This was a nice surprise. Malcolm @Gladwell and @BillSimmons spent some time discussing our (w @julieliss) 2017 paper on concerning declines in linguistic complexity in NFL players on @RingerPodcasts.
Paper is here: https://t.co/33rshGC1rr
Thanks @ETSshow for letting me know.
@sayashk@random_walker@RishiBommasani I agree C2PA is a step forward but not foolproof—you can still capture AI content on C2PA devices.
We need deeper provenance, like verifying at the time of recording that recorded speech is coming from a live human moving their lips (similar for images), to ensure authenticity.
@julieliss and I have been working together for over a decade on clinical speech analytics. This perspectives paper highlights the insights we’ve gained along the way.
How do we responsibly build clinical speech AI models that translate?
By bridging the gap between speech science & clinical research... A new article by @visarASU & @julieliss shows how this reduces training data size and improves model interpretability.
https://t.co/ywVqlt5PBo
New article! Read Responsible development of clinical speech AI: Bridging the gap between clinical research and technology https://t.co/LCeCMY7gVn in npj Digital Med
One key takeaway from this post is how the hype cycle fuels publication bias in ML science, leading to overoptimistic results. Our new paper provides a solution to estimate realistic model performance across various fields: https://t.co/XwKaNwEB4S
CC: @GautamDasarathy ,@visarASU
Powered by the 🇺🇸 US Department of State @StateDept and administered by @ASU, GIST Innovates the Balkans 2024 kicks-off tomorrow in 🇲🇰 Skopje, gathering 12 companies from the W. Balkans. https://t.co/wHeng8tHyK
@GISTNetwork@USEmbassySkopje@visarASU@uranikbegu
Amy Coney Barrett shocked during oral arguments that Idaho law would keep a doctor from performing an abortion unless the mother was immediately dying right at the exact moment she’s at the hospital.
Didn’t she vote to allow this to happen in the first place? 😒
Meta AI claims to have a child in a NYC public school and share their child's experience with the teachers! The reply is in response to a question looking for personal feedback in a private Facebook group for parents. Also, Meta's algorithm ranks it as the top comment! @AIatMeta
Introducing our first 2024 #TIME100 cover star: Dua Lipa (@DUALIPA).
The pop star talked to TIME about her new album, her childhood in Kosovo, and the power of manifestation https://t.co/Z9EgXmS8ah
The crappiness of the Humane AI Pin reported here is a great example of the underappreciated capability-reliability distinction in gen AI. If AI could *reliably* do all the things it's *capable* of, it would truly be a sweeping economic transformation.
https://t.co/h3Pu4fFtH0
Submitting health-related work to @ISCAInterspeech? Remember to select Topic 14.05, "Speech and Language in Health: From Remote Monitoring to Medical Conversations (Special Session)", in the submission system!
#ASUEngineering has 4 new associate deans to guide @ASUEngineering's rapid growth in inclusive excellence, research and commercialization, global engagement and undergrad programs. 2 are electrical engineering faculty members @visarASU and Michael Goryll.
https://t.co/lInnVhOI2Q
I'm excited to announce our special session on "Speech and Language in Health: from Remote Monitoring to Medical Conversations" is going to be held for a 3rd time @ISCAInterspeech this year. Details at: https://t.co/eUgsNFCApZ.
👋 Greetings from 🇺🇸 Arizona, where six (6) innovative startups from 🇽🇰🇲🇰🇷🇸 are attending the two-week acceleration program at the Arizona State University @ASU. Part of the "GIST Innovated the Balkans" program, Powered by the @StateDept@GISTNetwork and ASU.
Nominations for #ISCA’s Distinguished Lecturers 2024-25 now open !
This outreach scheme supports leading researchers in Speech Communication to deliver lecture tours in regions where our field is under-represented.
Nominations : [email protected] before 31/01/24
This article looks at how good the predictions of AlphaFold actually are. The tldr is in the title: "AlphaFold predictions are valuable hypotheses and accelerate but do not replace experimental structure determination"
https://t.co/FHmSkDHNHp