C2PA produces signals to measure “trustworthiness” of content. C2PA has a technical “Trust Model” involving “Trust Lists” (and a “conformance program” in development).
Makaya McCraven &band were well worth wait. Won’t forget it!
And Shudder to Think (I love Dischord era) were so good. Craig still rules. Great to see Sohrab from Obits too :)
www.makayamccraven.com/offtherecord
Makaya McCraven &band were well worth wait. Won’t forget it!
And Shudder to Think (I love Dischord era) were so good. Craig still rules. Great to see Sohrab from Obits too :)
www.makayamccraven.com/offtherecord
purplishrecords.bandcamp.com/album/batch-...
purplishrecords.bandcamp.com/album/batch-...
Sound familiar? Except this was more than 30 years ago:
Sound familiar? Except this was more than 30 years ago:
I'm digging into layers upon layers of metrics baked into one trademarked set of psychometric measures.
I'm digging into layers upon layers of metrics baked into one trademarked set of psychometric measures.
When I started my journalism career, "bloggers" were discredited by some. Are today's Substackers journalists? What about indie media outlets?
When I started my journalism career, "bloggers" were discredited by some. Are today's Substackers journalists? What about indie media outlets?
Played this one as an encore, a fave:
Played this one as an encore, a fave:
"Some experts are skeptical if holographic avatars are even needed."
"Some experts are skeptical if holographic avatars are even needed."
Sharing this insightful talk re: problems with those AIAs & their actual use -
Sharing this insightful talk re: problems with those AIAs & their actual use -
Submissions to huge AI conf NeurIPS MORE THAN DOUBLED since 2020:
9,467 in 2020
⤴️Whopping 21,575 in 2025!
Datasets & benchmarks track submits (data to train, measures AI makers devise to grade own homework) MORE THAN DOUBLED since 2023:
987 in 2023
⤴️Whopping 1,995 in 2025!
Submissions to huge AI conf NeurIPS MORE THAN DOUBLED since 2020:
9,467 in 2020
⤴️Whopping 21,575 in 2025!
Datasets & benchmarks track submits (data to train, measures AI makers devise to grade own homework) MORE THAN DOUBLED since 2023:
987 in 2023
⤴️Whopping 1,995 in 2025!
Dangerous situation? You be the judge:
Dangerous situation? You be the judge:
story in @404media.co www.404media.co/pissed-off-f...
story in @404media.co www.404media.co/pissed-off-f...
I kinda wound down my digital political ad reporting around time these political ad transparency things started. My post Obama '08 book captures brief moment in time in the evolving history of online politics (www.goodreads.com/book/show/65...).
I kinda wound down my digital political ad reporting around time these political ad transparency things started. My post Obama '08 book captures brief moment in time in the evolving history of online politics (www.goodreads.com/book/show/65...).
I'm also considering design of C2PA as content provenance/ AI disinfo infrastructure in this context. worldprivacyforum.org/posts/privac...
I'm also considering design of C2PA as content provenance/ AI disinfo infrastructure in this context. worldprivacyforum.org/posts/privac...
"I need to upgrade my image editing software, but my old favourite now uses AI, and I'm not paying money for that. What are people's recommendations for human-centred image editing software?"
"I need to upgrade my image editing software, but my old favourite now uses AI, and I'm not paying money for that. What are people's recommendations for human-centred image editing software?"
Interested in this in relation to my C2PA research. C2PA-based CAWG creator assertions do something similar:
Interested in this in relation to my C2PA research. C2PA-based CAWG creator assertions do something similar:
TRUST SCALE SCORES ARE RELATED SIGNIFICANTLY TO POSITION IN THE FAMILY, SOCIOECONOMIC LEVEL, RELIGION, AND RELIGIOUS DIFFERENCES BETWEEN PARENTS.
TRUST SCALE SCORES ARE RELATED SIGNIFICANTLY TO POSITION IN THE FAMILY, SOCIOECONOMIC LEVEL, RELIGION, AND RELIGIOUS DIFFERENCES BETWEEN PARENTS.
C2PA produces signals to measure “trustworthiness” of content. C2PA has a technical “Trust Model” involving “Trust Lists” (and a “conformance program” in development).
C2PA produces signals to measure “trustworthiness” of content. C2PA has a technical “Trust Model” involving “Trust Lists” (and a “conformance program” in development).