Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
jonny saunders Profile picture
jonny saunders
@json_dirs
Find me on the free web: 🐘personal: @jonny@social.coop work: @jonny@neuromatch.social bsky: https://t.co/uVO6PaxwQu
Dec 5, 2022 • 6 tweets • 2 min read
this is a breathtakingly clueless account of mass production. "they will only displace everything but the most esoteric" is not a defense, but a direct indictment of the technology - everyone but the obscenely wealthy eat out of machine-made ceramic bowls. "This technology will only automate every place that language and computation are done routinely, leaving only the truly artisanal uses of language and computation with value." So who in this fantasy world survives? what is the vision of the future from the brainwashed AI class?
Dec 5, 2022 • 9 tweets • 2 min read
gpt3 is going to make the world unlivable so quickly. we already have ghost kitchens and dropshipping, adversarial swarms of bots fine tuned on stock market data hallucinating entire industries against each other with remotely operated human labor are literally inevitable.
!/ academics blithely dismissing this must have never seen the contemporary landscape of scams. a shocking proportion of digital retail is already dropship bots doing arbitrage, but what happens when the bots are convincing enough to hire people and manage a workforce? !/
Dec 3, 2022 • 4 tweets • 1 min read
personality quiz: academic workers are on strike for living wages. You are a PI, your role is to:
- Condescendingly explain to the foolish ECRs that the NIH's budget is set by Congress and of course there is nothing you can do, personally, by yourself.
- Do any organizing at all weird how committed some people are to believing they're powerless that it's easier to assume 48,000 people formed unions and organized a strike because they don't understand the system they literally have to understsnd to keep their jobs
instead of...
"maybe I can do that too"
Nov 30, 2022 • 7 tweets • 3 min read
seeing people wondering how to pay exorbitant APCs and how they'll pay their grad students back to back on the tl. They are of course related: the result of shunting the costs of infrastructural deficits onto grad students almost always paid poverty wages.
arxiv.org/abs/2209.07493 Impacts on the daily experience of researchers include: A pThe complicity of scientists in rendering our collective intIndustry capture has crept into ever more of the little gran needing to pay ever increasing proportions of your grant to APCs, cloud, hosting, and SaaS subscriptions, grad students needing to take several extra years to finish because of our entirely dysfunctional incentives for toolbuilding are all symptoms of infrastructural deficits.
Oct 3, 2022 • 4 tweets • 1 min read
uncle phil "campus for privatization of the commons" and "arena of pricing a generation of students out of public education through housing auxiliary loan obligations" journalists pull the 990 for the UO foundation during the years the track and field stadium was build and compare the disregarded entities section to the public statements of cost challenge 2022
Oct 3, 2022 • 9 tweets • 2 min read
the virgin dataviz blog -> data science -> adtech pipeline vs. the chad dataviz blog -> what the hell am i doing -> trying to abolish software pipeline standing under the bleachers after every R bloggers post being like "hey kid if ya wanna be cool you'll smoke this direct action against the surveillance industry"
Oct 3, 2022 • 7 tweets • 2 min read
so I'm in the process and writing a guide for deplatforming myself from everyday surveillance, stuff like notes, calendars, file backup, etc. does anyone have any requests or recommendations? !/ so far thinking about a two-part of rented hosting for public things and sbc with some external drives for private stuff, would love to see how low we can get the barriers to entry by being able to extend it out to multiple levels of documentation via wiki format !/
Oct 2, 2022 • 10 tweets • 5 min read
the logic of capital is just 🤌 in this 1989 paper about manufacturing pharmaceuticals in space and needing to invent new pharmaceuticals to cure the diseases caused by manufacturing pharmaceuticals in space !/
doi.org/10.1177/002580… In addition to the above re... [wall street guy voice] space factories? ya gadda problem. space insurance. ya can't get it. the satellites keep exploding. costs a fortune. Space Insurance Insurance c...
Oct 2, 2022 • 5 tweets • 1 min read
very curious what the legal case against scraping a database like this and rehosting it for free would be since at least in the US you can't copyright facts. I assume it would be related to the database technology and curation process, so maybe for the thought experiment assume you launder the data enough that it's unrecognizable and is just a really big database of chemical data
Oct 1, 2022 • 4 tweets • 1 min read
distributed triplet graph backbone for scholarly communication let's go been freakin screaming about this approach and it is time to build it!!!!!
