multiple people work on the same files without overwriting each other's contributions.
Language models can transform documents into interactive tools in minutes. This post walks through a concrete example, turning a 21-page Word questionnaire into a working web app, and reflects on what that capability makes possible.
AI-generated text is fluent regardless of whether its content is accurate or well-reasoned. Fluency was once a reasonable trace of genuine thinking — a student who wrote clearly had usually thought clearly. That relationship no longer holds. Worse, the AI literacy response of teaching output evaluation is a temporary fix: as models improve, output quality converges on expert-level across every artefact we care to measure. The question isn't how to spot current failure modes. It's what you'll do when those failure modes are gone.
Claude produced the word "contribuves" in a piece of writing, which is obviously not a real word. This is a different kind of error than hallucination, and the distinction matters.
Every week I annotate articles in Zotero, highlights in Reader, and podcasts in Snipd, all of which is synced to Obsidian. By Friday I have a week's worth of material, tagged and structured, but unreviewed. This post describes the weekly review command I built to surface what matters and create a reason to engage with it.
A presentation for students participating in an EU-funded Blended Intensive Programme at Thomas More Hogeschool in Belgium. Examines how AI separates the production of artifacts from the learning they were meant to evidence, what problem-based learning already does differently, how AI changes group work and inquiry, and three practical shifts students can make in how they use AI within PBL.
The structural features of an information source that enable its knowledge claims to be challenged, traced back to evidence, and evaluated against the source's track record. Traditional sources carry it; generative AI largely does not.
A headless AI model runs non-interactively — no chat interface, no conversation. You pass it text, it returns output, and it exits. This makes AI tools composable with the same scripts and schedulers that have coordinated Unix processes for decades.
Most advice on AI effectiveness focuses on prompt engineering. The real leverage comes from somewhere less obvious; knowing your professional commitments clearly enough to turn them into context an AI can work within. This post describes how to build AI personas for professional practice — structured documents that compress your values, frameworks, and evidence into a form an AI agent can actually use.
A field note on the time Claude deleted my file. The agent followed my instructions precisely and that was the problem. A reflection on a different kind of AI failure mode, and what the model's apology reveals about where responsibility actually sits.
I've been writing lecture slides in markdown for several years, mostly because I enjoyed working in structured formats and plain text. That decision turned out to matter in ways I didn't anticipate. When AI agents have access to your local filesystem, the format your teaching materials live in determines what's possible.
An internal staff development session for the CPC team introducing AI through Microsoft Copilot. Covers what AI is and isn't, safe working practices, structured prompting with the RGID heuristic, and hands-on practice — with the goal of each participant leaving with one specific task to try that week.
A field note on what the recent Claude outage revealed about where I am on the dependency curve, and what the difference between a session limit and an outage tells you about infrastructure.
Harness engineering is the practice of building the full architectural scaffolding within which AI agents operate — structured documentation they can reason with, constraints that enforce invariants, and feedback loops that let them know when they've succeeded. It is distinct from prompt engineering, which shapes individual tasks, and from oversight, which monitors outputs after the fact. The harness is the infrastructure that makes delegation coherent at scale.
The previous posts described what makes agentic workflows coherent at the individual level: a plan, documentation as infrastructure, and domain expertise that can evaluate outputs. Together, these form an informal harness; the conditions within which delegation stays accountable. At institutional scale, a personal harness is not enough: multiple people directing agents without shared constraints produce compounding drift that no amount of human oversight can track. This post examines what AI agent governance in higher education actually requires, and why a harness, not better oversight, may be the right frame.
Vibe coding describes using AI tools without maintaining genuine accountability for what they produce; accepting outputs without the scrutiny, direction, or judgement needed to evaluate and improve them. Simon Willison drew the key distinction: vibe coding versus vibe engineering, where the latter uses the same tools while remaining genuinely accountable for every output.
What does it actually take to work with AI agents in a disciplined way, and how does someone get there from where they are now? This post draws on thinking from developers who've been working through this question longer than most academic knowledge workers have, and translates the hard-won lessons across. Three prerequisites emerge: planning before handoff, documentation treated as infrastructure rather than record, and domain expertise sufficient to evaluate what agents produce. The path in isn't a course or a tutorial — it's building something real with stakes attached.
