4 ms·

I have never trolled on the internet… but I almost gave our future AI overlord my phone number just so I could have him compose a response to your post. I gues

by gregw2 4y ago

I have never trolled on the internet… but I almost gave our future AI overlord my phone number just so I could have him compose a response to your post. I guess I am more human than I thought.

rpigab 4y ago
That was my only response to this post, seriously, he's begging to be trolled...
Eisenstein 4y ago
I am glad I am not the only one who hasn't gotten past the phone number barrier.
- wila 4y ago
  Same, I am certainly NOT going to give a phone number for this free service.
gregw2 4y ago
Now if ChatGPT had written that response I would be impressed!
vonunov 4y ago
FYI, it wouldn't take my Gvoice number (natch) but it does accept Flowroute DID numbers. This was what finally made me get around to figuring out how to accept SMS over callback/webhook/whatever it is (instead of paying however much FlowSMS is charging now). I can share nginx config if anyone is trying to do the same
- death916 4y ago
  I'd be interested in seeing that actually.(the config)
  - vonunov 4y ago
    Flowroute: DIDs -> select number -> Choose a DID Action -> Set callback URL -> Apply action -> check SMS Callback, SMS DLR callback (?) -> enter URL complying with https://developer.flowroute.com/api/numbers/v2.0/set-a-messaging-callback-url-for-phone-number/ https://developer.flowroute.com/api/numbers/v2.0/set-a-messa... (I had to have it exactly as specified with /numbers/ starting from the webroot and not in a subdirectory), see also https://developer.flowroute.com/api/messages/v2.1/receive-an-sms/ https://developer.flowroute.com/api/messages/v2.1/receive-an... Using SMS webhook v2.0 in Flowroute Preferences -> API Control In Arch Linux I installed "community/nginx-mod-echo 0.62-5" and restarted nginx nginx.conf: load_module "/usr/lib/nginx/modules/ngx_http_echo_module.so"; http { log_format postdata $request_body; } sites-available/example.com.conf: server { location /numbers/1NPANXXxxxx/relationships/sms_dlr_callback { access_log /var/log/nginx/example.com/postdata.log postdata; echo_read_request_body; } } Reload nginx, then I just did `tail -f postdata.log` and in near realtime received output like "{\x22body\x22: \x22Your OpenAI API verification code is: 438301\x22, \x22to\x22: \x221903866xxxx\x22, \x22from\x22: \x2218559090331\x22, \x22id\x22: \x22mdr2-9e551f88de31449096782f9a9f32080f\x22}"
    - itsyaboi 4y ago
      Thank you for sharing this!
    - death916 4y ago
      Thanks
Keyframe 4y ago
Here's what actual response to his post, fully cited, from ChatGPT would be: It is not helpful for Person B to simply paste the response of a language model without providing any additional context or explanation. By doing so, Person B is not contributing to the conversation and is not providing any value to the discussion. Additionally, this behavior can be confusing and misleading for other readers who may not be aware that the response is from a language model and not from a real person. It is important for all participants in a conversation to contribute their own thoughts and ideas in order to have a productive and meaningful discussion. offer a suggestion One suggestion for Person B could be to provide their own thoughts and ideas in response to the question, rather than simply pasting the response of a language model. Person B could also provide some context or explanation for the response of the language model, to help other readers understand the relevance of the response and how it relates to the original question. Additionally, Person B could engage with the other participants in the conversation and ask follow-up questions or provide additional insights to further the discussion. By actively participating in the conversation and providing value through their own contributions, Person B can help make the conversation more productive and enjoyable for everyone involved. how would we enforce that? Enforcing the suggestion for Person B to provide their own thoughts and ideas in a conversation would be difficult to do in practice. One potential way to encourage this behavior would be for the community to establish norms and expectations around participation in discussions, and to provide positive reinforcement for those who contribute meaningfully to the conversation. For example, other participants in the conversation could provide feedback and recognition to those who contribute their own ideas and insights, and could avoid engaging with those who simply paste responses from language models without providing any additional value. Over time, this could help to establish a culture of active and meaningful participation in discussions, and could discourage the use of language models as a substitute for real human engagement. effectively, ChatGPT is advising us to use voting.

by gregw2 4y ago

I have never trolled on the internet… but I almost gave our future AI overlord my phone number just so I could have him compose a response to your post. I guess I am more human than I thought.

rpigab 4y ago

That was my only response to this post, seriously, he's begging to be trolled...