Sep 8, 2022 • 4 tweets • 2 min read
RELX just trying to normalize AI peer review, nothing to see here, AI peer review is more fair. not about making self-fulfilling scientometrics built into hiring and granting decisions to control the process of science. not at all.
elsevier.com/connect/resear… Al and peer review For many, peer review is the phase of thTo read or not to read Al reviewed articles When we asked r it's the metrics and it's the infrastructure y'all, publications are mostly a byproduct of the industry -- get ready for the next (now) shit where every part of your work gets evaluated by some hostile algorithm checking compliance with their profit-extraction machine
Sep 7, 2022 • 7 tweets • 2 min read
so ya when i am saying piracy is a means of protest against the vertical integration of streaming/movie production platforms that is a culturally preserving, rather than destroying act, this thread is the kind of thing I mean context
Sep 7, 2022 • 7 tweets • 2 min read
"your research is not research because it looks weird and the writing isn't totally sterile" volume 75429544 In this case, our moderator... what is this trash, there's not even a top-level heading called "results!"
Sep 6, 2022 • 4 tweets • 2 min read
"so you have lots of undergrads, and grad students too. but the grad students will do things like demand you pay them and make neural nets that mock your writing style. so you're really very much focused on the undergrads." And then when I went to the... might revitalize the @schillbot3000 for one last ride since I'm out of here anyway
Jul 25, 2022 • 7 tweets • 3 min read
extremely depressing to me that anyone thinks microcommodifying every action of their life to serve Brands would be a *good* thing. the web3 dream is to have people self-organize the company store. !/ web3 is a nightmare but I think good dreams and shared purpose can be a salve for the overwhelming dread for our shrinking future. I think we underrate how platforms like these built on personal branding and viral outrage bubbles foreclose that possibility
Apr 14, 2022 • 4 tweets • 1 min read
scholar is a trap and siphons off the energy needed to build the systems of collective knowledge organization needed to truly challenge the traditional prestige journal system. we can dream bigger than citation networks, and even beyond the scientific paper as such. these are just traditionalized forms of links, and beyond gatekeepers maintaining them for us we can rebuild scientific communication such that we can navigate and negotiate them ourselves.
Feb 8, 2022 • 5 tweets • 1 min read
is it ok to click a phishing link if you think its concept is very charming [screenshot of a text message] Limited Discount of 50% on Wi now that is a deal
Feb 8, 2022 • 10 tweets • 4 min read
NIH STRIDES is predicated on short-term, multi-million dollar contracts with to AWS, Google Cloud, and MS Azure to teach scientists how to lock themselves into cloud storage and compute. We're paying money for classes on how to pay money so we can pay them more money later. every time a well-resourced lab says "we're busy with science and throwing it on AWS is cheap," we take another step towards slicing off another publishing-industry sized dead weight loss off all our (public) grant funding.
Jan 25, 2022 • 15 tweets • 6 min read
More fun publisher surveillance:
Elsevier embeds a hash in the PDF metadata that is *unique for each time a PDF is downloaded*, this is a diff between metadata from two of the same paper. Combined with access timestamps, they can uniquely identify the source of any shared PDFs. [A list of metadata for a PDF, the important field being two You can see for yourself using exiftool.
To remove all of the top-level metadata, you can use exiftool and qpdf:

exiftool -all:all= <path.pdf> -o <output1.pdf>
qpdf --linearize <output1.pdf> <output2.pdf>

To remove *all* metadata, you can use dangerzone or mat2
Dec 4, 2021 • 21 tweets • 6 min read
Of course Elsevier's "enhanced pdf viewer" tracks where you click, view, if you hide the page, etc. and then transmits a big base64 blob of events along with ID from University proxy when you leave. I'm sure straight to SciVal for sale.
Is this the way we want science to work? Screenshot of a paper opene...Screenshot of JSON of a HTT...Screenshot of a split apart... genuinely sad that avoiding/gaming surveillance to keep your Bench Performance Rankings in the fundable range might have to become part of basic scientific training.
Oct 29, 2021 • 13 tweets • 5 min read
find paper. click button. read paper. have paper. auto-update .bib-liographies. share annotations. talk. build knowledge together. remember this is the floor of what's possible, constantly crushed by information monopolies. wonder why it has to be so hard, how we become complicit If you use Zotero first of... the amount of money being spent to try and cram our collective imagination into a box where we somehow actually want to be able to log in with a federated graph identity and be tracked across the web just for some peace is way more than you think.