A few months ago, my colleagues @wimcarton.bsky.social @nilsom.bsky.social and I published an article on racial capitalism and carbon removal. Elsevier now has a GenAI dropdown embedded in the article itself, incentivizing people to not actually read it www.sciencedirect.com/science/arti... - ThreadSky

hollyjeanbuck.bsky.social • 37 days ago

A few months ago, my colleagues @wimcarton.bsky.social @nilsom.bsky.social and I published an article on racial capitalism and carbon removal.

Elsevier now has a GenAI dropdown embedded in the article itself, incentivizing people to not actually read it

https://www.sciencedirect.com/science/article/pii/S1462901124001990

Comments

hollyjeanbuck.bsky.social•37 days ago

while I do understand that we are now producing work for AI agents to read, I have strong concerns about the publisher being the one to try to capture that space.

hollyjeanbuck.bsky.social•37 days ago

(1) a reader who doesn't want to actually read it could put it into Claude and get a summary they can interact with the agent about. that would be better because it is more interactive.

hollyjeanbuck.bsky.social•37 days ago

(2) if your thought is, "great, this will make the piece more accessible" - we already wrote this to try and be much more accessible than an average social science article. we should keep encouraging people to write in a way that encourages accessibility and theoretical advancement at the same time.

hollyjeanbuck.bsky.social•37 days ago

(3) my main concern is that pieces like these make their analysis based on evidence and examples. the reader needs to look at those to evaluate their strength, and what might be missing, to see if they agree. you can't bypass this step if you want to be a critical thinker.

hollyjeanbuck.bsky.social•37 days ago

(4) if you just read the AI Q-and-A, you would come away with the impression that this is one more article about how CDR is terrible. but for me personally - not to speak for my co-authors here - part of the point of the analysis is to inform an alternative future for CDR. that part is omitted.

hollyjeanbuck.bsky.social•37 days ago

in summary - this is a warning that if you choose to publish with Elsevier (and I hear @davidho.bsky.social cautioning you get what you pay for),

know that your readers will be offered what AI thinks are important questions and answers before they are even confronted with your text.

lfpromethee.bsky.social•37 days ago

My first thought was also: "summarizing is what abstracts already do, right?"

Comments

Posting Rules

Reply