What are good examples of concrete acts of resistance against "AI"?

[alineblankertz@indieweb.social]

@resistAI Will do! This is related to a book and other things I am writing/talking about, it is in my direct interest to share everything that can be shared

[aepasek@scholar.social]

@alineblankertz @asrg @danmcquillan

Getting limits on the use of AI put into one's union contract!

Limiting the collection of data that can be used in training/the legal access of such data for training/the legal commercialization or copyright of model outputs.

Tarrif structures that ensure that data centers have to pay their full share of grid infrastructure costs.

Synthetic text limitation policies at work/in journals.

Scorn and shame

[resistai@mastodon.social]

@alineblankertz Thanks and all the best with your book.

[aepasek@scholar.social]

@alineblankertz @asrg @danmcquillan Most of these are from my little commentary on a discard studies approach to AI. Basically, it's helpful to also attack the enabling conditions rather than (just) regulate consumer use.

Paywalled (️): https://intellectdiscover.com/content/journals/10.1386/jem_00138_1

Preprint: https://drive.google.com/file/d/1gg9PclZewhipEOknbCHHN1VI9Hpt27SI/view?usp=sharing

[alineblankertz@indieweb.social]

@Aepasek
Thank you!

[alineblankertz@indieweb.social]

@lennybacon
Thank you!

[hadokowa@mas.to]

@alineblankertz

@Aepasek

I install this ublock list on every machine, it (among many things) removes ai results from search engines, which are often the entry door for people..

GitHub - laylavish/uBlockOrigin-HUGE-AI-Blocklist: A huge blocklist of manually curated sites that contain AI generated imagery for uBlock Origin & uBlacklist.

A huge blocklist of manually curated sites that contain AI generated imagery for uBlock Origin & uBlacklist. - laylavish/uBlockOrigin-HUGE-AI-Blocklist

GitHub (github.com)

[megaphon@mastodon.social]

@alineblankertz @gerrymcgovern @asrg @danmcquillan

Refusal.

For example, people have sent me documents with the comment "hey please look over this text for this work thing we're doing together. The first attachment is my draft, the second is what chatgpt did with it".
I flat out email back "thanks for your draft, here are my comments. I refuse to read the other one."

(Probably easier for me than others; I'm a freelance traditional artist – people expect me to be weird anyway.)

[ortie@corneill.es]

@alineblankertz small but fun, tumblr users had a sort of spam campaign to get ai overview to suggest to Tesla owner to wash their cars with lemon and britte

[luc0x61@mastodon.gamedev.place]

@alineblankertz @gerrymcgovern @asrg @danmcquillan I'm already beyond that, with my thoughts.
My questions are: will we recognize in the future such a scam? Will we be able to sustain new technologies, or anything new is to be considered "rotten as the AI promise"?
It has serious consequences on the future, no matter how you take it - e.g. environmentally.

[megaphon@mastodon.social]

@alineblankertz @gerrymcgovern @asrg @danmcquillan

I'm lowkey politically active. When I spot some government entity or office using AI illustration for brochures, handouts and the like, I call them out.

"Is that AI? Why did you use that? Are you aware of the implications regarding the environment? Isn't that against the policy of your department?"

I do this in a friendly way. Mostly, they admit they were in a hurry and didn't really think about it, since everyone is using AI, too.

[thomasfuchs@hachyderm.io]

@alineblankertz @gerrymcgovern @asrg @danmcquillan@kolektiva.social I know that some people make honeytraps for AI scrapers that specifically inject bullshit into training data; both for code and for prose.

[stefan@stefanbohacek.online]

@alineblankertz Not sure if this fits what you're looking for, but sharing in case it might be relevant.

People-powered technology | Stefan Bohacek

People over (lying) machines.

(stefanbohacek.com)

[lennybacon@infosec.exchange]

@alineblankertz
@jaredwhite is also here

[malte@radikal.social]

@alineblankertz Ridicule, humour, memes against slop. Turning the culture against the delusional ideas being spread by the industry.

[malte@radikal.social]

@alineblankertz Baiting data crawlers to poison the large language models with noise and nonsense. I've heard about some tricks to make the crawlers load the same page over and over again - not sure how.

[alice@lgbtqia.space]

@alineblankertz data poisoning (such as prompt attacks, Glaze & Nightshade), and resource thrashing (making it too expensive to run their LLMs).

@gerrymcgovern @asrg @danmcquillan

[thoe@snac.9space.no]

@alineblankertz@indieweb.social

I really liked @mntmn@mastodon.social 's suggestion to bring back the e-mail signatures highlighting the environmental impact of printing e-mails, only for "AI" instead. It's not much, but it's something. If I had the courage I'd use something like this at work:

Before using "AI" to summarize, respond to, or process this email or its attachments, please consider the environmental impact and the potential effects on your own cognitive abilities. "AI" consumes significant natural resources, and relying on it can diminish your own thinking skills.

CC: @gerrymcgovern@mastodon.green @asrg@tldr.nettime.org @danmcquillan@kolektiva.social

Edit: spelling

[mariamannone@fediscience.org]

@megaphon @alineblankertz @gerrymcgovern @asrg @danmcquillan Maybe they do not want to hire and pay artists to draw. Same happens with music. Very few people are interested in new artworks. Even less are willing to pay for them.

[penguinrebellion@tldr.nettime.org]

@alineblankertz

In addition to what others have said, you can resist as a human being:

• Reply with an AI;DR to emails or messages containing #AIslop.

• Change your email signature to something along the lines of:

Using so-called "Artificial Intelligence" (AI) is a waste of energy and exacerbates climate change.
If you are short on time, there is no need to generate an reply to this. Raw bullet points are very acceptable. Stay human!

@gerrymcgovern @asrg @danmcquillan