Eric R. Scott
@leafyericscott.hachyderm.io.ap.brid.gy
Scientific Programmer & Educator | Data Engineer with grantwitness.org | Mentor for @Posit Academy | Ecologist | Tea geek Posts are mine and do not represent my […] 🌉 bridged from ⁂ https://hachyderm.io/@LeafyEricScott, follow @ap.brid.gy to interact
"of the seven cited articles, two appeared to be "entirely made up and three others did not exist in the journals they were attributed to."" Judge humiliates RFK Jr. over filing loaded with AI-generated studies that don't exist - Raw Story https://www.rawstory.com/rfk-jr-2677804529/
The termites seem to know today is a special day. First overnight rain of the summer! #tucson
Revising a conference proceedings paper (from over a year ago...) and when I submitted it there were fewer than 8,500 Starlink satellites in orbit, out of about 12,500 total satellites. Today there are 11,102 Starlinks out of 16,978 total active satellites. Fuck.
About all those "jobs that will be created in your community if you let us build a data center there" https://www.wired.com/story/inside-metas-experiments-with-data-center-robots/
Inside Meta’s Push to Put Robots to Work in Data Centers
The company is testing robots that can swap cables, reset servers, and take on other tasks performed by technicians, fueling concerns among some workers that their jobs could be at risk.
wired.com
Update. "Trump cannot deport students for criticizing Israel, judge rules." https://apnews.com/article/israel-protests-deportations-ideological-d3694ba08ab714eb9441ca51a83fa57f #Israel #Trump #USLaw #USPol #USPolitics
Trump cannot deport students for criticizing Israel, judge rules
Citing the importance of the First Amendment, a federal judge in California says the U.S. government is unconstitutionally silencing critics of Israel and others in its quest to deport noncitizens who the government says have disrupted college campuses as they expressed their views.
apnews.com
Hey #bostodon ! How about a Tootup this Sunday at Notch Brewing in Brighton? Starting around 4pm? Special guest from the West @skinnylatte will be here!
Huh. So #Debian voted for Proposal E regarding the use of LLM generated code, which people seem to like to describe as a "middle ground" but which, in fact, places absolutely no limitations on the use of AI in any way, not even merely requiring disclosure. That sounds more like complete […]
Original post on chaosfem.tw
chaosfem.tw
So who is down for a mission.. I bet we can get more than 1 stuck!! 🤣😂
The degree of xenophobia towards south Asian people everywhere is disgusting. Regarding the missing people in Nepal’s recent floods, I’m seeing Chinese people in Singapore say that Indian Singaporeans are not really Singaporeans; and white Aussies saying Indian Aussies are not Aussies. I’m […]
Original post on hachyderm.io
hachyderm.io
I getting Really Fucking Sick And Tired of people saying, "So, you don't want to use AI? Then I guess you'd better stop using Linux or you're a hypocrite!" Fuck you, honestly, if you're saying that. There is no ethical consumption under capitalism, and some […] [Original post on federate.social]
I love that Trader Joes has a cereal called just "High Fiber Cereal". They're like, "look, we're not even going to pretend you'll enjoy this breakfast"
People can only take in information that they have capacity to deal with the emotional consequences of. The tone that you share information in greatly influences who will take it in. An optimistic tone will decrease the chance that cynics will take it in, and vice versa. A shaming or righteous […]
Original post on social.coop
social.coop
Talking to my colleagues about students using "AI" generated output sometimes feels like this... John: Why don't we use DDT to improve crop yields on farms anymore? Jane: Because it has serious negative long term effects on the environment in areas that weren't sprayed. John: I didn't ask […]
Original post on fediscience.org
fediscience.org
Grant Witness has a new website, with agency dashboards, reports, and a feed of our investigations you can follow with our newsletter or via @grantwitness.org. Give a follow! U.S. science needs all the eyes and hands we can muster as we enter election and budget season. bsky.app/profile/did:...
