#ssp2024 — Public Fediverse posts

Oh, one more message about the #SSP2024 conference. I think I forgot to mention that the closing plenary session was an Oxford-style debate between me and @polka on the proposition "The open access movement has failed".

Modesty forbids that I say who "won" (i.e. shifted most votes between Before and After), but I DO think that between us Jessica and I laid out a pretty convincing record of past OA movements and a compelling vision for the future. To a room full of scholarly publishers.

Just overheard at #SSP2024: "So nice to meet you! You're real!"

A lovely moment.

Yet more talk of impact factors in this "community" session at #SSP2024. Not for me, thanks!

This is on me, for choosing this #SSP2024 session based only on its title, rather than navigating through the website maze to find the session description.

Over all this #SSP2024 session is deeply disappointing. Its title suggests it's about building communities, but it turns out to be all about marketing magazines.

I made the wrong choice in this session: I should have gone to "The diminishing importance of the article: an overview of the emerging modular, multi-modal research landscape".

Now this person at #SSP2024 is telling us that she doesn't bother reading the whole of RFP responses. Well, great. Then why should we bother WRITING the whole response?

Someone who can't be bothered with the process shouldn't initiate it. Don't waste our time. We have work to do.

This #SSP2024 session is now telling the story of an extraordinarily stupid procurement process. The person telling this story thinks it was very smart. I don't know what to tell you.

Disppointing in this #SSP2024 session on "three paths to community" to hear the first speaker place so much emphasis on raising the impact factor of the journal he's involved in. It's as though Sick Of Impact Factors and the #SFDora never happened. http://occamstypewriter.org/scurry/2012/08/13/sick-of-impact-factors/

This #SSP2024 session on "three paths to comunity" sounds like it should be interesting. But the first few minutes are taken up with one of the presenters reading the Code Of Conduct out loud for some reason.

Got to hand it to #SSP2024, the timekeeping in this session has been amazing. Five-minute slots, and after another, all perfecly to time. We're about to end an hour's session having heard a dozen presentations.

These short #SSP2024 presentations each have slide decks, but the bottom 20% of the screen is covered by a (pretty good) real-time transcription of what's being said. It's infuriating how often this is obscuring important information.

Lesson (and this is likely to become more important in future as transcription becomes more common): don't put important stuff in the bottom 20% of your slides.

Next presentation at #SSP2024: arXiv on their newish ability to offer content as HTML as well as PDF (largely for accessibility reasons). This is great, but so far as I can make out, this is only for paper submitted as LaTeX — which is ubiquitous in maths, astronomy, physics and CS, but unheard of in most other fields.

A recurring problem with presentations as #SSP2024: lots of slides have TINY TINY fonts — often because they are screenshots of web-based apps taken from huge screens. Sitting through a presentation/demo now that is literally incomprehensible for this reason.

#SSP2024 short presentation on a tool intended to help handling editors find peer-reviewers. My take is that the problem is too fundamental for any tool to help much: there are just too many new papers needing reviews, and not enough people will time available to do those reviews.

@emmatonkin Haha, yes! Genuinely, I think a lot of my unhappiness about things being said about AI here at #SSP2024 would be ameliorated if people just said "LLMs" when that's what they mean. In too many people's minds, AI means Magic.

A poster from #SSP2024. The phrase "dynamic synergy of AI and XML" might be the worst thing I've ever heard.

(Though, to be fair, the application it describes might actually be useful.)

Choosing from half a dozen afternoon sessions at #SSP2024 and it looks like four of them are about #AI. There really seems to be collective delusion taking hold in this community.

Day 2 of #SSP2024! Come pick up some ]u[ Ubiquity goodies from booth #205 #OA #OpenAccess #scholarlypublishing

If anyone at #SSP2024 is listening to this AI panel discussin and in danger of swallowing the kool-aid, I refer you to this case-study in which an LLM answered my question in a way that contained eight factual statements and EVERY SINGLE ONE OF THEM was incorrect — but stated with impacable certainty.
https://svpow.com/2024/05/23/another-day-another-catastrophic-ai-failure/

@paulwalk Yep. Three quarters of an hour into the session, and I've yet to register any one of the three panelists even mention to gross unreliability of LLMs. (It's possible I tuned out for a moment and missed a passing reference.)
#SSP2024

Now on to "Research integrity and AI". #SSP2024. Hmm.

I'm hearing utterly deluded ideas from the platform, such as "ChatGPT will write our papers for is, and do our reviews for us". It's hard for me to believe that I'm hearing from an expert.
#SSP2024

Panel discussion on #AI entited "Rise of the Machines" at #SSP2024. So far I'm hearing a lot of cheerleading from the three panelists. I'm hearing the analogy that LLMs will help us with scholarship just as calculators help with arithmetic. This analogy would work better if calculators gave wrong answers half the time.

We are at #SSP2024 this week! Come to the @crossref booth in the exhibit hall (number 312) to learn more about ROR or find and say hi to ROR Technical Community Manager @amandafrench. #PIDs #metadata

@chartgerink … and #SSP2024, though that's not getting much traction.

@thatandromeda Good to meet you! That makes three of us using #SSP2024 now (including @crossref) so maybe it's not too late for it to take off.

This was OK. turns out that "vested" here primarily means that the service provider charges the publisher a share is revenue rather than charging per page or per paper, so it had an incentive to make the publisher successful. #SSP2024

Next up at #SSP2024, Newgen on "reimagining the vendor-publisher service relationship: the case for vested publishing services".