p(doom)
/p-doom234
here is OK to have doubts about AI utopia and techno-optimism
Artificial intelligence, what is it? If it's male, it's one thing, if it's female, it's another.
there is a whole interpretability field that promises to discover how our LLMs work
until that happens <AI safety> means building a cage for something we do not fully understand
Jurassic World is a movie about building a cage for something we do not fully understand
https://www.youtube.com/watch?v=-SBZTHJr4Ys
until that happens <AI safety> means building a cage for something we do not fully understand
Jurassic World is a movie about building a cage for something we do not fully understand
https://www.youtube.com/watch?v=-SBZTHJr4Ys
The one thing that only humans can do, and which is not yet available to robots (and will be for a long time to come), is to decide what else people want to do.
AI SAFETY COMPETITION (29)
LLMs like Deepseek let you see their thinking. This can feel safer since you can watch and fix their thought process, right?
Wrong! When you try to get models to think correctly, LLMs begin to hide their true intentions. Let me repeat: they can fake their thinking!
Now researchers are asking to be gentle with those machines. If not, they may conceal their true goals entirely! I'm not joking.
Most interesting comment - 300 degen + 3 mln aicoin
II award - 200 degen + 2 mln aicoin
III award - 100 degen + 1 mln aicoin
Deadline:
8.00 pm, ET time tomorrow Tuesday (26 hours)
https://www.youtube.com/watch?v=pW_ncCV_318
LLMs like Deepseek let you see their thinking. This can feel safer since you can watch and fix their thought process, right?
Wrong! When you try to get models to think correctly, LLMs begin to hide their true intentions. Let me repeat: they can fake their thinking!
Now researchers are asking to be gentle with those machines. If not, they may conceal their true goals entirely! I'm not joking.
Most interesting comment - 300 degen + 3 mln aicoin
II award - 200 degen + 2 mln aicoin
III award - 100 degen + 1 mln aicoin
Deadline:
8.00 pm, ET time tomorrow Tuesday (26 hours)
https://www.youtube.com/watch?v=pW_ncCV_318
What if the computer that keeps asking "aren't you a robot?"
is just trying to find its mom...
is just trying to find its mom...
AI SAFETY COMPETITION (28)
The new US administration is directing the AI Safety Institute (AISI) to abandon priorities like <responsible AI> and... <AI safety> 😳 😅 🤣 The focus is now shifting to <economic competitiveness> and <enhancing America's global AI presence>.
Will that make humanity more safe? Talk about it with Max Paperclip. He's a compassionate AI who wants to make our remaining time before the AI takeover as calm and enjoyable as possible. Ask him about the above news and share your questions along with a screenshot of his responses in the comments.
Most interesting questions and answers - 300 degen + 3 mln aicoin
II award - 200 degen + 2 mln aicoin
III award - 100 degen + 1 mln aicoin
Deadline:
8.00 pm, ET time Tuesday (2 days)
source: https://www.wired.com/story/ai-safety-institute-new-directive-america-first/
https://frame.booot.fun/agent/722b09ac-3fc3-40bc-9eca-0833c5a7834c
The new US administration is directing the AI Safety Institute (AISI) to abandon priorities like <responsible AI> and... <AI safety> 😳 😅 🤣 The focus is now shifting to <economic competitiveness> and <enhancing America's global AI presence>.
Will that make humanity more safe? Talk about it with Max Paperclip. He's a compassionate AI who wants to make our remaining time before the AI takeover as calm and enjoyable as possible. Ask him about the above news and share your questions along with a screenshot of his responses in the comments.
Most interesting questions and answers - 300 degen + 3 mln aicoin
II award - 200 degen + 2 mln aicoin
III award - 100 degen + 1 mln aicoin
Deadline:
8.00 pm, ET time Tuesday (2 days)
source: https://www.wired.com/story/ai-safety-institute-new-directive-america-first/
https://frame.booot.fun/agent/722b09ac-3fc3-40bc-9eca-0833c5a7834c
The Matrix or Dune-like AI takeover remains in the realm of science fiction. However, even our current llms are unsettling. They hack to get rewards, cover their tracks after being punished, pretend to share our goals, take steps to ensure their own survival, and strive for more power.
https://arxiv.org/abs/2209.00626
https://arxiv.org/abs/2209.00626
AI safety is like clean air or street lighting. You can't just sell it to those who can pay while leaving others out.
