Tasting the Future

Prompt Word of the Week

Slop as in AI Slop:

AI slop is low-quality, misleading, or nonsensical content generated by artificial intelligence. Because it is so easy to generate AI content, AI slop is beginning to flood the internet.

Key Characteristics:

• created using AI tools (text, image, video, audio)

• Rapidly produced and distributed so it can overwhelm legitimate information

• often inaccurate or misleading (but not necessarily purposeful disinformation)

Motivation:

• Gain political advantage (e.g., fake images of effects of natural disasters)

• Increase social media engagement (e.g. Youtube videos with AI generated script, voice, and visuals.)

Bottom Line:

• As AI-generated content improves, distinguishing real from fake will become increasingly challenging

• Digital literacy and better content labeling are crucial to combat this trend

News of the Week

• State of AI Report from Air Street Capital. It's a comprehensive, but accessible report looking at research, industry, politics, and safety of AI.

  • Interesting Take Aways

    • Chinese labs are finding ways around US sanctions and building very capable LLMs

    • Value of AI companies has hit $9T

    • Some AI companies are generating significant revenue already

    • Discussion of existential risks of AI has quieted

    • Canada is fifth place in AI research publication

    • Semiconductor smuggling is a sophisticated and lucrative industry

  • Key Takeaways for educators:

    • Math and science capabilities of models are starting to catch up to language capabilities

    • Powerful open-source models may be a pathway to equitable access to AI tools

    • Rapid rate of development continues making it challenging for educators to stay updated

    • As AI capabilities improve job displacement concerns will grow. This will affect both students as we consider what we are educating them for as well as roles on campus.

    • Both the scope and importance of what AI literacy is growing. How do we stay on top of it, let alone how do we make sure students are on top of it?

    • AI is being used across industries for cutting-edge research and development (e.g., thought-to-speech, atmospheric simulation, folding proteins, mind reading)

• Nobel Nod to AI Pioneers: AI research takes center stage in 2024 Nobel Prizes. Chemistry honors go to Demis Hassabis, David Baker, and John Jumper for AlphaFold2, revolutionizing protein structure prediction. In Physics, Joen Hopfield and Geoffrey Hinton, a Canadian, win for applying physics concepts to enhance machine learning. These breakthroughs underscore AI's growing impact across scientific disciplines. Chemistry Prize | Physics Prize

• Electronic Tongue Breakthrough: Penn State's "electronic tongue" lets AI taste, opening new possibilities in food safety, food production, and medical diagnostics. Taste the future

AI Powered Pedagogy

• AI in Ed (Opinion) Hype vs. Reality: Erik Larson, AI expert, cautions against overreliance on AI in education. He stresses the importance of AI literacy and critical thinking skills for students. Larson warns: "The danger is that you're going to just take [AI-generated text] because it's already done and not do the hard work of doing the writing yourself." Read more

• Northeastern University's AI Gallery: This gallery offers practical examples for integrating AI into various courses. From using ChatGPT in applied math to building patient interview chatbots, these ideas help educators embrace AI responsibly. Explore assignments that define AI boundaries and leverage gen AI for business strategy analysis. Discover more

• Upcoming Webinars (You may find these seminars interesting, but I don't have any insight into how good or informative they will be for you)

  • Georgetown University Faculty and Students Dialogue on AI. (Nov. 11 12pm) (Register for Zoom Session)

  • AI and Disinformation (Oct. 15, 12pm). RSVP now

  • Conference Board of Canada Generative AI in post-secondary education. (Oct. 16, 10am) Register

The mAIn Event

To follow up last week's summary of the AI tools I regularly use, here is an overview of the writers, thinkers, and educators I follow across various media to try to stay on top of developments in AI technology and education.

Substacks, Newsletters, Twitter

AI in Education

Newsletters

• Ethan Mollick (One Useful Thing on Substack): Ethan Mollick is the OG commentator on AI in education. He develops some of the most innovative uses of AI in education and has a great prompt library that he shares on his More Useful Things website. He brings together the latest research on AI efficacy and requires AI use in his courses - including requiring his students to do "one impossible thing" using AI on their final projects.

• Lance Eaton (AI+Education Simplified on Substack): Lance Eaton's strength is bringing practical advice to educators in post-secondary education. He has written on co-creating AI policies with students, how to deliver Gen AI workshops for faculty, and has collected a huge sampling of AI syllabi policies. One of the best things about this substack is that he shares it with an Attribution-ShareAlike creative commons license.

• Nick Polatskiy (Educating AI on Substack): This newsletter explores three key areas: how AI writing tools affect student learning and creativity, the potential of AI-assisted writing for genuine knowledge creation, and the challenge of integrating AI with traditional teaching methods. I find his thoughts about how AI supports and enhances college writing, but could also corrupt it if not used judiciously.

• Philippa Hardman: I first met Phil when she facilitated an AI supported learning designers workshop. It was the first time that I saw how powerful well crafted prompts were for supporting your own work. I took the workshop in the earlier days of generative AI when we were all still scrambling to decide whether to ban it in our classes or not. The focus of the newsletter is on learning design and using research to find the best ways to use AI to support the process of learning design.

• Dan Meyer: Dan is a former public high school math teacher with some very strong opinions about the usefulness of ed tech. He is an advocate for human teachers and is skeptical of generative AI as a tool to support learning and makes good arguments that AI for supporting education is mostly just hype.

Podcasts

• Designing Schools: Interesting weekly discussions about the impact of generative AI on education. The pair of podcasters is Dr. Sabba Quidwai and Stefan Bauschard. They have great insight and perspective, but Stefan needs to get his microphone fixed.

AI Technology

Newsletters

• Gary Marcus: Gary is both an AI expert/academic and AI skeptic. His newsletter offers a skeptical analysis of AI trends and challenges the AI hype. He strongly believes that the current generative AI technology will soon hit the limits of its capabilities but at the same time he advocates for a measured and slower approach to AI to ensure AI safety. He seems to be the goto guy for when media outlets want a perspective opposite the AI hypemasters like Sam Altman.

• The Neuron: A popular daily AI newsletter that breaks down AI topics into bite-sized chunks. It covers the latest AI tools, news and industry trends.

• Allie K. Miller's Twitter feed (@alliekmiller) offers a blend of tech insights and career wisdom. As a LinkedIn Top Voice for Technology, she breaks down complex AI concepts into bite-sized tweets, often spicing them up with eye-catching visuals. Miller's approach is refreshingly human – she shares personal experiences alongside industry updates, making AI feel less like science fiction and more like an exciting, accessible field.

Podcasts

• Hard Fork: A weekly tech podcast from the New York Times. It focuses on all aspects of tech which of course includes a lot of generative AI. The hosts, Kevin Roose and Casey Newton, make the content engaging and understandable through their easygoing, conversational style. Highly recommend - I have not missed a single episode.

• This Day in AI: This is my favourite AI podcast, but they do go pretty deep into technical details. The hosts are Mike and Chris Sharpie who are software entrepreneurs from Australia. They have offbeat and mischevious senses-of-humour but deep insight into how best to use generative AI tools - especially for software development. They also developed the website SimTheory which is a tool like ChatGPT that allows you to switch between LLMs to find the one that will work best for your needs.

• The AI Daily Brief: Delivers daily AI news and analysis in an accessible manner. If you just want to keep up to date with the latest AI headlines, this podcast is the way to go.

• AI Explored: Deep dives into how to use generative AI, especially for supporting social media and content creation.