4 ms·
"The goal of this study is to better understand children’s everyday learning experiences and to develop Al tools that can help assess classroom interaction qual
by pj_mukh 5mo ago
"The goal of this study is to better understand children’s everyday learning experiences and to develop Al tools that can help assess classroom interaction quality"
This..feels like a worthy goal. Just switch the study to opt-in instead of opt-out and 404media can't get the "ER-MEH-GERD AI" click bait. This is not generative AI and the lead-researcher is an early-childhood education specialist [1].
My kid went to a UC Berkeley adjacent pre-school and had researchers come by all the time, filming and taking notes (after consent forms were signed). They would share their studies with us later, it was very illuminating.
[1]: https://education.uw.edu/about/directory/gail-e-joseph https://education.uw.edu/about/directory/gail-e-joseph
- krupan 5mo agoStudies are great. Free training material for commercial AI products, not so great
- pj_mukh 5mo ago“ material for commercial AI products” Where was there any indication this was that?
- Eddy_Viscosity2 5mo ago> "to develop Al tools that can help assess classroom interaction quality". This..feels like a worthy goal. Does it though? Do you really want everything, including caring for young children, to be quantified into metrics assessed by a computer?
- pj_mukh 5mo agoWe don’t have to make that number a goal. Contrary to popular pessimistic thought, not everything we measure becomes a goal. Most things we measure are actually ignored. I think to pre-emptively disregard research like this because “well it could be misused” is folly
- Eddy_Viscosity2 5mo agoThings that can be measured easily are often transformed into goals. Hence standardized test scores for measuring education, and number of papers published to measuring a researcher's performance. AI makes it even easier while simultaneously removing accountability to the managers using it. And there will always be managers who will use such tools. Couple this with AI vendors and their cash piles lobbying every public and private institutions like crazy and the foreseeable is easy to foresee.
- asdff 5mo ago>My kid went to a UC Berkeley adjacent pre-school and had researchers come by all the time, filming and taking notes (after consent forms were signed). They would share their studies with us later, it was very illuminating. My undergrad had something similar. I took a class through it for some general credit. For one of our assignments we'd observe and take notes, sitting in the gallery space, which was a sort of catwalk above the classroom hidden behind a one way mirror. The whole classroom was microphoned and we could tune in to certain areas of the room to listen in. I am not sure if the preschoolers themselves were aware of the surveillance as probably that would influence their behavior.
- skinfaxi 5mo agoHow does opt-out work? How can I guarantee my kid's likeness will not be recorded by some company? Slapping a sticker on then won't cut it since they would still be collecting the video/audio and then having to do something after the fact, right?
- pj_mukh 5mo agoYou can’t guarantee any of that in any footage taken in public. Or public school. Just a good faith attempt. Again just want to point out this is a university researcher trying to measure better ways to teach kids not Flock trying to get started early on their school to prison pipeline. I feel like we’ve lost the plot here.
- skinfaxi 5mo agoYou are conflating things with your first sentence. There are federal laws about schools and their agents recording students as it relates to their education. We are talking about filming children in a classroom not a highschool basketball game open to the public. We haven't lost the plot. How do you think things like Flock came to be? Incrementally or all at once?
- pj_mukh 5mo ago“There are federal laws about schools and their agents recording students as it relates to their education” And there is no indication that the professor here planned to violate any of them. “How do you think things like Flock came to be?” Not from a benign study about better teaching methods, or of anything akin to it. Video is a rich data source. I hope researches get to use it more. I’m not selling all that out just because one completely unrelated private company misuses data.
- rolph 5mo agohow much propriety over information does a teacher mounted with someone elses camera, have available, to counter mispropriety by a third party?
- onetokeoverthe 5mo ago[dead]