Grant Witness@grantwitness.org · 2w agoHello all! Grant Witness has a new website at grantwitness.org, and you can now subscribe to our newsletter and follow us at @grantwitness.org, or on whatever social feed you prefer (find them at linktr.ee/grantwitness)
@davidgerard this is where a lot of people in tech are right now and the cognitive dissonance is incredible “oh I’d never do slop because I” type lgtm in all the PRs, type the code myself on the chatbot’s instructions, only use it for tests, only use it for docs, blah blah blah and meanwhile […]
Original post on mas.to
mas.to
“…the pattern we found using data on more than 11,000 firm-year observations of large, publicly traded companies in the US between 1998 and 2013. Firms led by racializd women posted the strongest financial performance, followed by white women, then racialized men. White men, who made up most […]
Original post on mastodon.social
mastodon.social
"Pentagon slashes staff of R&D repository by nearly 80%." https://www.defenseone.com/policy/2025/08/defense-tech-office-cuts-staff-down-40/407209/ "The Pentagon will lay off nearly 80 percent of the staff of its Defense Technical Information Center [#DTIC], its repository of […]
Original post on fediscience.org
fediscience.org
Hello all! Grant Witness has a new website at grantwitness.org, and you can now subscribe to our newsletter and follow us at @grantwitness.org, or on whatever social feed you prefer (find them at linktr.ee/grantwitness)
Grant Witness
An independent, up-to-date, grant-level record of federal science funding disruptions in the United States.
grantwitness.org
@grantwitness.org has a new website (and a bridged Mastodon account!): https://grantwitness.org Follow for news on federal research grant disruptions and restorations. Our previous Quarto site got the job done, but this is a nice improvement in design, branding, and interactivity. #uspol […]
Original post on hachyderm.io
hachyderm.io
I co-authored a paper, "Testing and Evaluation of Agentic AI Systems In Military Command and Control", with an excellent crew. I had so much fun working on this and it's finally publication day. https://arxiv.org/abs/2608.20597
Testing and Evaluation of Agentic AI Systems In Military Command and Control
Agentic AI systems are being procured for military command and control (C2) under public commitments to rigorous testing and human oversight. Whether such commitments can be discharged depends on their supporting assurance case, which requires three elements: claims specifying the conditions for acceptability, evidence bearing on those claims, and an argument connecting the two. Through a structured review of 240 documented Testing and Evaluation (T&E;) practices, spanning eight evaluation dimensions and three lifecycle stages, we identify eight assumptions that established methods make about their test article, grouped into four clusters: system specifiability, stability, composability, and supervisability. Agentic properties weaken all eight assumptions. This erosion affects the argument connecting evidence to claims, not the claims or evidence themselves. As a result, test results may satisfy process requirements, but they do not warrant the inference from tested to fielded behavior. We derive ten assurance claims for the first three assumption clusters and assess whether current and emerging methods can address each, mapping operational consequences through five C2 scenarios. Supervisability is identified but not assessed here, since evidencing it depends on system stability results and human factors T&E; methods beyond the present scope. The documented record does not support broad claims about system-level behavior, but narrower claims remain recoverable in principle, contingent on mature methods: bounded mission envelopes, trajectory-grounded correctness, executable runtime constraints, and characterized run-to-run variance. Part of the evidentiary burden shifts into deployment, making the determination to field a continuing act. Where evidence cannot be generated, the residual uncertainty can be governed through defined expiry conditions and assigned ownership.
arxiv.org
Potting bench is taking shape! Still needs slats for the shelf and table top. I’ll paint it or stain and finish it in the winter when it’s not 120° in this room. #woodworking
I ate three new-to-me fruits today, so it was a good day! 1. Granadilla (Passiflora ligularis) 2. Ice cream bean (Inga edulis) 3. Momoncillo (Melicoccus bijugatus)
The most popular pattern on ravelry right now is simply called The Garment
I've added my name to the open letter calling for a public inquiry into the tragic death of Professor Jason Arday. Will you join me? goodlaw.social/625453
Jason Arday: Demand a public inquiry now | Good Law Project
Commission an immediate public inquiry into how to ensure a responsible press given its contribution to Dr Arday’s tragic death.
goodlaw.social
"Over the past year and a half, at least 28 federal datasets have been deleted and another 338 data collections have been modified, according to an online tracker released last week from dataindex.us, a group of data policy experts, researchers and developers that formed last year to monitor […]
Original post on fediscience.org
fediscience.org
Rising levels of carbon dioxide in the atmosphere may already be influencing human biology. New research has identified long-term changes in blood chemistry that appear to track rising atmospheric CO2, raising concerns that an important blood marker could approach the upper end of its healthy […]
Original post on c.im
c.im
