4 ms·

Someone made a funny, novel toy, and they are seriously talking about "harm" it brings? We seriously need to reevaluate the nature of communication on the inte

Someone made a funny, novel toy, and they are seriously talking about "harm" it brings? We seriously need to reevaluate the nature of communication on the internet. Something about it just facilitates the social contagion of oversensitivity and hurt feelings.

Kye 6y ago
For example: oversensitive people like to dismiss concerns before they've taken a single second to hear or understand them. Actually listening to people and engaging with their words instead of the reader leaning into their biases and hurt feelings about the subject would improve communication on the internet. It wasn't a: >> "funny, novel toy" This was someone trying to build a business on guessing gender based on meaningless data. It did a very poor job even before considering the ethical issues. It couldn't even get it right with the name of the person behind it.
- TheAdamAndChe 6y ago
  > This was someone trying to build a business on guessing gender based on meaningless data This claim is baseless. We don't know their dataset or how it was used to train the model. > It did a very poor job By what metric? I didn't realize people have done stastical models to show that it was inaccurate. > It couldn't even get it right with the name of the person behind it. Yeah because it's machine learning. It's imperfect. You ran circles around what you want to say. You implied that I was being overly sensitive about how upset people were, but then you went on a tangent. To address the main point of conflict head-on: making assumptions of people based on a limited dataset. Some people think it's okay, some people think it isn't. I think in this situation, it is fine! They aren't determining who is being sent to prisons or something, they just made a mathematical model to make fuzzy predictions to help with things like advertising.

Kye 6y ago

For example: oversensitive people like to dismiss concerns before they've taken a single second to hear or understand them. Actually listening to people and engaging with their words instead of the reader leaning into their biases and hurt feelings about the subject would improve communication on the internet. It wasn't a: >> "funny, novel toy" This was someone trying to build a business on guessing gender based on meaningless data. It did a very poor job even before considering the ethical issues. It couldn't even get it right with the name of the person behind it.

TheAdamAndChe 6y ago
> This was someone trying to build a business on guessing gender based on meaningless data This claim is baseless. We don't know their dataset or how it was used to train the model. > It did a very poor job By what metric? I didn't realize people have done stastical models to show that it was inaccurate. > It couldn't even get it right with the name of the person behind it. Yeah because it's machine learning. It's imperfect. You ran circles around what you want to say. You implied that I was being overly sensitive about how upset people were, but then you went on a tangent. To address the main point of conflict head-on: making assumptions of people based on a limited dataset. Some people think it's okay, some people think it isn't. I think in this situation, it is fine! They aren't determining who is being sent to prisons or something, they just made a mathematical model to make fuzzy predictions to help with things like advertising.