Ancestors

Toot

Written by ungual on 2025-01-31 at 22:09

i'm just going to drop that here but since i'm getting haunted by conversation about AI on all side professionnals and intimates, i'd like to adress something too few people seems to care about.

we always speak about AI in pros/cons, the addict will say how it's fast, surprising, opening possibilities, and the wise will say how it's stolen data, computationnally intensive, corporations owned, etc.

but to me it's also about everything you DON'T do while you use AI. it's just like people loose their spatial feeling of a city because they always use google maps. it's creating generative numbness and dessensitivations.

even if an AI was the most pure "ethical" thing, open-source, people-owned, self-trained, well i don't want it to write my emails. even though it's doing it "well". it matters to me to write my email. yes slower, but to devellop an intricate knowledge on how to handle written conversations, to think about the words i use, to reflect on what i wanna say to someone and change my mind in the middle of it, to open one i wrote 2 years ago and realise how much i changed, to have a vivid memory of a random expression i used and use it back somewhere else. this is practice. i wanna practice stuff, not just getting them done.

=> More informations about this toot | More toots from doriane@post.lurk.org

Descendants

Written by ungual on 2025-01-31 at 22:48

i have conversations about students who rely on AI to make projects. they get things done sure, but they don't practice. they don't explore the materials or their own sensibilities. and it shows.

when they have to present a project in front of the classroom they often struggle to say anything about it, they're not confident. confidence is rooted in practice.

i don't know if it is a cliché but are younger generations going to be less confident? is AI going to give them big imposter syndromes for the rest of their lifes because they just learned how to output stuffs.

maybe they feel AI is doing a great job when the goal is to "get things done". but at some point in your life it is ok to have higher expectations: to know yourself - where you fail and where you're good, to develop an instinct for meaningfullness, to build through complexe balances of sensibilities and collaborations. you need to practice for that: to learn, to search, to be patient, to look into yourself, to look into the vastness of the world and not into a single grey interface.

i don't want students to "restraint" from using AI because it's morally "wrong". I want students to practice because it's like acquiring super powers. to learn things so cool that those who just use AI envy them for their intricate craft and conceptual eloquence.

=> More informations about this toot | More toots from doriane@post.lurk.org

Written by ungual on 2025-01-31 at 23:09

finally, even if you're using LLMs chatbot to ask them to explain something and not do it, i think it's sad. search, searching is good practice, develop your own mental tools for understanding. even if it is "so slow" you'll get so rich from it.

in your quest, you'll end up on wikipedia rabbit hole and maybe one day starts editing a page? you'll find bad tutorial and ask yourself why they're bad and how you would explain it better. you'll develop a critical point of view in how knowledge is given to you, you'll start asking by who and for what reason? you'll take a sheet of paper and make some drawing to explain it to yourself and it will become a tool you'll often use until you need less and less paper to figure out things. you'll end up discovering niche blogs with some cool comments and start following the person who wrote that on social media and a few years later meet them and work with them. maybe you'll buy a book on the topic and that book contains wonderfull illustrations that will inspire you more than what you where trying to learn in the first place. you'll annotate the book and give it to a friend. they will give it back to you 10 years later.

So whenever you use AI, also think about what else you're NOT doing, in the meantime i'll romanticize the fuck out of practicing.

=> More informations about this toot | More toots from doriane@post.lurk.org

Written by YHANCI~1.TXT on 2025-02-01 at 00:39

@doriane the conclusion we have reached so far (in discussions at St-Luc that are admittedly difficult because our department is composed of people from very different backgrounds - it's precious but sometimes complex. Anyway) .. .the only conclusion we could agree on so far is that we need to more clearly and explicitly give value to the process, as a way to resist tools that are solely focused on outputting a result/product. It's maybe more naturally already present at erg ; it definitely can turn out to be more of a challenge at St-Luc, but I'll keep pushing the idea :p

=> More informations about this toot | More toots from yhancik@octodon.social