3 responses to “A3: Chats with Clarisse – What’s up with ChatGPT4?”

  1. Joel Flanagan

    Hello Clarisse,

    Thank you for submitting your A3 project. It has sparked much reflection, especially on the impact of technological advances on various industries and environments. The substantial scale of computational power and electricity needed to maintain the servers and AI capabilities you discussed raises crucial questions about sustainability and resource management. I can’t help but feel that there will be another digital divide in the future with access and feasibility of these generative models put into question.

    I appreciate your acknowledgment of the teacher’s role in classroom learning in your podcast, which remains crucial despite the push toward increased technological integration. Your project presented a balanced view of the opportunities and challenges, though I think a more critical focus on an aspect could have added depth to the forecast. Nonetheless, thoughtful ideas contribute to the ongoing conversation about technology in education.

    Thank you for sharing your perspective.
    Joel


    ( 0 upvotes and 0 downvotes )
  2. Kirsten

    Hi Clarisse,

    Thank you for including respective and respected AI and ecology experts into our 523 community! I am not yet using ChatGPT4 (primarily because I refuse to pay) but I have begun to delve deeper into the secondary functions of the free existing online version. I have discovered ‘other GPTs’ within ChatGPT3 where users can “discover and create custom versions of ChatGPT that combine instructions, extra knowledge, and any combination of skills”(Explore GPTs chatgpt.com). My greatest fear from an educational perspective is that this technology bypasses the first 5/6 scaffolds of Bloom’s taxonomy expecting human users to be level six content creators from first contact: namely remember, understand, apply, analyze, evaluate. Digital Literacy pushes this traditional framework to elicit a new shape with new verbs that can support today’s learners.

    Your interview with Brandon was interesting and I would have loved to have a glimpse inside of his Meta-vision computer! You mentioned a series of videos that we should all watch but I could not locate them on your website. Could you please direct me to their location for further information? I did a bit of digging to decipher the actual connection between Meta (or Facebook, as I will forever call it) and GPT 4. I was drawn to this article (Can Meta AI Beat GPT4? I Tested Them to Find Out. – Tremendous) where Chat-GPT4 clearly beat Meta AI in task execution. Brandon mentioned the improved multimodality of GPT4 to accepting video input and the capacity to decipher intonation in context. Could this be the main reason why or are there other factors? Either way, incroyable!!

    As an expert in the ecological impacts of AI, I would have liked to hear more of what Robyn had to say, although I realize your 9-minute podcast already exceeded time expectations for this assignment. She reminded us that large language models (LLMs) are “worse than the aviation industry” and that most people do not even think of the environmental consequences because it is built with reusable computer hardware (vs disposable paper, plastic, etc). One fascinating statistic that I came across in the multiple resources I engaged with for 523 was that GPT uses 25% more energy than a Google search (ChatGPT consumes 25 times more energy than Google (brusselstimes.com). If that was common knowledge, would people be more mindful of what they post? Without even knowing it, our society is starving for standardization to prevent frivolous and malicious sharing, if for nothing else but for environmental reasons.

    I loved your final question: Do we think ChatGPT4 is a potential solution to an existing problem? Or are we on our way contributing to new problems? I recently read that businesses have fronted a significant amount of capital in the AI industry, estimated at 2 trillion USD but it has not yet paid off (When will Big Tech investments on AI pay off? – Marketplace). The disruption has not yet hit in a way that is solving the Wicked Problems of the world. Will it ever if we are literally removing the educational foundation upon which solutions are created? The main reason schools still endorse creative and critical thinkers is because we still have massive existing major world problems (as fires and floods tell the story of our 2024 BC summer ☹). Until a technology is created to disrupt the environmental problems into a solution, none of this technological culture is sustainable and we should all shut our devices down immediately to go work in our gardens.

    PS. Can you ask Brandon if there is any humanly possible way to retrieve an unretrievable password on Meta for my 15-year old Facebook account, pretty please with a cherry on top? ????


    ( 0 upvotes and 0 downvotes )
  3. Shannon Wong

    Hi Clarisse,

    Great idea to incorporate a couple of special guests in the creation of your podcast. I enjoyed hearing their different perspectives!

    I appreciate the points Robyn raises as I made a similar comment, raising issues around sustainability and environmental impact, when we explore the Internet of Things as part of the Movable Feast. That’s wild that the use of ChatGPT is worse for the environment than the aviation industry – I wonder how many people would stop (or limit) their use knowing this. Currently, based on my own use and knowing how my peers use the tool, it is used quite freely, without much or any thought put into the impact of every prompt.

    Do you have any insight on how the environmental impacts of ChatGPT compare to say, a Google search?


    ( 3 upvotes and 0 downvotes )

Leave a Reply

You must be logged in to post a comment.