I'm going full luddite mode on GPT, copilot, all that shit. Get it out of here. I'm not suffering a computer to feign knowledge.
Instead of misinformation, etc we should be talking about information pollution. And people are making machines whose entire job is to spew information pollution.
@cliff i haven’t figured out the hook yet but i have half a talk sketched out contrasting it with the industrial revolution and https://en.wikipedia.org/wiki/Great_Smog_of_London
Great Smog of London - Wikipedia

@phillmv i like the comparison, but at least the one thing you could say for coal burning is that it could do some productive work. the smog is the primary output of generative AI which is the fucked up part

@cliff i’m not convinced it’s _useless_; there’s a bit of heat there.

but it’s totally smog everything & it’s being pushed irresponsibly.

so with the industrial revolution came great migrations, to the new world & to the cities, and folks brought with them their lore: fairies and elves and spirits and gnomes, that explained the world around them.

imagine telling a mfer from 1809 that you caught an elf in a jar, and that it’s wicked smart, lies constantly, and yr gonna start doing what he says

@phillmv @cliff
Like to hear your talk or essay.

@phillmv @cliff

I think the purpose of ChatGPT and LLM is to generate plausible bullshit.

If you can't tell the difference, what's the difference?

@cliff oof yeah, that and it will be weaponised pollution as well.
@cliff This is also how Keith Stanovich diagnoses the problem of increasing myside bias in modern life; an explosion of information. The more information, the easier to (often unconsciously) pick and choose only what reinforces our pre-existing views and reject or downplay the rest.

@cliff

Info smut and smog
Together block our vision
Rule by misreason

#haiku

@cliff I think with the rise of AI it's going to get even worse. I feel like the next cat and mouse game is going to be AI on one side trying to filter the info pollution while AI on the other creating it.
@cliff @cordelya This is very much the argument I’ve been trying to make. We’re polluting the internet with layers and layers of garbage that makes it difficult to find what is real. And on top of that, these systems will start ingesting the garbage and the data quality will continue to drop.
@nazgul @cliff @cordelya How long before we can't trust articles on the internet to not be AI generated BS? I give it a few months.
I'll also be checking the dates of articles for pre 2023 timestamps going forward. (Pre-AI contamination)

@samhainnight @cliff @cordelya It doesn’t help that all the commercial publishers are weekly changing summaries and dates for SEO.

I don’t think we were more than a few weeks into 2023 than my search for “best xxx” returned pages of “best xxx of 2023”.

It doesn’t bother me so much that everyone’s trying to game the system…of course they are. What bothers me is that the search engines seem to have stopped caring.

But it all makes sense when you apply Doctorow’s “enshittification” model. They’ve completely stopped focusing on users. All they care about is advertisers. So who cares if the search results are garbage, so long as they get to put ads on them?

@cliff @emc2 The research and literature speak of mis/dis/ and mal (MDM) information. The study is more nuanced so the pollution concept would have to be adapted.

From https://proprivacy.com/privacy-news/what-is-mal-information

Misinformation – false information thought to be true by those disseminating it, often linked to sincerely held yet erroneous beliefs.

Disinformation – false information being purposefully disseminated by those who hope it will resonate with a particular type of audience and achieve some behavioral outcome.

Mal-information – Information circulated in order to create some negative outcome such as sowing division, which can either be true or false information.

A response guide from CISA:
https://www.cisa.gov/sites/default/files/publications/mdm-incident-response-guide_508.pdf

What is Mis/dis And Mal-information And What Can We Do?

Several members of our team here at ProPrivacy recently attended a webinar about mis/dis and mal-information. Here's what you need to know about it.

ProPrivacy.com

@cliff Good metaphor. Related to "Data Smog" (Shenk, 1997)

https://en.wikipedia.org/wiki/Data_Smog

I think the difference is that most real pollution is a bad side effect of producing or doing something of actual value. Today, "AI" producers are creating pollution and telling us that the pollution itself is great stuff.

Data Smog - Wikipedia

@cliff
Whitney Phillips and Ryan M. Milner have done some work along these lines in You Are Here
A Field Guide for Navigating Polarized Speech, Conspiracy Theories, and Our Polluted Media Landscape and elsewhere - they talk about it in a Project Information Literacy literacy interview here:
Thinking Ecologically about Our Polluted Information Networks

https://projectinfolit.org/smart-talk-interviews/polluted-information-networks/

Thinking Ecologically about Our Polluted Information Networks – Project Information Literacy

@cliff Agreed. Whatever happens with this trash, I'm long on actually knowing shit.
@cliff same, but I assume that’s not news to you

@cliff I'm with you.
So far, GPT has shown itself to be at least as gullible as the sort of people who make sure that misinformation spreads faster amd further than verifiable truth.

And it seems to be even better at the "loud, confident and wrong" act than humans.