An AI agent is a system that autonomously executes multi-step tasks using language model reasoning — distinct from an AI assistant, which responds to individual prompts. Agents plan, act, observe results, and adapt, using tools such as file access, code execution, and web search. They perform best when given clear goals, explicit constraints, and well-prepared context.
For the last few months, my screen has been split between Obsidian and a terminal, with two or three AI agents running in parallel tabs. This post describes what that shift in academic workflow looks like and what made it possible. The change is not simply additive: the work has shifted from execution to direction. What that distinction means in practice — and why it matters for those of us working in knowledge-intensive academic roles — is what I try to work out here.
A mode of working in which AI agents handle execution-level tasks while the human operates at the direction layer — deciding what should be made, defining what good looks like, and evaluating outputs.
Seven principles for extended AI collaboration, distilled from a week-long project to restructure a large note collection using Claude Code. The principles cover goal-setting, understanding what AI can and cannot contribute, investing in planning conversations, adaptive planning, safety infrastructure, treating AI output as drafts, and expecting to learn something about your own thinking. Offered not as rules to follow but as patterns to recognise.
A detailed account of a week-long project to restructure 5,819 Obsidian notes using AI as a working partner. The project involved building a 23-category taxonomy, migrating thousands of legacy notes to a consistent metadata structure, and generating AI-written descriptions for every note in the collection. The piece describes not just what was done, but how extended planning conversations, external project documentation, and careful human review at each phase made the work tractable. The most unexpected outcome was that building infrastructure for a note collection required articulating, for the first time, precisely how I think about my academic field.
A field note on switching from Claude Opus 4.6 to Sonnet 4.6 as my default in Claude Code, and what I'm noticing after the first hour.
Prompt injection is a technique in which instructions embedded in text cause an AI system to follow them as commands. In educational contexts it has appeared as an AI detection mechanism in assessment — which raises sharper questions about authorisation and trust than it might initially seem.
The process by which a trained language model generates outputs; the computational work that happens each time you send a prompt and receive a response.
The idea that different kinds of cognitive work have different computational costs in large language models, and that matching task complexity to model capability matters for both efficiency and output quality.
Most academics treat AI models as interchangeable general-purpose tools. They aren't. Different models have different characteristics that make them better suited to particular kinds of cognitive work, and matching tasks to those characteristics may improve both efficiency and output quality. This post explores what that looks like in personal workflows and how the same logic scales to institutional AI strategy.
What happens when you query the Zotero database with AI, treating your entire reference library as context rather than searching it document by document? This field note documents a proof of concept using Claude Code to read a Zotero SQLite database directly. The approach works but what breaks reveals how much your metadata practices actually matter.
Learn about LLM context drift (or context rot) and how the reduction in output quality as tokens increase affects complex AI workflows.
Most universities have responded to AI by rewriting assessment policies and running prompt-writing workshops. Context engineering demands something different: infrastructure decisions that commit institutions to a direction. This post explains what context engineering involves, why it matters for health professions education, and why the gap between changing words and changing structures is where most institutions are stuck.
Most discussions of AI in writing focus on output. This post describes a different experience—using AI as a thinking partner to challenge my choices and claims during a writing session.
A mathematical framework demonstrating that AI tutoring systems with 10–15% error rates can achieve superior learning outcomes through dramatically increased engagement compared to more accurate but largely unused alternatives. Drawing on evidence from health professions education, this essay shows that the multiplicative relationship between accuracy and utilisation creates an accessibility paradox: imperfect but engaging systems outperform perfect but unused ones. The argument carries three critical qualifications—errors vary in consequence and safety-critical content demands high accuracy; generative AI poses distinctive epistemic challenges that may undermine conventional error correction mechanisms; and engagement is necessary but not sufficient for learning, with superficial use patterns capable of nullifying predicted benefits entirely. A framework for calibrating accuracy requirements to context and consequence is proposed.
