28 ms·

For those who aren't aware, this blog is infamous for sloppy work. https://www.lesswrong.com/posts/7iAABhWpcGeP5e6SB/it-s-probably-not-lithium https://www.lessw

For those who aren't aware, this blog is infamous for sloppy work. https://www.lesswrong.com/posts/7iAABhWpcGeP5e6SB/it-s-probably-not-lithium https://www.lesswrong.com/posts/7iAABhWpcGeP5e6SB/it-s-proba...

niplav 3y ago
This is a submission to a [contest](https://slimemoldtimemold.com/2023/01/01/mysterious-mysteries-of-unsolved-mystery-call-for-entries/ https://slimemoldtimemold.com/2023/01/01/mysterious-mysterie...) SMTM are running, so the author is very likely not the normal authors of SMTM.
- axiomaticdoubts 3y ago
  I still think it's important to note that the blog's work is shoddy. The fact that this blog post is #1 on HN means that multiple people will read SMTM's other blog posts, which contain several falsehoods that they've refused to fix. See, for example, this page https://manifold.markets/Natalia/how-many-of-these-falsemisleading-s https://manifold.markets/Natalia/how-many-of-these-falsemisl... for an incomplete list. I have been trying to get them to remove falsehoods from their blog for over one year, but it's impossible.
  - gruez 3y ago
    >The fact that this blog post is #1 on HN means that multiple people will read SMTM's other blog posts, which contain several falsehoods that they've refused to fix. That was prescient: https://news.ycombinator.com/item?id=36854850 https://news.ycombinator.com/item?id=36854850
qt31415926 3y ago
Thanks, didn't know. This article seems to have been written by a reader though so not the same authors of the sketchy lithium work
Zetice 3y ago
[flagged]
- camel-cdr 3y ago
  In the tread: > > Primarily he focuses not on developing a Strong AI (AGI), but rather focusing on safety issues that such a technology would pose. > That's absurd at worst, science fiction at best, akin to worrying about manned flight safety in the 1500's. It sure seems more warranted now.
  - Zetice 3y ago
    "Safety issues" are to him more like, "the AI will kill everyone to fill its objectives" and less like, "deepfakes are concerning." Though I'm sure he hopped on the "deepfakes are bad" bandwagon at some point to further his time in the spotlight.
    - cubefox 3y ago
      > "Safety issues" are to him more like, "the AI will kill everyone to fill its objectives" Yes, and he was way ahead of the curve here, since similar positions got a lot more mainstream in the past years and months. E.g. two of three AI Turing award winners (Hinton and Bengio) now say that superintelligent AI poses a serious extinction risk. OpenAI has also identified the alignment problem as a major issue. Even former AI skeptics like Douglas Hofstadter now broadly agree with this assessment. Yudkowsky's opinion was quite prescient.
      - Zetice 3y ago
        So in other words, people who've spent their entire lives thinking and talking about AI praise other people who also have spent their entire lives thinking and talking about AI. This is a bubble conversation that makes no sense to people outside of that bubble, and for good reason; it doesn't matter outside of that bubble.
        3y ago
notJim 3y ago
I follow Natália on twitter, and she's repeatedly tried to get them to address her criticisms, to no avail as far as I've seen. It's not a good look for a supposedly scientific blog, especially given that her criticisms are detailed and data-based. I don't think we should dismiss things based purely on the source, but caution seems warranted (as always, I suppose.)

niplav 3y ago

This is a submission to a [contest](https://slimemoldtimemold.com/2023/01/01/mysterious-mysteries-of-unsolved-mystery-call-for-entries/ https://slimemoldtimemold.com/2023/01/01/mysterious-mysterie...) SMTM are running, so the author is very likely not the normal authors of SMTM.

axiomaticdoubts 3y ago
I still think it's important to note that the blog's work is shoddy. The fact that this blog post is #1 on HN means that multiple people will read SMTM's other blog posts, which contain several falsehoods that they've refused to fix. See, for example, this page https://manifold.markets/Natalia/how-many-of-these-falsemisleading-s https://manifold.markets/Natalia/how-many-of-these-falsemisl... for an incomplete list. I have been trying to get them to remove falsehoods from their blog for over one year, but it's impossible.
- gruez 3y ago
  >The fact that this blog post is #1 on HN means that multiple people will read SMTM's other blog posts, which contain several falsehoods that they've refused to fix. That was prescient: https://news.ycombinator.com/item?id=36854850 https://news.ycombinator.com/item?id=36854850

qt31415926 3y ago

Thanks, didn't know. This article seems to have been written by a reader though so not the same authors of the sketchy lithium work

Zetice 3y ago

[flagged]

camel-cdr 3y ago
In the tread: > > Primarily he focuses not on developing a Strong AI (AGI), but rather focusing on safety issues that such a technology would pose. > That's absurd at worst, science fiction at best, akin to worrying about manned flight safety in the 1500's. It sure seems more warranted now.
- Zetice 3y ago
  "Safety issues" are to him more like, "the AI will kill everyone to fill its objectives" and less like, "deepfakes are concerning." Though I'm sure he hopped on the "deepfakes are bad" bandwagon at some point to further his time in the spotlight.
  - cubefox 3y ago
    > "Safety issues" are to him more like, "the AI will kill everyone to fill its objectives" Yes, and he was way ahead of the curve here, since similar positions got a lot more mainstream in the past years and months. E.g. two of three AI Turing award winners (Hinton and Bengio) now say that superintelligent AI poses a serious extinction risk. OpenAI has also identified the alignment problem as a major issue. Even former AI skeptics like Douglas Hofstadter now broadly agree with this assessment. Yudkowsky's opinion was quite prescient.
    - Zetice 3y ago
      So in other words, people who've spent their entire lives thinking and talking about AI praise other people who also have spent their entire lives thinking and talking about AI. This is a bubble conversation that makes no sense to people outside of that bubble, and for good reason; it doesn't matter outside of that bubble.
      - 3y ago

notJim 3y ago

I follow Natália on twitter, and she's repeatedly tried to get them to address her criticisms, to no avail as far as I've seen. It's not a good look for a supposedly scientific blog, especially given that her criticisms are detailed and data-based. I don't think we should dismiss things based purely on the source, but caution seems warranted (as always, I suppose.)