Towards a code of practice for AI in AT.pptx

Towards a code of practice for AI in AT.pptx

• 1.
  Workshop: Towards acode of practice to support the use of AI (artificial intelligence) in AT March 2024 Rohan Slaughter
• 2.
  Rohan Slaughter Snr. Lecturerat the University of Dundee from Nov 2020 - • Developing and teaching MSc Educational Assistive Technology (EduAT) Sept 2015- October 2020 was employed by Jisc as a subject specialist • Provided consultancy support on assistive technology, accessibility, infrastructure, technology strategy, leadership and management, use of technology in the curriculum Previously worked at Beaumont College 1999-2015 • Assistant principal, (curriculum, quality, technology, facilities) • Head of technology • Mainstream information technology (IT) • Assistive Technology (AT) • E-learning / Technology Enhanced Learning / EdTech • MIS (Management Information Systems)
• 3.
  Session Aims (1) •We want to hear from members of the ATN community to gather your views on AI in AT systems • We want to consider the potential benefits of the use of AI in AT systems, as well as discussing what safeguards may be needed • We wish to understand specific concerns you may have and where you’re optimistic about the potential of AI in AT systems
• 4.
  Session Aims (2) •This is an interactive workshop; we mainly want to hear from you! • This afternoon in the ideas room - we would be particularly keen to hear the views of ATN event delegates on whether a code of practice or code of ethics might be needed for those looking to make use of AI in AT systems
• 5.
  Why Are WeDoing This? • The AAC/AT Research Group at Dundee has been working on the development of effective AAC systems for 30+ years, which has involved the use of "intelligent" computing. • In September 2023 The UoD MSc EduAT team (Rohan, Tom Griffiths and Annalu Waller) undertook a workshop on the use of AI in AAC systems at Communication Matters conference at Leeds University, the resulting paper has been submitted to JET (Journal for Enabling Technologies), currently in review. • We want to understand what the community thinks about the use of AI in AT systems to guide future research priorities.
• 6.
  Background - WhyNow? • Recent rapid growth in large language models (LLM) such as ChatGPT, Bard Gemini, Bing Co-Pilot • Generative AI is now in the public consciousness • Critical thinking around AI use is not yet well developed: • What does this mean for HE/FE? • Academic Misconduct (cheating) • Authentic assessment? • Institutional AI policies and guidance materials development.
• 7.
  Background – interestingthings to read 1? Very recent reading around the development of legal precedent and regulation or legislation of AI: • Copyright and potential mind-wiping of LLM’s that have ingested copyrighted works without licencing: • Why The New York Times might win its copyright lawsuit against OpenAI | Ars Technica • A new Copyright Settlement in the age of AI? • Desai, Deven R. and Riedl, Mark, Between Copyright and Computer Science: The Law and Ethics of Generative AI (February 16, 2024). Available at SSRN: https://ssrn.com/abstract=4735776 or http://dx.doi.org/10.2139/ssrn.4735776
• 8.
  Background – interestingthings to read 2? Very recent reading around the development of legal precedent and regulation or legislation of AI: • It’s all fine Microsoft is just making VCR’s: ¯_(ツ)_/¯ • Microsoft argues Supreme Court’s VCR ruling should doom NYT’s OpenAI lawsuit | Ars Technica • Meanwhile in the sensible part of the world: • EU votes to ban riskiest forms of AI and impose restrictions on others | Ars Technica
• 9.
  Background - Scope •This session focuses on the use of AI by, with or for AT users with a view to: • Increasing text input speed (super-charged auto-complete / predictive text) • Richness of output (augmentation based on a short input or prompt) • Producing summaries of academic texts etc.
• 10.
  Questions for Today 1.What are the opportunities for AI to support AT users? 2. How might you see these opportunities being realised? 3. What do you see as the risks for the use of AT to support AT users? 4. How would you propose mitigation of the risks? And in the ideas room (later this afternoon): 5. Do you think a code of practice is helpful, and if so, what should it contain?
• 11.
  Question 1 What arethe opportunities for AI to support AT users?
• 12.
  Question 2 How mightyou see these opportunities being realised?
• 13.
  Question 3 What doyou see as the risks for the use of AI to support AT users?
• 14.
  Question 4 How wouldyou propose mitigation of the risks?
• 15.
  Question 5 (IdeasRoom) Do you think a code of practice is helpful and, if so, what should it contain?
• 16.
  Thank You RSlaughter001@dundee.ac.uk @rohanslaughter MScEduAT@dundee.ac.uk
• 17.
  MSc EduAT links •At the Natspec TechAbility 2021 Conference (resources here: TechAbility 2021 Resources - TechAbility), I provided a 5 minute video overview of MSc EduAT: The EduAT Masters course – YouTube • Main course website: Educational Assistive Technology (part time) MSc | University of Dundee • The MSc EduAT scholarship has been renewed to support the 2023 intake. • There is a recording available of our 2021 MSc EduAT open afternoon, the most interesting / useful part is our students giving their views at around 1:07 in this recording: MSc Educational Assistive Technology Open Day - YouTube There is a recording of me going through the core modules towards the start of the video. • There is a set of slides for the EduAT modules on the TechAbility 2020 conference website. There is also a shorter presentation that I gave on the course, see MSc EduAT recording on the Call Scotland website.