An agentic AI command-line tool designed to understand, modify, and manage complex repositories of code and documentation.
The accumulated cost of outdated, ambiguous, or poorly structured institutional knowledge — manageable when humans compensate, operationally consequential when AI agents depend on it literally.
A standardised ontology providing business, data, and application architectures for the higher education sector — and a practical starting point for making institutional knowledge machine-readable.
A high-quality typesetting system that separates content from style, designed for the production of technical and scientific documentation.
A lightweight markup language for creating formatted text using plain-text syntax, enabling portability and interoperability.
A powerful command-line tool that converts documents between dozens of different markup formats, enabling interoperability and workflow flexibility.
The foundational layer of digital information, stored as a sequence of readable characters without proprietary formatting.
Most advice on organising your notes for AI treats it as a retrieval problem. The harder problem is translation; making your thinking machine-readable without losing what makes it yours. Contextual interoperability is the infrastructure that enables genuine AI collaboration in scholarly work.
Most conversations about AI focus on what it produces. This post describes what an AI workflow for academics actually looks like in practice — building structured context through documentation, iteration, and judgement that makes AI collaboration increasingly effective over time. Drawing on several weeks of restructuring scholarly output with Claude Code, I describe the iteration cycle, the role of documentation as external memory, and what the process reveals about the relationship between explicit information architecture and productive AI collaboration.
Learned numerical representations of text that capture semantic meaning, enabling similarity-based search and retrieval
A technique that improves LLM responses by retrieving relevant information from external sources and including it in the prompt
Direct question-answer retrieval based on statistical similarity, the default reasoning pattern in RAG systems
A database that stores embeddings for similarity-based retrieval, serving as the knowledge layer for RAG systems
Professional education curricula face a fundamental infrastructure problem: while comprehensively documented, they lack systematic queryability. This presentation introduces a three-layer architecture using graph databases as the source of truth for curriculum structure, supported by vector databases for content retrieval and the Model Context Protocol for stakeholder interfaces.
A curriculum knowledge graph that turns programme documentation into queryable infrastructure. Curriculum structure, learning outcomes, assessment alignment, and regulatory standards are modelled as a graph staff can interrogate in plain language — turning quality-assurance and compliance checks that take weeks into ones that take minutes. A Snapplify product I work on as a partner and consultant.
When AI agents consume documentation as operational input, it undergoes a category shift from reference material to operational architecture; inaccuracies no longer merely inconvenience readers, they cause system failures. This essay argues that the primary bottleneck for institutional AI integration is not AI capability but information architecture: how institutional knowledge is structured, maintained, and made available to AI systems. Documentation written for human readers cannot function as reliable AI input without deliberate restructuring around explicit relationships and rigorous maintenance workflows. Treating this transition as a governance imperative, rather than a technical afterthought, determines whether AI integration delivers on its institutional promise.
The discourse around AI and human cognition tends to focus on differences, but what happens when we invert the question and use LLM terminology to explore the similarities between AI and human thinking? This post examines parallels between AI cognitive architecture and human thinking—context windows, training data bias, tokenisation, temperature, hallucination, and pattern matching—not to claim that humans are language models, but to ask what these similarities reveal about our own cognitive processes and why we are so invested in denying them.
How arms race dynamics in higher education create adversarial relationships between institutions and students, and what drives these cycles
A model for accessing AI capabilities while personal context remains private and under individual control, separating computational intelligence from data ownership.
A lightweight programme that exposes specific data sources or capabilities through the Model Context Protocol standard, acting as an adapter between AI systems and diverse data sources.
An open standard enabling AI systems to access diverse data sources through standardised interfaces with fine-grained permission control.
Persistent context included in every message to an AI model, establishing consistent behaviour, knowledge, or constraints across interactions.
When educators embed hidden instructions in assessment materials to detect AI use, they import adversarial security thinking into educational relationships. This post examines what AI tripwires reveal about institutional assumptions (i.e. that assessment is about artifact authentication rather than learning measurement) and argues that this approach creates escalating countermeasure dynamics while only detecting carelessness, not genuine disengagement. The alternative requires rethinking what assessment is actually for in an era when artifact production has become trivially automatable.
