3 ms·

It looks like those who love LLMs (which I do) are forcing themselves to come up with narratives that mitigate the proposed negative effects. I have a differen

It looks like those who love LLMs (which I do) are forcing themselves to come up with narratives that mitigate the proposed negative effects. I have a different suggestion. It’s likely it is real and painful and sad. Transitions are hard, disruptive and can often cause very real and negative problems, but they are unavoidable and historically technology brings humanity to a better place in the long run. That reminder about printing press effects in the Ng and LeCunn video was good example. A lot of crazy bad stuff happened that likely the printing press triggered but on the other side was a Renaissance. I think the narrative in this post is likely accurate. It’s not great on an individual human level and that is hard but the last possible thing we want is governments stepping in to control this situation. I don’t want to be in a global version of the Ottoman’s restricting the printing press.

ChatGTP 4y ago
I keep hearing this but I’ve never heard a coherent plan about how we’d go forwards with AI, like what is the actual strategy and do we actually think this will end. By all accounts it’s probably not going to be ideal?
- alchemist1e9 4y ago
  Perhaps you need to let go of the idea that humanity actually has direct control of how things unfold. Instead focus on your own situation. The idea there is a possibility of an “actual strategy” is I think the real flaw. Government can try to regulate products but given the 8 billion people and numerous advanced economies on the planet any locations that attempts to restrict products that use LLMs for proposed harm instead of realized harm will probably put themselves at a disadvantage.
  - blibble 4y ago
    > Government can try to regulate products but given the 8 billion people and numerous advanced economies on the planet any locations that attempts to restrict products that use LLMs for proposed harm instead of realized harm will probably put themselves at a disadvantage. we put ourselves at a competitive disadvantage with child labour laws too data privacy law, banning slavery, emission laws, etc. to have a fair and just society some things have to be banned or highly regulated, regardless of the economic cost AI is no different
    - ImHereToVote 4y ago
      >we put ourselves at a competitive disadvantage with child labour laws too You are in luck. We are about to do an Atlas Shrugged on this in many states. We now have kids working in meat packing plants. https://www.theguardian.com/us-news/2023/mar/08/arkansas-bill-child-labor-protections https://www.theguardian.com/us-news/2023/mar/08/arkansas-bil... https://www.theguardian.com/law/2023/feb/17/underage-child-labor-working-slaughterhouse-investigation https://www.theguardian.com/law/2023/feb/17/underage-child-l...
gls2ro 4y ago
> Transitions are hard, disruptive and can often cause very real and negative problems, but they are unavoidable and historically technology brings humanity to a better place in the long run. First: Why does it have to be that way? Is it maybe because that the top does always impose the costs of disruption to the bottom? Second: Now that we have AI that seems capable to do this kind of disruption can we kindly ask AI to do this transition smoother?

ChatGTP 4y ago

I keep hearing this but I’ve never heard a coherent plan about how we’d go forwards with AI, like what is the actual strategy and do we actually think this will end. By all accounts it’s probably not going to be ideal?

alchemist1e9 4y ago
Perhaps you need to let go of the idea that humanity actually has direct control of how things unfold. Instead focus on your own situation. The idea there is a possibility of an “actual strategy” is I think the real flaw. Government can try to regulate products but given the 8 billion people and numerous advanced economies on the planet any locations that attempts to restrict products that use LLMs for proposed harm instead of realized harm will probably put themselves at a disadvantage.
- blibble 4y ago
  > Government can try to regulate products but given the 8 billion people and numerous advanced economies on the planet any locations that attempts to restrict products that use LLMs for proposed harm instead of realized harm will probably put themselves at a disadvantage. we put ourselves at a competitive disadvantage with child labour laws too data privacy law, banning slavery, emission laws, etc. to have a fair and just society some things have to be banned or highly regulated, regardless of the economic cost AI is no different
  - ImHereToVote 4y ago
    >we put ourselves at a competitive disadvantage with child labour laws too You are in luck. We are about to do an Atlas Shrugged on this in many states. We now have kids working in meat packing plants. https://www.theguardian.com/us-news/2023/mar/08/arkansas-bill-child-labor-protections https://www.theguardian.com/us-news/2023/mar/08/arkansas-bil... https://www.theguardian.com/law/2023/feb/17/underage-child-labor-working-slaughterhouse-investigation https://www.theguardian.com/law/2023/feb/17/underage-child-l...

gls2ro 4y ago

> Transitions are hard, disruptive and can often cause very real and negative problems, but they are unavoidable and historically technology brings humanity to a better place in the long run. First: Why does it have to be that way? Is it maybe because that the top does always impose the costs of disruption to the bottom? Second: Now that we have AI that seems capable to do this kind of disruption can we kindly ask AI to do this transition smoother?