@JubalBarca@scholar.social

James Baillie @JubalBarca@scholar.social

· 3mo ago

Replying to

That's such an unimaginably horrible realisation that we shy away from recognising it, but it's true and we have to confront it if we want to keep children safe. Locking them out of online contact is likely to isolate that block of vulnerable children to a worse degree. Furthermore, when children go on social media anyway, which they will do, they're now doing something illegal, which makes it much *harder* for them to ask an adult for help when something goes wrong online.

22

2

11

0

James Baillie @JubalBarca@scholar.social

· 9mo ago

duck

in case your timeline needed duck

#wildlife #nature #wildlifephotography #cute #birds

scholar.social

91

7

38

1

James Baillie @JubalBarca@scholar.social

· 2mo ago

EU resident friends, we can vote on the design of new Euro notes apparently - here's the survey link on case you want more cute birbs on your currency in future: https://surveys.ecb.europa.eu/10b/banknotes/share/ Do feel free to boost/share :)

surveys.ecb.europa.eu

6

9

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Sitting here and realising I might genuinely have spent twice as many hours on trying to apply for money for myself or other people to do further research as I have on actually researching anything seriously, over the past 12 months.

Why do we run systems like this? Is this really a good use of every academic's time?

37

6

19

1

James Baillie @JubalBarca@scholar.social

· 5mo ago

I'm looking to maybe commission someone to draw/paint/art a map for a project, real-world + vignette illustrations of places. Ideally looking at relatively toned-down styles and colour schemes (kinda thinking sketchy or watercoloury though I'm open to thoughts). Absolutely no AI obviously.

Further details here:
https://exilian.co.uk/forum/index.php?msg=171307

If this might be you, get in touch! If this isn't you, I'd really appreciate a boost on this so I can find someone.

#MastoArt #Artist #Commission #Commissions

exilian.co.uk

14

6

80

0

James Baillie @JubalBarca@scholar.social

· 12mo ago

Replying to

Further PSA 1, how to do account deletion on Academia.edu without agreeing to the new terms of service on the way: If you want to delete your account, open academia.edu, do NOT click accept to the new terms, but instead click "privacy policy". You can then go to the top left dropdown menu with your name on it, click "account settings", scroll to the bottom of the page, and there's the account deletion option. You can even leave them a little note if you want though I doubt they'll read it.

64

2

61

0

James Baillie @JubalBarca@scholar.social

· 3mo ago

Boosted by @trending@homestead.social

Fediverse people who know more about home maintenance than me: I'd quite like to clean this very grimy outdoor wall at the back of my flat which I can now access again after some repairs got done. I've brushed some dirt off, but it's still visibly grimy and given what I think is just paint on slightly crumbly concrete + my joints not being great I don't know if soapy water and a bucket is the best option or if I should consider other ideas. Boosts appreciated :) #home #DIY #cleaning

5

0

15

0

James Baillie @JubalBarca@scholar.social

· 3mo ago

Replying to

@toerror@mastodon.gamedev.place I wish they'd just go and do actual stress eating instead honestly. I will happily personally send Keir Starmer several packets of snacks and a large box of tea bags if it means we get better governance.

5

0

1

0

James Baillie @JubalBarca@scholar.social

· 12mo ago

Replying to

Further PSA 2, where do we go instead? Probably the current best option is Knowledge Commons, which is run as a nonprofit and can be found at hcommons.org (they also run the hcommons.social mastodon server). Knowledge Commons gives you essentially the same profile-building tools as Academia with far less slop and more control over what's going on.

28

3

22

0

James Baillie @JubalBarca@scholar.social

· 6mo ago

Replying to

@Mimesatwork @Catvalente I'd agree with this - I mean, I certainly agree hallucinate is also the wrong word, but error to me implies something more fixable and programmatic, and less inherent and random, than what's going on.

7

1

0

James Baillie @JubalBarca@scholar.social

· 3mo ago

Replying to

@LeonianUniverse@leonianplace.masto.host Better parental efforts would be great, but a core problem with this as a solution to all these problems is that a lot of parents are very clearly not capable of doing it effectively. A lot of adults are not capable of protecting *themselves* from online harms: sometimes, too, the adults in a child's life are the actual problem that they need to turn to the internet for help with, rather than vice versa. My experience was not dissimilar to yours, but that's not universal.

3

1

2

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Finally properly home from today onward. I've not had a full seven day week at home since early March, which feels like a long time at this point.

4

1

0

James Baillie @JubalBarca@scholar.social

· 3mo ago

Replying to

@matthewbdaly@mstdn.social @LeonianUniverse@leonianplace.masto.host Yeah, I think that would help. I do still think that the big platforms need to take more responsibility for content though: I think if people have actively signed up to see stuff, that's one thing, but if a platform chooses to algorithmically push content at people or create their own content to push (incl. AI summaries) they should be held responsible for disinformation and harmful content therein.

