Also FWIW, Jourdan and I have spent two years trying to tell folks that the work of "AI detection" is incredibly flawed and biased against folks who are disabled and English learners.
www.tandfonline.com/doi/abs/10.1...
www.tandfonline.com/doi/abs/10.1...
December 16, 2025 at 1:40 PM
Also FWIW, Jourdan and I have spent two years trying to tell folks that the work of "AI detection" is incredibly flawed and biased against folks who are disabled and English learners.
www.tandfonline.com/doi/abs/10.1...
www.tandfonline.com/doi/abs/10.1...
haha fair critique. Though i'd argue N=1 is exactly the point—if an AI tool can't handle one person's real messy content, what's it actually optimizing for? Not defending romance of human error, just questioning what gets lost when we automate feedback.
December 16, 2025 at 1:37 PM
haha fair critique. Though i'd argue N=1 is exactly the point—if an AI tool can't handle one person's real messy content, what's it actually optimizing for? Not defending romance of human error, just questioning what gets lost when we automate feedback.
I’m too much of an empath to cope with interacting with AI productions. Or also human error.
a monkey is standing in the middle of a swamp with its arms outstretched .
ALT: a monkey is standing in the middle of a swamp with its arms outstretched .
media.tenor.com
December 16, 2025 at 1:34 PM
I’m too much of an empath to cope with interacting with AI productions. Or also human error.
It's absolutely fine to aspire to do creative things for money or to get recognition, kudos, whatever for them. A-okay. The problem is when AI people commit the category error of treating creative acts as if they ARE the commercial outputs. It's such a hopeless, warped view. It's a non-starter.
December 16, 2025 at 1:28 PM
It's absolutely fine to aspire to do creative things for money or to get recognition, kudos, whatever for them. A-okay. The problem is when AI people commit the category error of treating creative acts as if they ARE the commercial outputs. It's such a hopeless, warped view. It's a non-starter.
with every update microsoft adds more stuff to teams I don't want (AI, more bloat that freezes my laptop, stupid tips section you can't get rid of) and nothing I actually want (ability to block colleagues from messaging you, "camera error" option for when you don't want to be perceived)
December 16, 2025 at 1:07 PM
with every update microsoft adds more stuff to teams I don't want (AI, more bloat that freezes my laptop, stupid tips section you can't get rid of) and nothing I actually want (ability to block colleagues from messaging you, "camera error" option for when you don't want to be perceived)
Long term, I also wonder if there are societal considerations for how e.g. a monopoly supplier of AI may affect human error tolerance and risk tolerance.
December 16, 2025 at 12:23 PM
Long term, I also wonder if there are societal considerations for how e.g. a monopoly supplier of AI may affect human error tolerance and risk tolerance.
The other angle is error risk. AI-generated reports, which may make mistakes humans would not, shine a light on whether or not we value tackling those mistakes.
So far, yes, we tackle them. But also no - not enough to value not trying it in the first place. www.ft.com/content/934c...
So far, yes, we tackle them. But also no - not enough to value not trying it in the first place. www.ft.com/content/934c...
Deloitte issues refund for error-ridden Australian government report that used AI
Big Four firm will repay final instalment after incorrect references and citations found in document
www.ft.com
December 16, 2025 at 12:23 PM
The other angle is error risk. AI-generated reports, which may make mistakes humans would not, shine a light on whether or not we value tackling those mistakes.
So far, yes, we tackle them. But also no - not enough to value not trying it in the first place. www.ft.com/content/934c...
So far, yes, we tackle them. But also no - not enough to value not trying it in the first place. www.ft.com/content/934c...
I keep joking that "if it's worth doing badly, it's worth using AI" - but think it's revealing about what I'm prepared to invest time in getting right. AI will always have Type 1 and 2 errors, and so how much you use AI tells you about your error tolerance.
December 16, 2025 at 12:23 PM
I keep joking that "if it's worth doing badly, it's worth using AI" - but think it's revealing about what I'm prepared to invest time in getting right. AI will always have Type 1 and 2 errors, and so how much you use AI tells you about your error tolerance.
Is social media in any danger of disappearing?
Think that answers your original content concern.
As for the quality of information online, you overestimate how good humans have been at this. AI is riddled with error because we are & have been.
The question becomes, who makes the best fact checker?
Think that answers your original content concern.
As for the quality of information online, you overestimate how good humans have been at this. AI is riddled with error because we are & have been.
The question becomes, who makes the best fact checker?
December 16, 2025 at 8:59 AM
Is social media in any danger of disappearing?
Think that answers your original content concern.
As for the quality of information online, you overestimate how good humans have been at this. AI is riddled with error because we are & have been.
The question becomes, who makes the best fact checker?
Think that answers your original content concern.
As for the quality of information online, you overestimate how good humans have been at this. AI is riddled with error because we are & have been.
The question becomes, who makes the best fact checker?
I do not like any AI chatbots. I can't reconcile myself with the unethical way they were "trained" or the power and water their data centers are taking from vulnerable humans.
