@dahukanna gpt is handy in helping break down skills into subskills for lesson planning.
Prompted it with my project details and then asked it to unpack a craft/technical skill in it's minutiae subskills and then map a set of competencies in a progression, with weekly or biweekly milestones across 3-4 months
Followed up with a prompt to get more granular and this is what it suggested.
@impactology Q! How do you know the output is reliable? Does it repeat the same structure the next 4 times you run the prompt or does the output change?
Also how do you know if it covered everything or missed something critical? Did you adapt your workflow to check it against something you found perfect. Is there a pattern to what is being overlooked or skipped?
@impactology ps - did the scheduling match the workload - for ex. does the chunk in week 1 be significantly more time-consuming than week 4 even though quantitatively they are the same number of points etc?
ps also does the subskill and competency match properly to hierarchy - or is there a pattern to where it mismatches & puts one in the other?
For eg: utilize twt meaningful engagement feels like outcome while copywriting a skill you can act on, do they both fall under competency as per you?
@impactology I think these questions can help contextualize the tool, is it for someone who knows enough to do it on their own and evaluate someone els'e output (less reliable output is fine in this case as the expert can correct it - with caveat that some quality outputs, the edits take longer than doing it yourself)
or can this be endorsed for non-experts (the expertise then is within the LLM which i am skeptical of)
> ps - did the scheduling match the workload - for ex. does the chunk in week 1 be significantly more time-consuming than week 4 even though quantitatively they are the same number of points etc?
Good question, again don't know. Will have to try. Like I said this is just meant as a springboard for further exploration, not a one-off final solution.
@impactology Gotcha! Another Q, how is this a better springboard for your mental process than say using the course structure from an existing / lateral program that you admire or rely on? (considering in that case, the output passes reliability etc upstream vs gpt where it is shifted downstream)
and a more mythical Q : is the tradeoff enough considering the ethics of open and your stances on ethics.
@impactology Ah, using openAI product considering their ethics, as a way to combat restrictive uni pedagogy which you also have criticised properly before. Weighing out the criticism and how you decided to invite one in spite of it while stay away from other. Curious about the trade off judgement you made. (unless i misread your stance on openAI then no follow up Q)
@missmythreyi >how is this a better springboard for your mental process than say using the course structure from an existing / lateral program that you admire or rely on? (considering in that case, the output passes reliability etc upstream vs gpt where it is shifted downstream)
Their structure is made for university structure. A place where groups gather around to share knowledge usually in a building with has chairs, tables and tools for experimentation
@impactology No, i understood the intent, my curiosity was from knowing if that translates to the LLM understanding this? Does its output include this apsect ? For example which part of prompt underlines this necessity and does this output become a uni model if you run the prompt without that phrase? Just curious,
@missmythreyi the part in the prompt which uses the pronoun "I" "my" instead of "we" when I describe it my project, the specific blend of fields I mentioned : like educational psychology, the aim I mentioned : to invent new ways of teaching and finally the specific milestone projects
All this brings some amount of personalization, although springboard-like that is hard to find in existing programs because they have their own approach, aims and restrictions in interdisciplinarity
@missmythreyi like I haven't found any existing programs that are trying to bring together the fields I am, in the way that I aim and for the purpose I am.
>ps also does the subskill and competency match properly to hierarchy - or is there a pattern to where it mismatches & puts one in the other?
Again good question, I think it didn't unpack the competencies for each sub-skill well enough and so definitely didn't recognize the different level of skill progression for each level/hierarchy, rather that part was quite generic
@dahukanna I know it's nowhere near in terms of the creativity it takes in crafting and segmenting meaningful assignments but its useful to simply get the ball rolling, to just get started and then to iterate. Filling up a blank canvas which you can then personally modify manually
Add comment