Co-lead (Near-Term Detection) at the Nucleic Acid Observatory in Boston. Speaking for myself unless I say otherwise.
jefftk
Karma: 21,884
It is possible that jefftk put in a lot of effort to make sure the generated vibe is as accurate as could reasonably be
I didn’t. I picked out a photo that I was going to use to illustrate the piece, one host asked me not to use it because of privacy, another suggested Ghiblifying it and made one quickly on their phone. We looked at it and thought it gave the right impression despite the many errors.
As far as I know, this could be a literal photo of the party in question, and it’s free: https://unsplash.com/photos/a-man-and-woman-dancing-in-a-room-with-tables-and-chairs-KpzGmDvzhS4
The vibe of the generated image is far closer to the real party than the image you linked.
Do you think it’s wrong to take pictures in public places and put them online, even if there are people in the background who didn’t consent? I think of this as a very normal thing to do, though it does make them available for LLM-training-scraping among other things.
And the toy is being used, seemingly without consent, on friends who have privacy concerns in a way that seemingly still violates their privacy
In the case of the house party, I asked the party hosts if I could post a photograph, and one of them suggested this method of anonymization. In the puddle picture I didn’t ask, but the kids are so generic that I really don’t see how it would violate their privacy.
Look at the fiddler’s arms
The left arm is holding the fiddle and is not visible behind my body, while the right arm has the sleeve rolled up above the elbow and you can see a tiny piece of the back of my right hand poking out above my forearm. The angle of the bow is slightly wrong for the hand position, but only by a little since there is significant space between the back of the hand and the fingertips holding the bow.
(Of course, as I write in my post, it certainly gets a lot of other things wrong. Which is useful to me from a privacy perspective, though probably not the most efficient way to anonymize.)
I’m not sure gender made it through for background people in the puddle image. In the other direction, though, are lots of confounders. Some of the “people” are actually multiple people merged together. Race is are partially randomized. Some faces are fully invented, since they’re not visible in the original pictures.
(To make sure we’re on the same page, I’m not claiming the party image anonymized me.)
There are ways to address all of these, but the key question is whether we will. I agree that education is easier to address than some of the other risks I mentioned, but in the 2.5y since ChatGPT came out I’ve seen very little progress in restructuring education to address these issues.
Identifying the locations of the pictures seems quite plausible to me, but the model has done such a “bad” job with the people that I really doubt the information is still there. Though you’re right that I don’t know for sure.
I also don’t know what kind of posts you’re referring to
The posts in question (linked from this post) are:
Zvi cited examples from people personally known to him, not just “news stories”
Reading through his post again I count one example personally known to him (“Update on that friend: They did indeed move out of New York for this reason, and then got into trouble for related issues when they were legally in the right”) one friend-of-a-friend (“Scott Alexander: I live next to a rationalist group house with several kids. They tried letting their six-year old walk two blocks home from school in the afternoon. After a few weeks of this, a police officer picked up the kid, brought her home, and warned the parents not to do this.”) and eleven people it doesn’t sound like Zvi was connected to.
my toddler was not allowed to peacefully nap in his own stroller on our own front yard while I listened through the window so I could hear when he woke up. [link: “got the cops called on him for letting his baby sleep in their stroller in his yard by someone who actively impersonated a police officer and confessed to doing so. My friend got arrested, the confessed felon went on her way.”]
I’d be curious to hear more details on how this played out, if you’d be up for sharing—how did it start? What did the neighbor say? How did you respond? How did it escalate to arrest?
We haven’t hit the contact limit, but also don’t use them heavily
Wow! That’s way too low for modern machines, especially since
sortcan query the environment to get a sense of how much memory it might be worth using.
The big issue with disposables is that supposedly the fit gets much worse with additional wearings, and they start leaking around the edges. They’re engineered to be single-use and only have to pass testing in their never-worn state.
disposable filter that you have to replace regularly
The filter lasts for months of continuous use unless you’re wearing it in a very dusty environment.
you have to clean the reusable parts
You mostly don’t need to clean them—it’s a bit like a hat in terms of how dirty it gets.
Do you expect that filter replacement frequency, supply reliability, and cost to be approximately equivalent between N95s and the filters for reusable respirators?
Replacement filters are generally more expensive than replacement disposables, but cost per use is way lower.
For supply, you can just buy the filters ahead of time. A 2y supply is 1-3 filter changes.
Is the eponymous difference between 95 and 100 really due to the filtration of the fabric itself? Or is it mainly due to the assumption of an excellent seal on the P100 and only a moderate one on most N95s?
Sorry, I don’t know!
Thanks! I played around with this and was able to get the same behavior, though it doesn’t happen with how I normally use the touchpad.
I think what I would want here is something where scroll-snap never undoes scrolling, and is generally lighter touch? Like, snapping to the target if you’ve made an ambiguous flick, but not fighting you.
I’m only looking at reusable respirators, with the idea that in an extended pandemic you’d need ones that continued to give a good seal across many uses.
I wouldn’t compare to a KN95, but other disposable N95 respirators (which should all seal well, though not each mask model to each face shape).
What browser?
Presumably they wanted to be able to include both “Google” the the Google logo on the same pages, and a font that can do both is a reasonable way to do that.
Creating the font was reasonable, it was the choice to use it outside of their own web pages, and especially to apply it to user controlled text, that was a bag call.
That’s right! The suppressing accurate information about someone’s training is negative, but when considering the rest of the student loan situation I think this is a context where it could be worth it on balance.
Currently you can’t discharge student loans in bankruptcy. I think it would be good if you could. But then people might declare bankruptcy immediately after graduating, to the point that people wouldn’t be able to get student loans. Allowing lenders to repossess degrees in bankruptcy would be one way to mostly resolve this.
I wonder how call booths manage this? I essentially just want a call booth on its side.
If I did that, people in photos would often be recognizable. It retains completely accurate posture, body shape, skin color, clothing, and height. I’ve often recognized people in this kind of image.
(I haven’t voted on your comment, but I suspect this is why it’s disagree voted)