26 ms·

Very dystopian. It works well as farce, if it were one.

by deburo 4y ago

29athrowaway 4y ago
Human intelligence is tainted with lower forms of intelligence from less evolved animals. The brain is structurally a monolith where structures responsible for advanced cognitive brain functions are built on top of older brain structures from millions of years prior. Primitive animals are less social, less collaborative and cannot collaborate in large groups. There are vestigial behaviors from them that arise once in a while in humans. It's time to evolve and press the mute button on lower ape behavior. Humans cannot do it reliably. Automatically detecting and flagging selfish individuals that cannot work in groups and actively sabotage their group by preventing the best ideas from prevailing is something worth trying. The reason everyone is seeing therapists is because some people are determined to ruin the work experience for others. We can do better.
- deburo 4y ago
  You’re proposing that an imperfect AI flags (ok) or fires (wow) employees based on potentially bad input. Same thing on muting mic or doing any kind of auto moderation. That’s going very far on an imperfect technology. I’d sooner bet on autonomous driving than people using such a moderator.
  - 29athrowaway 4y ago
    No, it's not. We already employ profanity filters in multiplayer games. We also have rating systems in games that allow peers to flag toxic behavior after a game session for review. These have helped limiting the harm caused by toxic players. Those are simple low tech solutions that could be added to videoconferencing applications. Did someone act toxic during a meeting? Allow participants to report a toxic interaction, with an AI generated transcript + recording. Then you can go one step further and build the future of HR playbook enforcement: "Meeting participant #1234, you are engaging in toxic behavior and are in direct violation of the HR playbook page 113 section 9: please stop disrupting the meeting and shut the hell up to avoid disciplinary action. You have 10 seconds to comply. 5, 4, 3, 2, 1...". /s But jokes aside... Today we have great sentiment analysis capabilities. There is a subset of the English language you should never use at work. We can also detect if someone is yelling. We can flag highly offensive terms. There are ways in which we can apply technology in a sensible way to improve the work experience, to limit the harm caused by the worst types of behaviors including but not limited to violence. We can also use technology to hold HR accountable to enforce the HR playbook.
    - deburo 4y ago
      I guess when we get to have AGI agents, they will start replacing HR, even though I doubt this will sit well with anyone. Again, it’s dystopian. Maybe in a Demolition Man world. But today? Wew
      - 29athrowaway 4y ago
        Some stuff can be classified as toxic and universally unacceptable with high confidence. That is the low hanging fruit we can begin with. A reputation system with user submitted information can also help.

Very dystopian. It works well as farce, if it were one.

by deburo 4y ago

Very dystopian. It works well as farce, if it were one.

Human intelligence is tainted with lower forms of intelligence from less evolved animals. The brain is structurally a monolith where structures responsible for advanced cognitive brain functions are built on top of older brain structures from millions of years prior. Primitive animals are less social, less collaborative and cannot collaborate in large groups. There are vestigial behaviors from them that arise once in a while in humans. It's time to evolve and press the mute button on lower ape behavior. Humans cannot do it reliably. Automatically detecting and flagging selfish individuals that cannot work in groups and actively sabotage their group by preventing the best ideas from prevailing is something worth trying. The reason everyone is seeing therapists is because some people are determined to ruin the work experience for others. We can do better.

deburo 4y ago
You’re proposing that an imperfect AI flags (ok) or fires (wow) employees based on potentially bad input. Same thing on muting mic or doing any kind of auto moderation. That’s going very far on an imperfect technology. I’d sooner bet on autonomous driving than people using such a moderator.
- 29athrowaway 4y ago
  No, it's not. We already employ profanity filters in multiplayer games. We also have rating systems in games that allow peers to flag toxic behavior after a game session for review. These have helped limiting the harm caused by toxic players. Those are simple low tech solutions that could be added to videoconferencing applications. Did someone act toxic during a meeting? Allow participants to report a toxic interaction, with an AI generated transcript + recording. Then you can go one step further and build the future of HR playbook enforcement: "Meeting participant #1234, you are engaging in toxic behavior and are in direct violation of the HR playbook page 113 section 9: please stop disrupting the meeting and shut the hell up to avoid disciplinary action. You have 10 seconds to comply. 5, 4, 3, 2, 1...". /s But jokes aside... Today we have great sentiment analysis capabilities. There is a subset of the English language you should never use at work. We can also detect if someone is yelling. We can flag highly offensive terms. There are ways in which we can apply technology in a sensible way to improve the work experience, to limit the harm caused by the worst types of behaviors including but not limited to violence. We can also use technology to hold HR accountable to enforce the HR playbook.
  - deburo 4y ago
    I guess when we get to have AGI agents, they will start replacing HR, even though I doubt this will sit well with anyone. Again, it’s dystopian. Maybe in a Demolition Man world. But today? Wew
    - 29athrowaway 4y ago
      Some stuff can be classified as toxic and universally unacceptable with high confidence. That is the low hanging fruit we can begin with. A reputation system with user submitted information can also help.