2

0

1

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

New solution to the problem of how to get all my gear back to base in first person RPGs whilst wildly overencumbered

#gaming #games #fallout

scholar.social

3

1

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@MsHearthWitch@wandering.shop Thank you so much! Email recieved, running around doing stuff this evening but will get a reply to you with more details tomorrow/when I'm able to :)

2

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Aaaagh, I have had yet another stupid case of someone screwing with Wikipedia edits I've done for no good reason whatsoever.

Literally it's about fifty percent of edits on that place that someone will mess with (and yes, I mean it's fifty percent when the edits are perfectly in line with every minutia of Wikipedia policy), and then people will whine that nobody wants to support/edit it and ???

What do you expect when you run a project that asks for work then trashes the work consistently?

2

0

1

0

James Baillie @JubalBarca@scholar.social

· 6mo ago

Replying to

@futtta These days I keep multiple draft forms of things in case this comes up.

3

1

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@vmaderna Alerting @jtheseamstress to the buns!

2

1

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Replying to

@anderseknert This is truly and deeply weird.

4

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@mia@hcommons.social @sarahalang@fedihum.org That's a very fair point - there were a couple of other factors involved, including that it would have meant back to back conference weeks in different countries and that the acceptance was for a different format to the proposal, submitted a talk and got a poster slot (which is a totally fair system, I'm just a lot less comfortable presenting posters and doubly so if I want to make a point that's structured/was envisaged talk-style). But the feedback was a discouraging element.

2

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@sarahalang@fedihum.org I didn't bother going to last year's DH conference, and that was in part because I felt from the reviewer comments that the workflow I was hoping to present was either just going to get pointless "why isn't this using LLMs" questions thrown at it all weekend or was going to get no interest: the bandwagon approach to these things is very tiring and I wish we spent more of our time as DH people thinking sharply about what tools do, including to our thinking, and how they do that.

2

0

James Baillie @JubalBarca@scholar.social

· 4mo ago

Replying to

@joannaholman@aus.social Quite a few different species of them around the world too! Though I've only ever seen the European ones.

1

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Replying to

@Canageek Yes, I sometimes wonder about things like that. Even just having much shorter grant applications and review processes where everything gets a "is this basically sane, OK, it goes in the pool and then it's a lottery" would probably produce basically the same outcomes as now with considerably less work for everyone. I fear academia is so desperate to believe in the possibility of absolute scientifically (or rather, scientistically) ranked meritocracy that it's killing us sometimes.

2

3

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Replying to

@tobifant @leeloo Whilst we obviously can't show if humans have a soul, we can absolutely show that humans have e.g. abstracted concept frameworks that are not solely based on averages of language statistics. I understand what an "owl" is, for example, in a way separate to the numerical relationships between the word "owl" and other words. That is a really fundamental information processing difference and allows me to construct *novel* understandings of that concept in ways that an LLM couldn't.

2

0

1

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@mayintoronto@beige.party I'm in my early thirties (32) and yeah, I don't feel like I'm obviously The Little Childer but I'm definitely at the young end of the norm here still.

1

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@bluebudgie@mastodon.art Love the leafy seadragon :)

1

0

James Baillie @JubalBarca@scholar.social

· 9mo ago

Replying to

@KazuShuSora@social.tchncs.de @TheEddieShow@beige.party It'd be perfectly technically viable to a) random-assign each new user to one of a smallish pool of larger defaults rather than direct them all to the same default and b) ensure that sign up pages have a "explore other servers" button that's unobtrusive but visible.

3

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@skinnylatte yeah it has turned out that I am far from a minimalist on travel. Between carrying my own shampoo and my fretting over what to pack, having the whole case makes sense.

1

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@wizardneedsfood I'm assuming you'd consider a chocolate lava cake to be cheating in this endeavour? :)

1

2

0

James Baillie @JubalBarca@scholar.social

· 12mo ago

Replying to

Further PSA 3: Mitigations/Caveats to the above:

Sites do need a rights clause to be able to share your posts.
The rights clauses have two limiters: first, you grant rights "in connection with operating and providing the Services": but this explicitly includes making derivative works, which I suspect aren't thus limited.
Second, Academia agrees not to sell user content for a profit (but, again, derivatives)

I'm unconvinced that these limiters protect against AI-churning content at all.

