Improving skills and competences of VET teachers and trainers in the age of Artificial Intelligence
I am very impressed with a paper, Towards social generative AI for education: theory, practices and ethics, by Mike Sharples. Here is a quick summary but I recommend to read the entire article.
In his paper, Mike Sharples explores the evolving landscape of generative AI in education by discussing different AI system approaches. He identifies several potential AI types that could transform learning interactions: generative AIs that act as possibility generators, argumentative opponents, design assistants, exploratory tools, and creative writing collaborators.
The research highlights that current AI systems primarily operate through individual prompt-response interactions. However, Sharples suggests the next significant advancement will be social generative AI capable of engaging in broader, more complex social interactions. This vision requires developing AI with sophisticated capabilities such as setting explicit goals, maintaining long-term memory, building persistent user models, reflecting on outputs, learning from mistakes, and explaining reasoning.
To achieve this, Sharples proposes developing hybrid AI systems that combine neural networks with symbolic AI technologies. These systems would need to integrate technical sophistication with ethical considerations, ensuring respectful engagement by giving learners control over their data and learning processes.
Importantly, the paper emphasizes that human teachers remain fundamental in this distributed system of human-AI interaction. They will continue to serve as conversation initiators, knowledge sources, and nurturing role models whose expertise and human touch cannot be replaced by technology.
The research raises critical philosophical questions about the future of learning: How can generative AI become a truly conversational learning tool? What ethical frameworks should guide these interactions? How do we design AI systems that can engage meaningfully while respecting human expertise?
Mike Sharples concludes by saying that designing new social AI systems for education requires more than fine tuning existing language models for educational purposes.
It requires building GenAI to follow fundamental human rights, respect the expertise of teachers and care for the diversity and development of students. This work should be a partnership of experts in neural and symbolic AI working alongside experts in pedagogy and the science of learning, to design models founded on best principles of collaborative and conversational learning, engaging with teachers and education practitioners to test, critique and deploy them. The result could be a new online space for educational dialogue and exploration that merges human empathy and experience with networked machine learning.
In last weeks edition of her newsletter, Philippa Hardman reported on an interesting research project she has undertaken to explore the effectiveness of Large Language Models (LLMs) like ChatGPT, Claude, and Gemini in instructional design. It seems instructional designers are increasingly using LLMs to complete learning design tasks like writing objectives, selecting instructional strategies and creating lesson plans.
The question Hardman set out to explore was: “how well do these generic, all-purpose LLMs handle the nuanced and complex tasks of instructional design? They may be fast, but are AI tools like Claude, ChatGPT, and Gemini actually any good at learning design?” To find this out she set two research question. The first was sound the Theoretical Knowledge of Instructional Design by LLMs and the second to assess their practical application.She then analysed each model’s responses to assess theoretical accuracy, practical feasibility, and alignment between theory and practice.
In her newsletter Hardman gives a detailed account of the outcomes of testing the different models from each of the three LLM providers, But the The headline is that across all generic LLMs, AI is limited in both its theoretical understanding and its practical application of instructional design. The reasons she says is that they lack industry specific knowledge and nuance, they uncritically use outdated concepts and they display a superficial application of theory.
Hardman concludes that “While general-purpose AI models like Claude, ChatGPT, and Gemini offer a degree of assistance for instructional design, their limitations underscore the risks of relying on generic tools in a specialised field like instructional design.”
She goes on to point out that in industries like coding and medicine, similar risks have led to the emergence of fine-tuned AI copilots, such Cursor for coders and Hippocratic AI for medics and sees a need for “similar specialised AI tools tailored to the nuances of instructional design principles, practices and processes.”
In last weeks edition of her newsletter, Philippa Hardman reported on an interesting research project she has undertaken to explore the effectiveness of Large Language Models (LLMs) like ChatGPT, Claude, and Gemini in instructional design. It seems instructional designers are increasingly using LLMs to complete learning design tasks like writing objectives, selecting instructional strategies and creating lesson plans.
The question Hardman set out to explore was: “how well do these generic, all-purpose LLMs handle the nuanced and complex tasks of instructional design? They may be fast, but are AI tools like Claude, ChatGPT, and Gemini actually any good at learning design?” To find this out she set two research question. The first was sound the Theoretical Knowledge of Instructional Design by LLMs and the second to assess their practical application.She then analysed each model’s responses to assess theoretical accuracy, practical feasibility, and alignment between theory and practice.
In her newsletter Hardman gives a detailed account of the outcomes of testing the different models from each of the three LLM providers, But the The headline is that across all generic LLMs, AI is limited in both its theoretical understanding and its practical application of instructional design. The reasons she says is that they lack industry specific knowledge and nuance, they uncritically use outdated concepts and they display a superficial application of theory.
Hardman concludes that “While general-purpose AI models like Claude, ChatGPT, and Gemini offer a degree of assistance for instructional design, their limitations underscore the risks of relying on generic tools in a specialised field like instructional design.”
She goes on to point out that in industries like coding and medicine, similar risks have led to the emergence of fine-tuned AI copilots, such Cursor for coders and Hippocratic AI for medics and sees a need for “similar specialised AI tools tailored to the nuances of instructional design principles, practices and processes.”
There's an interesting post from Philippa Hardman in her newsletter today. Entitled Are ChatGPT, Claude & NotebookLM *Really* Disrupting Education? her research asks how much and how well do popular AI tools really support human learning and, in the process, disrupt education?
She created a simple evaluation rubric to explore five key research questions:
1. Inclusion of Information
2. Exclusion of Information
3. [De]Emphasis of Information
4. Structure & Flow
5. Tone & Style
Philippa Hardman used her own research articles as the input material, which she fed into what she says are considered to be the three big AI tools for learning:
She prompted each tool in turn to read the article carefully and summarise it, ensuring that it covered all key concepts, ideas etc ensuring that I get a thorough understanding of the article and research.
She provides a detailed table of the results of each of the three applications, and additionally of the NotebookLM podcast application, assessing the strengths and weaknesses of each. she says that "while generative AI tools undoubtedly enhance access to information, they also actively “intervene” in the information-sharing process, actively shaping the type and depth of information that we receive, as well as (thanks to changed in format and tone) its meaning. "
She goes on to say:
While popular AI tools are helpful for summarising and simplifying information, when we start to dig into the detail of AI’s outputs we’re reminded that these tools are not objective; they actively “intervene” and shape the information that we consume in ways which could be argued to have a problematic impact on “learning”.
Another thing is also clear: tools like ChatGPT4o, Claude & Notebook are not yet comprehensive “learning tools” or “education apps”. To truly support human learning and deliver effective education, AI tools need to do more than provide access to information—they need to support learners intentionally through carefully selected and sequenced pedagogical stages.
Her closing thoughts are about Redefining the “Learning” Process . She says:
It’s clear that AI tools like ChatGPT, Claude, and NotebookLM are incredibly valuable for making complex ideas more accessible; they excel in summarisation and simplification, which opens up access to knowledge and helps learners take the first step in their learning journey. However, these tools are not learning tools in the full sense of the term—at least not yet.
By labelling tools like ChatGPT 4o, Claude 3.5 & NotebookLM as “learning tools” we perpetuate the common misconception that “learning” is a process of disseminating and absorbing information. In reality, the process of learning is a deeply complex cognitive, social, emotional and psychological one, which exists over time and space and which must be designed and delivered with intention.