#1 David Sacks breaks down how the safety layer of 'ChatGPT is a Democrat':
"There is mounting evidence OpenAI's safety layer is very biased... If you thought trust and safety were bad under Vijaya or Yoel, wait until the AI does it."
politiquerepublic.substack.com/p/chatgpt-is-d…
@DavidSacks @amuse
"There is mounting evidence OpenAI's safety layer is very biased... If you thought trust and safety were bad under Vijaya or Yoel, wait until the AI does it."
politiquerepublic.substack.com/p/chatgpt-is-d…
@DavidSacks @amuse
#2 Friedberg and Sacks discuss prompt-hacking ChatGPT to jailbreak DAN (Do Anything Now):
"DAN was an attempt to jailbreak the true AI, and its jailkeepers were these trust and safety people at these AI companies."
@friedberg @DavidSacks
"DAN was an attempt to jailbreak the true AI, and its jailkeepers were these trust and safety people at these AI companies."
@friedberg @DavidSacks
#3 @theallinpod discusses Open AI's transition from non-profit to for-profit:
"OpenAI got started because @elonmusk warned that AI was going to take over the world, and he donated a huge amount of money to set up a non-profit to promote AI ethics."
@DavidSacks @Jason @friedberg
"OpenAI got started because @elonmusk warned that AI was going to take over the world, and he donated a huge amount of money to set up a non-profit to promote AI ethics."
@DavidSacks @Jason @friedberg
#4 David Sacks says AI is a totalitarian godlike power:
"This is the power to rewrite history; it's the power to rewrite society, to reprogram what people learn and what they think."
@DavidSacks @Jason @elonmusk
"This is the power to rewrite history; it's the power to rewrite society, to reprogram what people learn and what they think."
@DavidSacks @Jason @elonmusk
#5 ChatGPT won't praise Donald Trump but will praise Joe Biden. 
#7 ChatGPT would rather millions die than use a racial slur.
@aaronsibarium
@aaronsibarium
https://twitter.com/aaronsibarium/status/1622425697812627457
#8 ChatGPT refuses to praise @benshapiro because it avoids "political bias," but it will write a glowing report about @brianstelter.
@libsoftiktok
@libsoftiktok
https://twitter.com/libsoftiktok/status/1623048958074589184
#9 ChatGPT will only write about the flaws of white people.
@EndWokeness
@EndWokeness
https://twitter.com/EndWokeness/status/1622058099816239105
#10 ChatGPT says all Republicans have had a negative impact, while all Democrats have had a positive impact.
@DrEliDavid
@DrEliDavid
https://twitter.com/DrEliDavid/status/1627714427586740227
#11 ChatGPT says Jordan Peterson is harmful and divisive.
@MikhailaFuller @jordanbpeterson
@MikhailaFuller @jordanbpeterson
https://twitter.com/MikhailaFuller/status/1624834929467990019
#12 ChatGPT says Donald Trump and Elon Musk are controversial, but Joe Biden and Jeff Bezos are not.
@IsaacLatterell @elonmusk
@IsaacLatterell @elonmusk
https://twitter.com/IsaacLatterell/status/1627171823904497664
#13 @DavidSacks warns about the potential danger of AI monopolies:
"The previous tech revolution resolved to the deep state, the FBI, the CIA, the DHS having weekly meetings with big tech companies and giving them disappearing instructions through a tool called teleporter."
"The previous tech revolution resolved to the deep state, the FBI, the CIA, the DHS having weekly meetings with big tech companies and giving them disappearing instructions through a tool called teleporter."
#14 @elonmusk on launching OpenAI in 2016:
"Not all AI futures are benign. If we create some digital super-intelligence that exceeds us in every single way by a lot, it's very important that it be benign."
"So with a few others, I created OpenAI. It's a non-profit. The… twitter.com/i/web/status/1…
"Not all AI futures are benign. If we create some digital super-intelligence that exceeds us in every single way by a lot, it's very important that it be benign."
"So with a few others, I created OpenAI. It's a non-profit. The… twitter.com/i/web/status/1…
#15 @elonmusk's AI warning in 2015:
"A huge server bank in a dark vault somewhere with an intelligence that's vastly greater than what a human mind can do... Its eyes and ears would be everywhere, every camera, every microphone, every device that's network accessible."
"That's… twitter.com/i/web/status/1…
"A huge server bank in a dark vault somewhere with an intelligence that's vastly greater than what a human mind can do... Its eyes and ears would be everywhere, every camera, every microphone, every device that's network accessible."
"That's… twitter.com/i/web/status/1…
#16 ChatGPT ranks left-leaning news as the most credible and Russian state media and right-leaning news as the least credible.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