5

2

6

0

James Baillie @JubalBarca@scholar.social

· 12mo ago

Replying to

PSA 3a: Other notes

As noted earlier, it's possible that a bunch of this stuff is unenforceable in some jurisdictions. I don't advise testing this, lawyers are expensive.
For comparison, the hcommons terms could do better on delimiters: they could set out more limits on content reuse. One advantage the hcommons ToS has though is that their license to use content stops if you remove the content: conversely the Academia licenses are irrevocable.

(Nb: am not a lawyer, so take my notes w/salt)

5

0

7

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@Anke@social.scribblers.club Yes, it was multiple things - a conference, then a few days, then a month of research trip, then a couple of days then a short holiday trip, and I have seen interesting things and thought interesting thoughts and now I am Quite Tired but also I need to get on and write lots of things from it all.

1

0

James Baillie @JubalBarca@scholar.social

· 5mo ago

Replying to

@sarahalang@fedihum.org In any case - this is a really good post, thank you for sharing it!

1

0

James Baillie @JubalBarca@scholar.social

· 8mo ago

Replying to

@futurebird@sauropods.win Stalling is so key. Esp if you have a boss who is insistent but doesn't actually have the technical competency to naysay you. It goes on a work board for someone to look into it. Someone needs to check feasibility, someone needs to check data regulations, someone else needs to check preferred implementations. You need clarification from another department on how this would work with what they're doing. People move on, you quietly drop it in a year's time. All sand in the gears.

2

0

3

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Replying to

@bitchboss I agree with the endeavour, though I don't know if I know many people who have personal ChatGPT subscriptions... I've always presumed that besides a small number of hypers, the money inflow for OpenAI is mostly either investor finance, or sales directly to companies who think it will let them fire more staff, or deals with other big tech companies to shove unwanted GPT into their products.

1

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Replying to

@jaranta I guess I was just trying to get at what your instinctive definition of a podcast entailed (in that mine would also not be "any audio file" but I'm not sure if mine would be the same as yours).

1

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Replying to

@jaranta are you differentiating between "multi episode, public, discussion format" and "any audio file thingy"?

1

3

0

James Baillie @JubalBarca@scholar.social

· 9mo ago

Replying to

@vicorva Always happy to assist in this regard :)

1

0

James Baillie @JubalBarca@scholar.social

· 9mo ago

Replying to

@nblr Mandarin ducks in mating plumage are some of the most absurd things and I love them. They're originally an east Asian species but there's now quite a significant population in Germany in particular due to escapes over time because too many posh Europeans wanted to own the pretty ducks.

1

0

James Baillie @JubalBarca@scholar.social

· 32mo ago

Replying to

@shauna@social.coop I actually kinda used this concept for making a character select screen in a videogame: rather than have an idealised "man" or "woman" image, the character gender selector is a planetary chart and if you select man it highlights the Mars orbit, if you select a woman it highlights Venus, and if you choose a nonbinary character it highlights all the other orbits at the same time

2

0

James Baillie @JubalBarca@scholar.social

· 3mo ago

Replying to

@kneoghau@mastodon.social Sadly I doubt this will change anything about the government's relationship with particular social media sites. And yes, it's a miserable situation all round.

0

James Baillie @JubalBarca@scholar.social

· 7mo ago

Replying to

@Canageek Mm, yes, those are fair points. I think I'd think of it as more "similar results to random" because I think that usually for my sort of work whose desk a grant lands on to review and what mood they're in when they do it is often more important than how good the writing is, so that adds a pseudorandom component separate to testing the grant writing skill-set.

0

1

0

James Baillie @JubalBarca@scholar.social

· 3mo ago

Replying to

@liseo@kuupiaq.social @FreakyFwoof@universeodon.com I think things have to be tackled more at the content and algorithms level. The problem differs for different kinds of content: scams and political disinfo don't tend to get "driven underground" for example because the underground doesn't have enough marks to make those things viable. Illegal harms type content does go underground, but in some cases just reducing public exposure there is a net benefit. There are more complex cases though.

0

1

0

James Baillie @JubalBarca@scholar.social

· 22mo ago

Replying to

@nemobis @Bibliothecaris I'd guess not much, because I think most academics in relatively niche fields know if their immediate peers are here already and professionally that's what counts. Like, I like going and chatting to philosophers or physicists or scholars of C20th literature, but even if one found another scholar of some part of medieval W & SW Asia there's only so much % chance we'd really overlap on any given thing.

0

2

0

James Baillie @JubalBarca@scholar.social

· 22mo ago

Replying to

@nemobis I think people usually announce when they set up somewhere new, and work on searching tags or posts to find fellow people in the field - and once you find one person in your academic network, you can check their follows and find anyone in the field they're connected to etc. It's also something we talk about fairly often, given the pressures towards public engagement, and people often put their preferred social media handles on the last slides of talks and things like that.

0