OLDaily

Balancing optimism with caution in higher ed AI
Kathe Pelletier, EdScoop, 2022/06/08

This is a short article but it makes several very good points administrators in higher education should keep in mind before adopting artificial intelligence:

• "match the tool to the purpose" - and indeed, have a purpose
• "invest in data capabilities first" - this includes data management, ethics and policies, and overall data literacy
• guide and leverage AI with human interpretation
• be informed and critical about AI, knowing what it can and cannot do
• "lean into improvement opportunities" in pedagogy and programming

I would also add the point that students need to be aware of how AI is shaping their own disciplines, removing much of the rote work and replacing a lot of creative, classification and recognition tasks with faster and cheaper tools.

Web: [Direct Link] [This Post]

A framework for enterprise AI adoption
Ben Dickson, TechTalks, 2022/06/08

Ben Dickson summarizes Demystifying AI for the Enterprise, a book that "shares best practices and practical case studies of how AI is being applied to different industries." Though not specifically addressing the educational context, the applicability is nonetheless clear. It offers what is called the IMPACT technology adoption model (standing for imagination, maturity, people, augmentation, amplification, automation, culture and transformation (if you squeeze your eyes you can see the words SAMR written there)). "A good AI strategy starts with knowing what problem you're trying to solve," writes Dickson. "Having a good understanding of the capabilities and limits of current AI technology will help you look at problems from a new perspective." I would also advise not dismissing AI or selling it short. It's real, it will have an impact, and it's happening now.

Web: [Direct Link] [This Post]

As Challenges in Education Persist, Our Coverage Will Elevate Educator Voices
EdSurge, 2022/06/08

I'm not sure how much I believe this, given what EdSurge has published in the past. It here touts its Voices of Change project, "creating opportunities for educators to reflect, share and learn from one another through journalism, storytelling and research." The problem is that this content is mixed almost indistinguishably with its sponsored content, which you can read about on its ethics page. "EdSurge allows Sponsored Content in two forms: (1) Content produced by EdSurge Studio on behalf of its advertising partners; and (2) Content produced by advertisers that meet our editorial standards." The publication says it posts disclaimers, but the reason advertisers don't just advertise is that they know readers will be deceived by it. But these are EdSurge's choices to make, and it's up to you to assess the publication according to your own standards.

Web: [Direct Link] [This Post]

End Student Debt. Own your University.
UniversityDAO, 2022/06/08

This is an interesting idea that might work or (more likely) is completely unworkable. Here's the Product Hunt pitch: "UniversityDAO is a decentralized autonomous university aiming to end student debt and empower Universities to own their university through the tokenization of the University system." Here's a white paper. The idea is that students buy tokens (using real money) and this is used to fund the university; the tokens, meanwhile, represent shares in the ownership of the university (or whatever is left after the students' education is funded). Related: Social Bees University DAO, University DAO Voting, Invisible College.

Web: [Direct Link] [This Post]

This is the worst AI ever
Yannic Kilcher, YouTube, 2022/06/08

I rarely post single-link YouTube (SLYT) posts, but you will want to watch this 20 minute video. In a very well produced presentation Yannic Kilcher describes how he trained a GPT bot on some of the worst and vile discussion threads on 4chan and produced a convincing result, including achieving a greater score on 'truthfulness' than other GPT language models. Here's a full set of links to the model, code and dataset. Even if you're not interested in this specific bot, the video provides good descriptions of how language models work, how he defeated 4chan's captcha, and a bit about how language models are evaluated for things like truthfulness.

Web: [Direct Link] [This Post]