What are the ethical and pedagogical implications of AI deployment?
That was one question Graham Attwell and Sophia Roppertz were discussing at the 30th Annual EDEN Conference hosted by the National Distance Education University UNED in Madrid 22-25 June 2021. Die Conference has held under the title „Lessons from a pandemic for the future of education”.
At the beginning of the workshop, Graham Attwell and Sophia Roppertz gave input on
- Why it is important to care about AI in (Vocational) education and training and
- How ethical issues around AI affect VET.
The development of AI-based systems is increasing in the education landscape. Many technologies already exist, others are being developed. We should think early on about how AI systems in the school context can be aligned with our value system. What technologies enrich educational practice? Just because something is technically feasible does not mean that this innovation is also an advance for pedagogical practice.
On the one hand, the development of AI systems and the self-learning algorithms behind them must be discussed. What data is used to train the system? How does one deal with bias, e.g. discrimination against groups of people? Who is involved in the development?
Secondly, the use of such systems in pedagogical settings must be discussed. What skills do people who use such systems need and which AI applications are helpful at all?
Four AI scenarios, which were presented in April 2021 in a UNESCO report, were discussed with the workshop participants:
Scenario 1: Educational chatbot
A chatbot facilitating student admissions (e.g. “What computing courses do you have?”) or providing 24/7 information (e.g. “When is my assignment due?”).
Scenario 2: Automated writing evaluation
Using AI techniques to provide automatic feedback on writing.
Scenario 3: Intelligent tutoring systems
As the student engages with the learning activities, the system uses knowledge tracing and machine learning to automatically adjust the level of difficulty and provide hints or guidance according to the individual student’s strengths and weaknesses.
Scenario 4: AI-driven discussion forum monitoring
A tool might triage the forum posts and automatically respond to the simpler ones; aggregate posts that raise overlapping issues.
Even if the discussion about these scenarios is very important, we suggest another perspective …
… first, we need to answer the questions about what our vision for education is and what are the major challenges. Then we can consider if an AI system could help to solve this problem or helps to achieve the vision.
Using AI for social good!
Do you want to get an impression of what we discussed?
Have a look at this concept board:
If you see further ethical or pedagogical implications, feel free to add them to the concept board (use the “sticky note” function).