Live data from Hacker News

Proof of care in the age of AI

jacobfilipp.com

31–40 of 119 posts

Re: Proof of care in the age of AI

#31
post #13

i think the point is good, hadwriting forces you to think more, even from typing the same.. BUT , i am unsure this would be proof you wrote it or AI genereated it , same with tatoos , AI can genereate picutres of said Tatoo...

Even though it doesn't prove the author wrote it himself, it at least proves he had to thoroughly read it before sharing.

Re: Proof of care in the age of AI

#33

While I appreciate the work put into this, I found it pretty hard to read because of the authors handwriting. I would never do this myself because I know that I have awful handwriting, and people would struggle to read it

Scroll to the end, text is intentionally copy-pasteable. Press ctrl+a and read it in Comic Sans :).

Re: Proof of care in the age of AI

#34
post #30

Am I the only one who thinks the ending is a non-sequitur? How is the hackneyed, "the kids are allright" [sic] related to the preceding content?

Yeah and if this were taken seriously, you would have Mechanical Turk style services where poors are paid pennies to hand-write submitted/generated slop, defeating the purpose.

Re: Proof of care in the age of AI

#35

Earlier quoted context omitted.

LLM slop is considered low value because it contains a low information/minute as well as a low effort/minute signal. You want to know that the reader put more effort in than you do, and that it is worth your time. The effort signal just points to a possible high information/minute return. When someone takes the laborious effort to provide a short paragraph on an insanely complex topic, precisely written without exces…

> I know they worked hard on it. That is still a valuable sign An infant scrawling the alphabet in its own excrement would have that "signal"...

And also has the hallmarks of "art." I suggest, however, if one were to actually implement this, that the 'excrement' should likely be a food-safe lookalike; maybe chocolate with granola and fruit hunks. Less likely to have trouble with child welfare authorities.

Re: Proof of care in the age of AI

#36
We need to normalize provenance tracking and sharing, similar to how git lets you separate the author from the committer.

I would go further and quantify how much of the message is AI in situations where humans edit it.

Re: Proof of care in the age of AI

#37

I love this problem and think it's super important. I've similarly noticed myself using a whiteboard to think critically for a while and then take a picture of the whiteboard as proof of deep thought, even if the next step is AI supplemented (a doc, a video, etc). I've also started noticing people annotating a whole doc "written by humans" to try to convey effort and care. That's fine for some things but do that too…

What do you mean "no way"? We've written long texts for as long as we've been writing as a species.

Re: Proof of care in the age of AI

#39

I love this problem and think it's super important. I've similarly noticed myself using a whiteboard to think critically for a while and then take a picture of the whiteboard as proof of deep thought, even if the next step is AI supplemented (a doc, a video, etc). I've also started noticing people annotating a whole doc "written by humans" to try to convey effort and care. That's fine for some things but do that too…

What do you mean "no way"? We've written long texts for as long as we've been writing as a species.

Not GP, but consider the view from the inside of a feckless pleb's skull. Effort is to be avoided, so its needless expenditure is unrelatable.

Re: Proof of care in the age of AI

#40
I'm unreasonably distracted by the fact that the illustrations of the tattoos are on the back of the "Subtraction" page.

That is consistent for both pages, but inconsistent with how they seem to be ordered within the text.

I guess the chapters were re-arranged post-script, with the "Storytellers" chapter inserted between them later?

Post reply on HN