DC resident, gov researcher in trustworthy and responsible AI, background in linguistics, phonetics, and human language technology, trying to help people understand what socio -technical even means, personal account (she/her) :dcflag:
In this week’s Disconnect Roundup, I talk about a technology I actually think is good and cool, plus provide a breakdown of what we know from the new Elon Musk documentary along with labor updates and other news you might have missed!
https://www.disconnect.blog/p/roundup-a-tech-that-might-save-us
Social media makes it sound like every single person in America either slavishly uses AI bots or passionately hates them, but it turns out that most Americans have never used an LLM, don't really know what it is, and couldn't care less. https://techcrunch.com/2023/08/28/survey-finds-relatively-few-americans-actually-use-or-fear-chatgpt/
«Automation blindness is inescapable – and it's the inconvenient truth that #AI boosters conspicuously fail to mention when they are discussing how they will justify the trillion-dollar valuations they ascribe to super-advanced autocomplete systems. Instead, they wave around "humans in the loop," using low-waged workers as props in a Big Store con, just a way to (temporarily) cool the marks.» @pluralistic’s great take on why human supervision of AI won’t work. https://pluralistic.net/2023/08/23/automation-blindness/
If you've ever wanted to consider joining federal service but aren't familiar with USAJobs, the government's website for job listings, now's the chance to learn all about it!
The Office of Personnel Management is hosting an information session on using USAJobs this Wednesday, August 16th from 11:30-1pm ET!
https://www.eventbrite.com/e/navigating-usajobs-tickets-691674094837?aff=oddtdtcreator
Interesting piece by @robinsonmeyer about a climate measure tucked into last year's appropriations bill that will make the U.S. government the first in the world to pay companies to remove carbon dioxide from the air. The initial funding isn't much, but Meyer explains that government funding over time can be an outsized lever to increase adoption of CO2 capture.
https://heatmap.news/economy/carbon-removal-us-government
1/
OpenAI quietly shut down its "AI" detector.
Did shutting it down undo the harms?
No, its Algorithmic Imprint lives on.
Here's how ⤵️
#AI #ML #chatgpt #mastodon #responsibleAI #AIethics #generativeAI
I am not enamored of arguments saying that the solution to AI problems is "educating the users" to be more skeptical, or to fact check, etc. Education is not going to solve the problems. We've been trying to educate people about account security and phishes and other scams for decades, and about misinformation for years, and it's all only getting worse. Most people are nontechnical, have busy lives, and do not have the background nor interest to learn about how these systems work, any more than how a toaster works. It is the responsibility of these systems to be safe, not to try push the responsibility onto the users.
Angela Davis says it best:
“Virtually every institution seized upon that term, ‘diversity.’ And I always ask, ‘Well, where is justice here?’ Are you simply going to ask those who have been marginalized or subjugated to come inside of the institution and participate in the same process that led precisely to their marginalization? Diversity and inclusion without substantive change, without radical change, accomplishes nothing.”
https://jezebel.com/angela-davis-on-racial-reckoning-diversity-and-inclusi-1844853369
@jcrabapple this was amazing
i see the future of computing is progressing well 🫠
@jcrabapple that’s great to know. We’re always looking for close-ish low light pollution areas.
@jcrabapple where on the eastern shore did you take these?
The Fourth of July in Petworth continues to be a spectacle.
new paper from @abebab shows that the larger you scale up the data in your web-scraped models the more hate and racism you get: https://arxiv.org/abs/2306.13141
@PollyWaugh 😆 bridging the gap!
Saturn’s rings steal the show in new image from Webb telescope
Webb turned its gold-coated mirror toward Saturn this week.
Ruha Benjamin on ‘deep learning’ at PublicSpaces conference: ‘Computational depth without historical and sociological depth is superficial learning.’ #pubconf2023
Algorithms like #ChatGPT
do not think,
they do not dream,
they do not lie,
they don't pretend,
they don't hallucinate, and
they don't imagine.
Because they are algorithms.
Using these deeply human metaphors is a fundamental mistake that distorts what we can talk about when we talk about "#AI". Unfortunately it's all over the media and academic research too. https://doi.org/10.7551/mitpress/14668.003.0008
“Put another way, ChatGPT seems so human because it was trained by an AI that was mimicking humans who were rating an AI that was mimicking humans who were pretending to be a better version of an AI that was trained on human writing.”