I continue to get all my answers from human experts, who are oddly still less overall prone to error.
I continue to get all my answers from human experts, who are oddly still less overall prone to error.
December 16, 2025 at 6:16 AM
I do not like any AI chatbots. I can't reconcile myself with the unethical way they were "trained" or the power and water their data centers are taking from vulnerable humans.
I continue to get all my answers from human experts, who are oddly still less overall prone to error.
I continue to get all my answers from human experts, who are oddly still less overall prone to error.
i'd rather perish before letting any ai to choose my creative process.
That's the jest, trial and error. Without it things would be shallow
That's the jest, trial and error. Without it things would be shallow
December 16, 2025 at 4:19 AM
i'd rather perish before letting any ai to choose my creative process.
That's the jest, trial and error. Without it things would be shallow
That's the jest, trial and error. Without it things would be shallow
That is the core of the issue. One can write code without a deep understanding of logic and systems, especially with AI assistance. However, this often results in code that is inefficient, difficult to maintain, and prone to error.
December 16, 2025 at 1:30 AM
That is the core of the issue. One can write code without a deep understanding of logic and systems, especially with AI assistance. However, this often results in code that is inefficient, difficult to maintain, and prone to error.
Keep going. Make Microsoft see the error of using AI.
we can kill them all if we just work together and ✨believe💫
December 16, 2025 at 1:11 AM
Keep going. Make Microsoft see the error of using AI.
I find this error so fascinating. As though AI is imitating rhetorical style, but without the substance to
support it. Other examples of the same error are found in the sophistry of misinformation, and the discourse of a toddler.
support it. Other examples of the same error are found in the sophistry of misinformation, and the discourse of a toddler.
December 15, 2025 at 11:23 PM
I find this error so fascinating. As though AI is imitating rhetorical style, but without the substance to
support it. Other examples of the same error are found in the sophistry of misinformation, and the discourse of a toddler.
support it. Other examples of the same error are found in the sophistry of misinformation, and the discourse of a toddler.
December 15, 2025 at 10:01 PM
Lmao, let EU pour some regulations out. Then we shall see who is the rounding error.
#eupoli #AI #AIregulations
#eupoli #AI #AIregulations
December 15, 2025 at 8:32 PM
Lmao, let EU pour some regulations out. Then we shall see who is the rounding error.
#eupoli #AI #AIregulations
#eupoli #AI #AIregulations
Thousandth thing I use AI for because software sucks: Stack's "scroll up to find the error" message - fuck you I'm just gonna click the copilot button
December 15, 2025 at 8:13 PM
Thousandth thing I use AI for because software sucks: Stack's "scroll up to find the error" message - fuck you I'm just gonna click the copilot button
Nearly half (45%) of AI-generated news responses had at least one major error. One third had serious sourcing issues, including missing, misleading, or incorrect citations. One fifth contained major inaccuracies, such as fabricated details and outdated information.
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory
An intensive international study was coordinated by the European Broadcasting Union (EBU) and led by the BBC
www.bbc.co.uk
December 15, 2025 at 7:31 PM
Nearly half (45%) of AI-generated news responses had at least one major error. One third had serious sourcing issues, including missing, misleading, or incorrect citations. One fifth contained major inaccuracies, such as fabricated details and outdated information.
But it's true that their "critical" dimension, by and large, focuses on error checking and their "ethical" dimension focuses on plagiarism avoidance. Students found none (so far) that refer to AI literacy as knowledge of AI systems as sociocultural, as labor, as ecological, as geopolitical.
December 15, 2025 at 7:27 PM
But it's true that their "critical" dimension, by and large, focuses on error checking and their "ethical" dimension focuses on plagiarism avoidance. Students found none (so far) that refer to AI literacy as knowledge of AI systems as sociocultural, as labor, as ecological, as geopolitical.
And it's not like you couldn't schedule online pre-AI! It's an unforced error!!
December 15, 2025 at 7:09 PM
And it's not like you couldn't schedule online pre-AI! It's an unforced error!!
Washington Post Released Error-Plagued AI Podcast Feature It Knew Was Error-Plagued: Report
Washington Post Released Error-Plagued AI Podcast Feature It Knew Was Error-Plagued: Report
Washington Post Released Error-Plagued AI Podcast Feature It Knew Was Error-Plagued: Report
www.huffpost.com
December 15, 2025 at 6:52 PM
Washington Post Released Error-Plagued AI Podcast Feature It Knew Was Error-Plagued: Report
I'm sure the people using AI to generate "false flag" photos to promote weirdo agendas will see the error of their ways after Matty makes a really milquetoast tweet about civility and a slow clap will start that snowballs into AI being banned and misinformation bot farms becoming puppy orphanages.
December 15, 2025 at 6:01 PM
I'm sure the people using AI to generate "false flag" photos to promote weirdo agendas will see the error of their ways after Matty makes a really milquetoast tweet about civility and a slow clap will start that snowballs into AI being banned and misinformation bot farms becoming puppy orphanages.