The article: https://www.nature.com/articles/d41586-025-04064-7 *>in August, I temporarily disabled the ‘data consent’ option because I wanted to see whether I would still have access to all of the model’s functions if I did not provide OpenAI with my data. At that moment, all of my chats were permanently deleted and the project folders were emptied — two years of carefully structured academic work disappeared. No warning appeared. * Well, that sucks…
And now I’m hooked on Angela’s videos… thanks! 👍
I’ve been following her for a couple of years. Her videos are great and often very funny.
“Compound interest is really interesting”, LMFAO Back to roasting “Picard”.
It’s Fun February! Compound interest is fun!
I appreciate the article link. I’m obviously not pro AI, but the framing in the video is really uncharitable. Without seeing the exact text of the prompt to disable data consent, I can totally see it being ambiguous about deleting data. If you’re accustomed to opt out prompts and it does not explicitly use “permanently delete all your chat history” or something similar it’s possible you thought it was as innocuous as a cookie opt out. If it is ambiguous then this article serves the exact purpose it was written for. She opens by saying she appreciates articles where people are open about embarrassing topics but then lambasts this guy, several times for things I didn’t see him admit to. She takes issue with him using the word conversations as though it implies he is besties with it but they are very casually referred to as conversations and she does so herself. I totally understand not liking the way this guy does his job, or even thinking this means he’s not even doing his job, but she seems to be making a lot of assumptions. She’s probably right on several points, but the video might as well have been written by AI because it just regurgitates my opinions on AI back to me with no real relevance to the article she’s talking about. She also on multiple occasions just casually drops that asking AI to write code for you, with the presumption that you don’t understand it, is fine actually. She does this unironically. I almost stopped watching but wanted to finish it in case there was a gotcha at the end about media literacy. There was not.
The media literacy lesson is, as I understood, when you disable data consent then you can expect them to sever the data connection (or whatever they wrote in their 11,344 page TOS, but one could ask AI for a summary). Also expect it to still be stored on their server to feed the AI though. Or, the delete is surprisingly doing what the delete stands for. This guy is a professor, not some hobbyist.
“I wanted to see whether I would still have access”, did he have a contingency plan in case he would indeed lose this access?
He’s a “professor of plant molecular physiology at the University of Cologne and member of the CEPLAS Cluster of Excellence”. He published articles contain stuff like RNA sequencing, HIV-1 strain binding, and stuff in that range. It’s not LEGO, so let’s hope he does his what-if experiments (if he does them outside the AI environment) on plant antibody combo’s in well regulated and closed environments… or-else.
I think her criticism is well deserved.
Him pressing the button had nothing to do with media literacy. Neither she nor the article bother to include a picture of the popup or its direct verbiage so we have no idea what it even said.
It’s almost like that sentence you quoted continues “I temporarily disabled the ‘data consent’ option because I wanted to see whether I would still have access to all of the model’s functions”. Though I’m sure you know how it ends and conveniently left it out because it doesn’t support your view. Plenty of sites or applications limit functionality if you decide not to share your data. The professor wanted to know what functionality they might be unable to use going forward. That generally has nothing to do with preexisting data. If I turn off location data for my photos, it doesn’t retroactively remove the data from previous photos. It’s not unbelievable to expect to still have access to previous chats.
Suggesting he completely fabricated or used AI to generate his data is unfounded based on the article and his statements.
You never touched on her commenting about coding and never even addressed the point that she didn’t bother to even look at what the button said before making a video about it. You’re free to think her criticism is well deserved, but I do not think she bothered to justify her video beyond “AI bad”. You can dislike AI while still being critical of people who agree with you on that point.
this dumbass was using ChatGPT like it was Obsidian ROFL
take his degrees away
A cautionary tale against not having backups***
