2 ms·

> To some extent, they may be right that open sourcing AGI would lead to too much danger. That's clearly self-serving claptrap. It's a leveraging of a false de

> To some extent, they may be right that open sourcing AGI would lead to too much danger. That's clearly self-serving claptrap. It's a leveraging of a false depiction of what AGI will look like (no ones really knows, but it's going to be scary and out of control!) with so much gatekeeping and subsequent cash they can hardly stop salivating. No strong AI (there is no evidence AGI is even possible) is not going to be a menace. It's software FFS. Humans are and will be a menace though and logically the only way to protect ourselves from bad people (and corporations) with strong AI is to make strong AI available to everyone. Computers are pretty powerful (and evil) right now but we haven't banned them yet.

afarviral 3y ago
That makes little intuitive sense to me. Help me understand why increasing the number of entities which possess a potential-weapon is beneficial for humanity? If the US had developed a nuclear armament and no other country had would that truly have been worse? What if Russia had beat the world to it first? Maybe I'll get there on my own if I keep following this reasoning. However there is nothing clear cut about it, my strongest instincts are only heuristics I've absorbed from somewhere. What we probably want with any sufficiently destructive potential-weapon are the most responsible actors to share their research while stimulating research in the field with a strong focus on safety and safeguarding. I see some evidence of that.
- cortesoft 3y ago
  Lots of people disagree on whether it is true or not, but basically the idea is mutually assured destruction https://en.wikipedia.org/wiki/Mutual_assured_destruction https://en.wikipedia.org/wiki/Mutual_assured_destruction
  - afarviral 3y ago
    I sense that with AGI all the outcomes will be a little less assured, since it is general-purpose. We won't know what hit until it's over. Was it a pandemic? Was it automated-religion? Nuclear weapons seem particularly suited to MAD, but not AGI.
- Jensson 3y ago
  > If the US had developed a nuclear armament and no other country had would that truly have been worse? Yes, do you think it is a coincidence that nuclear weapons stopped being used in wars as soon as more than one power had them? People would clamor for nukes to be used to save their young soldiers lives if they didn't have to fear nuclear retaliation, you would see strong political pushes for nuclear usage in everyone of USA's wars.
  - afarviral 3y ago
    Hmm, indeed
mitthrowaway2 3y ago
> there is no evidence AGI is even possible Reading this is like hearing "there is no evidence that heavier-than-air flight is even possible" being spoken, by a bird. If 8 billion naturally-occuring intelligences don't qualify as evidence that AGI is possible, then is there anything that can qualify as evidence of anything else being possible?
- stonogo 3y ago
  we also cannot build most birds
insane_dreamer 3y ago
> No strong AI (there is no evidence AGI is even possible) is not going to be a menace. It's software FFS. have you even watched Terminator? ;)
esafak 3y ago
Networking is a thing, so the software can remotely control hardware.

afarviral 3y ago

That makes little intuitive sense to me. Help me understand why increasing the number of entities which possess a potential-weapon is beneficial for humanity? If the US had developed a nuclear armament and no other country had would that truly have been worse? What if Russia had beat the world to it first? Maybe I'll get there on my own if I keep following this reasoning. However there is nothing clear cut about it, my strongest instincts are only heuristics I've absorbed from somewhere. What we probably want with any sufficiently destructive potential-weapon are the most responsible actors to share their research while stimulating research in the field with a strong focus on safety and safeguarding. I see some evidence of that.

cortesoft 3y ago
Lots of people disagree on whether it is true or not, but basically the idea is mutually assured destruction https://en.wikipedia.org/wiki/Mutual_assured_destruction https://en.wikipedia.org/wiki/Mutual_assured_destruction
- afarviral 3y ago
  I sense that with AGI all the outcomes will be a little less assured, since it is general-purpose. We won't know what hit until it's over. Was it a pandemic? Was it automated-religion? Nuclear weapons seem particularly suited to MAD, but not AGI.
Jensson 3y ago
> If the US had developed a nuclear armament and no other country had would that truly have been worse? Yes, do you think it is a coincidence that nuclear weapons stopped being used in wars as soon as more than one power had them? People would clamor for nukes to be used to save their young soldiers lives if they didn't have to fear nuclear retaliation, you would see strong political pushes for nuclear usage in everyone of USA's wars.
- afarviral 3y ago
  Hmm, indeed

mitthrowaway2 3y ago

> there is no evidence AGI is even possible Reading this is like hearing "there is no evidence that heavier-than-air flight is even possible" being spoken, by a bird. If 8 billion naturally-occuring intelligences don't qualify as evidence that AGI is possible, then is there anything that can qualify as evidence of anything else being possible?

stonogo 3y ago
we also cannot build most birds

insane_dreamer 3y ago

> No strong AI (there is no evidence AGI is even possible) is not going to be a menace. It's software FFS. have you even watched Terminator? ;)

esafak 3y ago

Networking is a thing, so the software can remotely control hardware.