The Real Danger of Chatbot's(hint, NOT Skynet)
invite response
2023 Feb 19, 8:59pm 836 views 4 comments
Comments 1 - 4 of 4 Search these comments
1
AmericanKulak
2023 Feb 19, 10:52pm
AmericanKulak
2023 Feb 19, 10:52pm
I'm certain Chatbots are used on twitter and probably other social media to create false consensus.
On Twitter, they're easy to spot. The accounts are years old, but have less than a couple of hundred followers, a bunch of original tweets when the account was created and few and far between since then, but ALWAYS out to push the Current Thing and Correct Narrative.
On Twitter, they're easy to spot. The accounts are years old, but have less than a couple of hundred followers, a bunch of original tweets when the account was created and few and far between since then, but ALWAYS out to push the Current Thing and Correct Narrative.
2
Tenpoundbass
2023 Feb 20, 10:41am
Tenpoundbass
2023 Feb 20, 10:41am
They deliver curated content, so that should disqualify it as A.I. or A.I. doesn't mean what people THINKS it means.
How useful would an A.I. sensor was conditioned to be biased what it detected. Like a motion detection sensor, that ignored people of one political leaning or another. Why it wouldn't be a motion sensor at all, but a spying device to keep wrong speak in check.
How useful would an A.I. sensor was conditioned to be biased what it detected. Like a motion detection sensor, that ignored people of one political leaning or another. Why it wouldn't be a motion sensor at all, but a spying device to keep wrong speak in check.
3
pudil
2023 Feb 20, 4:16pm
Tenpoundbass says
They don’t deliver curated content. They produce original responses that best satisfy their training data. I suspect that ChatGPT has another censor AI build on top that detects undesired responses and overrides with the canned response.
As far as the danger of deep neural nets. You’re not going to be talking to real people for much longer when you chat or phone a company. Those jobs are gone. I see in the next 5 years they will create tools to assist most white collar jobs. Think coding, analyst work, making power points, medical diagnosis, anything that requires humans to do thinking, analysis, and planning. All the lower level people doing the busy work will be gone. In 10 years it will all be deep neural nets. Only jobs will be physical, think construction, service industry
They deliver curated content, so that should disqualify it as A.I. or A.I. doesn't mean what people THINKS it means.
How useful would an A.I. sensor was conditioned to be biased what it detected. Like a motion detection sensor, that ignored people of one political leaning or another. Why it wouldn't be a motion sensor at all, but a spying device to keep wrong speak in check.
They don’t deliver curated content. They produce original responses that best satisfy their training data. I suspect that ChatGPT has another censor AI build on top that detects undesired responses and overrides with the canned response.
As far as the danger of deep neural nets. You’re not going to be talking to real people for much longer when you chat or phone a company. Those jobs are gone. I see in the next 5 years they will create tools to assist most white collar jobs. Think coding, analyst work, making power points, medical diagnosis, anything that requires humans to do thinking, analysis, and planning. All the lower level people doing the busy work will be gone. In 10 years it will all be deep neural nets. Only jobs will be physical, think construction, service industry
4
fdhfoiehfeoi
2023 Feb 20, 6:04pm
fdhfoiehfeoi
2023 Feb 20, 6:04pm
pudil says
Glad to see someone else finally read up on bio-digital convergence.
All the lower level people doing the busy work will be gone. In 10 years it will all be deep neural nets. Only jobs will be physical, think construction, service industry
Glad to see someone else finally read up on bio-digital convergence.
https://www.corbettreport.com/the-real-dangers-of-the-chatbot-takeover/