3 ms·

The article is from 2015, so, not this article. Perhaps I have been lucky, but I have seen far more accusations of using AI to generate writing than I have see

The article is from 2015, so, not this article. Perhaps I have been lucky, but I have seen far more accusations of using AI to generate writing than I have seen of obviously generated writing.

nkozyra 4y ago
> The article is from 2015, so, not this article. I didn't mean this article specifically, but it can absolutely be done. Make the LLM the "reader" and have them retell the story like low-quality blogspam that still generates clicks and gets promoted by Google Now-type features.
NovaDudely 4y ago
Yep, it is the same thinking that runs heavily in the conspiracy theory crowd. Anything that they even vaguely suspect as evidence is proof of the theory. It is a shame that we are now in the position of having to think like this but that is the new world we are in now.
- rendall 4y ago
  I remember several years back there was what I can only describe as a moral panic about "machine intelligence-driven", "automated" YouTube videos for children. The idea was that there were myriad channels using AI to pump out low-quality content, and that much of this content was bizarre and completely inappropriate for children. I was interested in the process of how someone could use AI to create content and, as is probably obvious with the perspective of hindsight, I could find no real examples of channels exploiting automated, machine-learning production. This is the article that seems to have kicked off the trend but does not actually have any examples of automation: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39c471271d2 https://medium.com/@jamesbridle/something-is-wrong-on-the-in...
  - rendall 4y ago
    ... the connection there being that undue paranoia and accusations about AI generated content has been around for a little while already.

3 ms·

The article is from 2015, so, not this article. Perhaps I have been lucky, but I have seen far more accusations of using AI to generate writing than I have see

by rendall 4y ago

The article is from 2015, so, not this article. Perhaps I have been lucky, but I have seen far more accusations of using AI to generate writing than I have seen of obviously generated writing.

nkozyra 4y ago
> The article is from 2015, so, not this article. I didn't mean this article specifically, but it can absolutely be done. Make the LLM the "reader" and have them retell the story like low-quality blogspam that still generates clicks and gets promoted by Google Now-type features.
NovaDudely 4y ago
Yep, it is the same thinking that runs heavily in the conspiracy theory crowd. Anything that they even vaguely suspect as evidence is proof of the theory. It is a shame that we are now in the position of having to think like this but that is the new world we are in now.
- rendall 4y ago
  I remember several years back there was what I can only describe as a moral panic about "machine intelligence-driven", "automated" YouTube videos for children. The idea was that there were myriad channels using AI to pump out low-quality content, and that much of this content was bizarre and completely inappropriate for children. I was interested in the process of how someone could use AI to create content and, as is probably obvious with the perspective of hindsight, I could find no real examples of channels exploiting automated, machine-learning production. This is the article that seems to have kicked off the trend but does not actually have any examples of automation: https://medium.com/@jamesbridle/something-is-wrong-on-the-internet-c39c471271d2 https://medium.com/@jamesbridle/something-is-wrong-on-the-in...
  - rendall 4y ago
    ... the connection there being that undue paranoia and accusations about AI generated content has been around for a little while already.