This article highlights how AI is transforming instructional design. A lot of the ideas brought up in the article like AI providing personalized or adaptive learning, tutoring, gamification, and AI-enhanced assessments can benefit students with learning disabilities or who are neurodiverse. As someone with ADHD receiving immediate feedback and lessons that use a variety of mediums to convey key concepts can help me stay engaged. #accessibility
AI generated gamification sounds horrible, though.
Feels very close to “we finally built the engine of horrors, from that classic novel, ‘Do not build the engine of horrors.’”
As ever, the current AI hype is stupid, and almost everyone is in for a disappointment. But…
This is an area that AI shines.
An important aspect of inclusive design is consistently considering folks whose needs are often forgotten. Modern AI’s regurgitation engine excels at catching things that feel obvious once pointed out.
But to step back onto my soap box, so does the average untrained random person (under expert guidance, given a use case to try to accomplish, etc.)
AI is genuinely fantastic for improving in first drafts before getting help from a second human.
Also, AI can bring in the most trivial impacts of expert guidance without access to the original expert. That’s a pretty big boon in a lot of corners of life where expert guidance isn’t highly available.
Accessibility is such an area. If AI can help non-experts apply even the most basic principles of accessibility, that’s a win for everyone.
This sounds like a really cool use of AI.
Some great examples, also a really important thing for less developed regions is being able to construct lessons and take questions in the person’s first language even if it’s something like tagalog.
We’re going to see things like the gamification they mention really grow in scope as people start exploring ideas and developing structures that can be easily adapted using ai tools like coding, image gen, etc to suit the purpose. I know a lot of the coding focused projects are currently working towards lora style training on your own code base and full project incorporation hopefully we’ll see some good open source games get the treatment so that it’s easy to mod to include lesson info or vocab from a list of things you’re trying to memorize.
Some people here are super anti LLM, no wonder you get downvoted. You use it as a tool? How dare you! They are stupid and uncreative and only copy stuff from the Internet!
Me:
Write python code that calculates K_eq for temperatures from 300 to 1000 °C given: d_R_H =172530 J/mol d_R_S =176.69 J/(molK) d_R_G =d_R_H-d_R_ST K_eq = exp(-d_R_G/(8.314T)) p_total_initial = 1.0 p_CO2_initial = 0.5 p_CO_initital = 0.0 x = 1/8(K_eq**0.5*(8p_CO_initital+16p_CO2_initial+K_eq)**0.5-4p_CO_initital-K_eq) p_CO2_eq = p_CO2_initial - x p_CO_eq = p_CO_initital +2x p_total_eq = p_CO2_eq + p_CO_eq Now iterate p_CO2_initial until p_total_eq = p_total_initial Then calculate mole fraction of both: mole_fraction_CO2 = p_CO2_eq / p_total_eq mole_fraction_CO = p_CO_eq / p_total_eq Then plot mole_fraction_CO (starting from 0 going up to 1) over T. Add a secondary y-axis containing the same y-values in reverse order (starting from 1 going up to 0) since this is exactly 1-mole_fraction_CO, but without drawing another line, since they would be ontop of each other. Then annotate both y-axis: Left hand side is “Mole fraction of CO” and right hand side is “Mole fraction of CO2”. Title is “Boudouard Equilibrium”. Unit of x-axis is “°C”.
And you know what I got first try? Exactly that. I did some more refining but it worked out of the box as specified.