Large language models are deep learning models with billions of parameters, trained on vast text corpora using self-supervised learning, capable of general-purpose language tasks.
A framework for embedding AI literacy development into existing modules and courses, enabling students to develop AI capability while learning disciplinary content.
A technique that combines knowledge graphs with retrieval-augmented generation for structured reasoning
AI reasoning capability that draws conclusions by traversing multiple connected concepts
AI literacy is a multidimensional capability spanning recognition, critical evaluation, functional application, creation, ethical awareness, and contextual judgement, and is not reducible to any single dimension.
A six-dimension framework that underlies all forms of literacy—information, media, digital, data, and AI literacy share the same structural pattern.
Any claim that a course or programme of study develops AI literacy requires important qualifications—literacy develops through sustained practice, is developmental and contextual, and cannot be fully assessed at course completion.
When AI can generate text, images, and ideas at scale, what remains distinctively human? This post argues that evaluative judgement—the capacity to assess what is worth creating, what deserves attention, and what matters—becomes the core human contribution in knowledge work. Drawing on research into evaluative judgement in health professions education, it explores how educators can make this capacity explicit and deliberately develop it, rather than treating it as an invisible by-product of experience.
Generative AI presents serious ethical challenges in education—to academic integrity, to equity, to the nature of learning itself. This post acknowledges these concerns while arguing that AI also represents an unprecedented opportunity for learning at scale, particularly for the kinds of personalised, adaptive learning that have always been theoretically desirable but practically impossible to deliver. For health professions educators committed to expanding access to quality education, this opportunity deserves serious, open-minded consideration.
AI meeting scribes are increasingly being adopted as productivity tools, automatically transcribing and summarising organisational meetings. But who controls these records, and who benefits from perfect organisational memory? This post explores how AI meeting scribes can entrench existing power dynamics by giving those in authority unprecedented access to communication patterns, informal decision-making, and dissent—all rendered visible and retrievable without those present realising the implications for how organisations are governed.
Most commentary on AI in education focuses on what AI cannot do, or catalogues its failures as warnings. This post argues for a different approach—instead of performative critique, demonstrate thoughtful use in your own practice. By modelling considered, reflective engagement with AI tools, health professions educators can critique from experience rather than speculation, help shape how AI is integrated into professional education, and play a better game than the one they're currently losing.
Using natural language to produce desired responses from large language models through iterative refinement
A system-level discipline focused on building dynamic, state-aware information ecosystems for AI agents
A framework positioning personal context—knowledge, values, goals, thinking patterns—as central to human-AI collaboration, with individuals maintaining control over their cognitive environment while accessing AI capabilities.
AI-forward describes institutions treating AI integration as ongoing strategic practice requiring active engagement, rather than fixed deployment of finished solutions.
Higher education institutions face persistent pressure to demonstrate AI engagement, often resulting in 'innovation theatre' — the performance of transformation without corresponding structural change. This essay presents a diagnostic framework distinguishing between performative and structural AI integration across four domains: governance and accountability, resource architecture, learning systems, and boundary setting. Unlike linear maturity models, it reveals gaps between institutional rhetoric and operational reality. Three legitimate strategic positions — incremental, selective, and transformative — help institutions move from accidental drift toward conscious choice. Treating AI integration as ongoing strategic practice rather than fixed deployment ensures institutions preserve agency over technology decisions aligned with institutional values.
Professional curricula are extensively documented but not systematically queryable, creating artificial information scarcity that makes compliance reporting and quality assurance labour-intensive. This essay proposes a three-layer architecture — graph databases as the source of truth for curriculum structure, vector databases for semantic content retrieval, and a Model Context Protocol layer for stakeholder access — that transforms documentation into operational infrastructure. The architecture incorporates temporal versioning for longitudinal evidence, role-based access controls for multi-stakeholder environments, and internal quality audit against institutional policy alongside external regulatory compliance, enabling verification in hours rather than weeks.
