4 ms·

This is totally false. Intelligence is strongly genetic. Traits are very often genetically additive. Although we won't be certain until do a huge survey of gene

This is totally false. Intelligence is strongly genetic. Traits are very often genetically additive. Although we won't be certain until do a huge survey of genetics, all the author is saying is that it might be possible. And if it is, then we do have the technology to "engineer" a child with the genes we want.

dreamweapon 12y ago
All the author is saying is that it might be possible. No; he doesn't merely say that it "might be possible." He says that it will happen. In fact he's so sure it will happen that it's basically inevitable. So sure that he takes great pains to say so, in bold letters (hundreds of pixels wide): "Super-Intelligent Humans Are Coming"
- Houshalter 12y ago
  That's just the title, and that claim is much weaker than the "1,000 IQ" claim I was referring to. The author is very clear this isn't a certainty.
  - dreamweapon 12y ago
    Which is to say: he makes an outlandish claim in the title, in order to get us to read what he has to say; then meekly backpeddles away from that claim in the body of the article.
    - Houshalter 12y ago
      Generally the editor writes the title. Regardless, the title isn't very controversial; "superintelligence" isn't well defined and it's certainly possible we could have humans with greater than genius IQ. It was the "1,000 IQ" claim that isn't certain, since it's just a simple extrapolation we don't really know how all of those genes would interact.
dreamweapon 12y ago
Intelligence is strongly genetic. Is it? Define. Traits are very often genetically additive. Therefore all traits are additive? And linearly so, in arbitrary combinations?
- gwern 12y ago
  > Therefore all traits are additive? And linearly so, in arbitrary combinations? No, not all variation is additive, but things like twin studies and GCTAs let you estimate the additive part of the heritability. (I can't tell whether you're completely ignorant of this topic & wildly overestimating the cogency of your comments, or just being a Socratic dick, so it would be helpful if you could clarify.)
  - dreamweapon 12y ago
    You can call me names, if you like. The guy above me seemed not to understand the implications of what he was saying, and I was doing my best to gently help him understand this fact.
    - gwern 12y ago
      I think he understood better than you did, given that your response did not address anything I said.
      - dreamweapon 12y ago
        I don't see a need to, given the way you started in.

dreamweapon 12y ago

All the author is saying is that it might be possible. No; he doesn't merely say that it "might be possible." He says that it will happen. In fact he's so sure it will happen that it's basically inevitable. So sure that he takes great pains to say so, in bold letters (hundreds of pixels wide): "Super-Intelligent Humans Are Coming"

Houshalter 12y ago
That's just the title, and that claim is much weaker than the "1,000 IQ" claim I was referring to. The author is very clear this isn't a certainty.
- dreamweapon 12y ago
  Which is to say: he makes an outlandish claim in the title, in order to get us to read what he has to say; then meekly backpeddles away from that claim in the body of the article.
  - Houshalter 12y ago
    Generally the editor writes the title. Regardless, the title isn't very controversial; "superintelligence" isn't well defined and it's certainly possible we could have humans with greater than genius IQ. It was the "1,000 IQ" claim that isn't certain, since it's just a simple extrapolation we don't really know how all of those genes would interact.

dreamweapon 12y ago

Intelligence is strongly genetic. Is it? Define. Traits are very often genetically additive. Therefore all traits are additive? And linearly so, in arbitrary combinations?

gwern 12y ago
> Therefore all traits are additive? And linearly so, in arbitrary combinations? No, not all variation is additive, but things like twin studies and GCTAs let you estimate the additive part of the heritability. (I can't tell whether you're completely ignorant of this topic & wildly overestimating the cogency of your comments, or just being a Socratic dick, so it would be helpful if you could clarify.)
- dreamweapon 12y ago
  You can call me names, if you like. The guy above me seemed not to understand the implications of what he was saying, and I was doing my best to gently help him understand this fact.
  - gwern 12y ago
    I think he understood better than you did, given that your response did not address anything I said.
    - dreamweapon 12y ago
      I don't see a need to, given the way you started in.