Michael Webb

@mwebb.bsky.social

848 Followers 824 Following

Mostly AI and technology in education. Director of AI at Jisc. Might post the odd music and cat thing too.

A swan landing in a flooded road
Don’t panic
Don’t panic!
View of the stage and a slide with 4 people.
Man climbing a bench to fix the track
Lighthouse
Tim Peak on stage
Image Title: Detection vs. Transparency Left side: Section titled "Algorithmic Logic" with the subtitle: "Black-box algorithms to estimate text sources." A faded image of a document with some text and graphics. At the bottom, there is a small icon representing a professor's coat with the text: "Only professors have visibility." Right side: Section titled "Transparent Chain of Ownership" with the subtitle: "Browser-based text categorisation." Two columns: Left column (Human typed): "Typed" "Typed then rephrased with Grammarly's gen AI" "Typed then edited with Grammarly" Right column (Sourced content): "AI-generated" "AI-generated then edited" "Copied from an online source" "Copied from an unknown source" The bottom right corner contains the text "Grammarly for Education."
Title: "Let us know if your institution is:" The poll results show the following options and their corresponding percentages: "Academic integrity forward" leading with 40%. "Responsible use forward" at 34%. "Other" at 12%. "Innovation forward" at 8%. "Detection forward" at 6%.
Cat
tomorrow and tomorrow and tomorrow book
Pens
Palm2 on the left hand side, saying it's using Gemini
Waves become wings
Robert Halfon on stage via video
Gig poster Plymouth 23rd Nov.
Jisc AI in Education Bias in AI, particularly within the context of educational technology, is a significant concern. AI bias occurs when an algorithm produces systematically prejudiced results due to erroneous assumptions in the machine learning process. This can happen because of the limitations in the design of the algorithm or the nature of the data it is trained on. Here are some key points about bias in AI, drawn from the "AI in tertiary education: A summary of the current state of play" document: Source of Bias: AI and machine learning models often replicate any biases present in their training data. For instance, if historical data reflects certain biases or exclusions of particular groups, the AI model will likely perpetuate these biases in its predictions or decisions. Generative AI Bias: Generative AI, like ChatGPT, has shown tendencies to produce biased outputs. This bias arises from both the training materials (often scraped from the internet, which can have inherent bia