Eisenstein 4y ago

I am glad I am not the only one who hasn't gotten past the phone number barrier.

wila 4y ago
Same, I am certainly NOT going to give a phone number for this free service.

gregw2 4y ago

Now if ChatGPT had written that response I would be impressed!

vonunov 4y ago

FYI, it wouldn't take my Gvoice number (natch) but it does accept Flowroute DID numbers. This was what finally made me get around to figuring out how to accept SMS over callback/webhook/whatever it is (instead of paying however much FlowSMS is charging now). I can share nginx config if anyone is trying to do the same

death916 4y ago
I'd be interested in seeing that actually.(the config)
- vonunov 4y ago
  Flowroute: DIDs -> select number -> Choose a DID Action -> Set callback URL -> Apply action -> check SMS Callback, SMS DLR callback (?) -> enter URL complying with https://developer.flowroute.com/api/numbers/v2.0/set-a-messaging-callback-url-for-phone-number/ https://developer.flowroute.com/api/numbers/v2.0/set-a-messa... (I had to have it exactly as specified with /numbers/ starting from the webroot and not in a subdirectory), see also https://developer.flowroute.com/api/messages/v2.1/receive-an-sms/ https://developer.flowroute.com/api/messages/v2.1/receive-an... Using SMS webhook v2.0 in Flowroute Preferences -> API Control In Arch Linux I installed "community/nginx-mod-echo 0.62-5" and restarted nginx nginx.conf: load_module "/usr/lib/nginx/modules/ngx_http_echo_module.so"; http { log_format postdata $request_body; } sites-available/example.com.conf: server { location /numbers/1NPANXXxxxx/relationships/sms_dlr_callback { access_log /var/log/nginx/example.com/postdata.log postdata; echo_read_request_body; } } Reload nginx, then I just did `tail -f postdata.log` and in near realtime received output like "{\x22body\x22: \x22Your OpenAI API verification code is: 438301\x22, \x22to\x22: \x221903866xxxx\x22, \x22from\x22: \x2218559090331\x22, \x22id\x22: \x22mdr2-9e551f88de31449096782f9a9f32080f\x22}"
  - itsyaboi 4y ago
    Thank you for sharing this!
  - death916 4y ago
    Thanks

Keyframe 4y ago

Here's what actual response to his post, fully cited, from ChatGPT would be: It is not helpful for Person B to simply paste the response of a language model without providing any additional context or explanation. By doing so, Person B is not contributing to the conversation and is not providing any value to the discussion. Additionally, this behavior can be confusing and misleading for other readers who may not be aware that the response is from a language model and not from a real person. It is important for all participants in a conversation to contribute their own thoughts and ideas in order to have a productive and meaningful discussion. offer a suggestion One suggestion for Person B could be to provide their own thoughts and ideas in response to the question, rather than simply pasting the response of a language model. Person B could also provide some context or explanation for the response of the language model, to help other readers understand the relevance of the response and how it relates to the original question. Additionally, Person B could engage with the other participants in the conversation and ask follow-up questions or provide additional insights to further the discussion. By actively participating in the conversation and providing value through their own contributions, Person B can help make the conversation more productive and enjoyable for everyone involved. how would we enforce that? Enforcing the suggestion for Person B to provide their own thoughts and ideas in a conversation would be difficult to do in practice. One potential way to encourage this behavior would be for the community to establish norms and expectations around participation in discussions, and to provide positive reinforcement for those who contribute meaningfully to the conversation. For example, other participants in the conversation could provide feedback and recognition to those who contribute their own ideas and insights, and could avoid engaging with those who simply paste responses from language models without providing any additional value. Over time, this could help to establish a culture of active and meaningful participation in discussions, and could discourage the use of language models as a substitute for real human engagement. effectively, ChatGPT is advising us to use voting.