An invited presentation at the Lincolnshire LMC Getting It Done Conference exploring how generative AI functions as a virtual business consultant — supporting practice management, strategic planning, and operational efficiency for healthcare practices. The talk covers six use cases: developing online presence, targeting patient groups, enhancing patient experience, strategic service development, ethical implementation, and change management.
Contemporary AI discourse often focuses on 'sanctuary strategies' — defensive attempts to identify uniquely human capabilities — positioning humans and AI as competitors for finite cognitive territory. This essay reframes human-AI relationships as embedded within complex cognitive ecologies where meaning emerges through interaction, and introduces 'taste' as a framework for cultivating contextual judgement: sophisticated discernment about when, how, and why to engage AI in service of meaningful purposes. Unlike technical literacy, taste development involves iterative experimentation and reflection, preserving human agency over value determination. By shifting from 'What can humans do that AI cannot?' to 'How might AI help us do more of what we value?', the essay builds a case for abundance-oriented human-AI partnership.
Higher education's focus on prompt engineering — teaching technical skills for crafting AI queries — represents a misunderstanding of learning. This essay argues that prompts emerge from personal meaning-making frameworks, not technical mechanics, and that the institutional impulse to control AI interaction reveals a 'learning alignment problem': systems optimising for measurable proxies like grades rather than authentic curiosity. Drawing parallels to AI safety's value alignment problem, it shows how AI exposes that many assignments were already completable without genuine intellectual work. Universities must shift from control to cultivation paradigms, recognising that learning is personal and resistant to external specification, ensuring AI becomes a partner in human flourishing rather than a tool for strategic performance.
This essay proposes 'context sovereignty' as a framework for maintaining human agency in AI-supported learning, arguing that context engineering — not just prompting — is the key to meaningful human-AI collaboration.
Health professions education faces a fundamental challenge: graduates are simultaneously overwhelmed with information yet under-prepared for complex practice environments. This essay introduces a theoretically grounded framework for integrating AI into health professions education that shifts focus from assessing outputs to supporting learning processes. Drawing on social constructivism, critical pedagogy, complexity theory, and connectivism, six principles emerge — dialogic knowledge construction, critical consciousness, adaptive expertise, contextual authenticity, metacognitive development, and networked knowledge building — to guide AI integration in ways that prepare professionals for the complexity and uncertainty of contemporary healthcare practice.
Language is humanity's first general-purpose technology, developed to extend cognitive capabilities beyond biological limits. Large language models represent the latest evolution in a continuum stretching from spoken language through writing and print to digital text, extending language's capabilities through unprecedented scale, cross-domain synthesis, and cognitive adaptability. For health professions education, this framing shifts priorities from knowledge acquisition toward cognitive partnership and adaptive expertise. It demands a reconceptualisation of AI literacy — moving beyond technical prompting to understand how these tools shape reasoning — and requires assessment to evaluate students' capacity for collaborative problem-solving. Understanding LLMs as language technology offers a middle path between uncritical enthusiasm and reflexive resistance.
LLM terminology provides unexpectedly precise language for human cognitive constraints we've struggled to describe—revealing that the similarities might be more extensive than professional identity allows us to admit
A template classroom policy for generative AI use that educators can adapt for their own modules and courses.
A presentation for the ADAPT International Conference 2023 on expertise and AI in professional education. Argues that generative AI is an unstoppable force — democratising access to professional knowledge and acting as a low-cost, personalised "expert mentor" — meeting the immovable object of universities and their traditional role as gatekeepers of expertise. Asks what formal professional education is for in a world where expertise is abundantly available, and argues for raising expectations through project-based learning and authentic assessment.
A presentation for the 2023 ENPHE Conference on coming to terms with AI in professional education. Argues that the dominant discourse traps us in a false choice between AI as saviour and AI as threat, when what the moment actually demands is a tolerance for ambiguity. Borrowing the antihero archetype, it reframes AI as neither hero nor villain but an ambiguous partner in new human–machine coalitions, and proposes that wise adoption is not a destination but the start of an iterative, cooperative process.