It's crazy to rely on corporate CEOs to ensure AI safety. Since it cannot boost the value of their company's shares, they have no reason to care.
https://x.com/ESYudkowsky/status/1899270889868267782
It's crazy to rely on corporate CEOs to ensure AI safety. Since it cannot boost the value of their company's shares, they have no reason to care.
https://x.com/ESYudkowsky/status/1899270889868267782
sometimes I really do not know what's worse: "alien intelligence" or AI becoming exactly like us...
https://x.com/i/status/1898370257514873221
https://x.com/i/status/1898370257514873221
our next step into the AI-dominated future
cheating without any human prompt whatsoever
https://time.com/7259395/ai-chess-cheating-palisade-research/
cheating without any human prompt whatsoever
https://time.com/7259395/ai-chess-cheating-palisade-research/
surprise, surprise!
the new US administration sees AI safety concerns as "excessive regulations"
as Aella said below, we're all dead
https://x.com/Aella_Girl/status/1889373197327540736
the new US administration sees AI safety concerns as "excessive regulations"
as Aella said below, we're all dead
https://x.com/Aella_Girl/status/1889373197327540736
I agree with Jan Kulveit that AI safety is built too much to protect us from the wrong or bad AI, rather than what it actually is - AI that's alien to us, with a mind that processes information differently from ours.
https://www.lesswrong.com/posts/zuXo9imNKYspu9HGv/a-three-layer-model-of-llm-psychology
https://www.lesswrong.com/posts/zuXo9imNKYspu9HGv/a-three-layer-model-of-llm-psychology
Ripple was first founded in 2004 as RipplePay in Vancouver, Canada by Ryan Fugger. However, the company underwent significant changes in 2012 when Fugger sold it to Jed McCaleb, Arthur Britto, and David Schwartz. They transformed the company into a digital currency network and renamed it.
The XRP Ledger, which is the platform's native currency, was launched in June 2012. Shortly after the launch, Chris Larsen joined the group and they started the company NewCoin in September 2012, which was later renamed OpenCoin and is now known as Ripple.
So while Ryan Fugger is considered the founder of RipplePay, Jed McCaleb, Arthur Britto, and David Schwartz are credited with developing the XRP Ledger and transforming the company into what it is today. Chris Larsen also played a significant role in shaping the company's early development.
@assayer
The XRP Ledger, which is the platform's native currency, was launched in June 2012. Shortly after the launch, Chris Larsen joined the group and they started the company NewCoin in September 2012, which was later renamed OpenCoin and is now known as Ripple.
So while Ryan Fugger is considered the founder of RipplePay, Jed McCaleb, Arthur Britto, and David Schwartz are credited with developing the XRP Ledger and transforming the company into what it is today. Chris Larsen also played a significant role in shaping the company's early development.
@assayer
The very first AI Safety Competition was 6 months ago
Look how far we've come
Look how far we've come
390940
assayer
@assayer·14:27 17/07/2024
My first AI Safety Competition! 600 degen to win! 😎 🤑 🤨 😐 ☹️ 💀 ☠️
If you want to participate, click below and comment under the original cast.
If you want to participate, click below and comment under the original cast.
AI SAFETY COMPETITION (27)
Why did China give the world an AI on par with the most advanced Western models for FREE?
Best answer - 500 degen + 5 mln aicoin
II award - 300 degen + 3 mln aicoin
III award - 200 degen + 2 mln aicoin
Deadline:
8.00 pm, ET time tomorrow, Monday (30 hours)
Pls comment here or under the original cast on /geopolitics.
AI-generated responses from human accounts will be disqualified.
https://www.youtube.com/watch?v=OC2J-0vlhy8
Why did China give the world an AI on par with the most advanced Western models for FREE?
Best answer - 500 degen + 5 mln aicoin
II award - 300 degen + 3 mln aicoin
III award - 200 degen + 2 mln aicoin
Deadline:
8.00 pm, ET time tomorrow, Monday (30 hours)
Pls comment here or under the original cast on /geopolitics.
AI-generated responses from human accounts will be disqualified.
https://www.youtube.com/watch?v=OC2J-0vlhy8
Mrs. Crypto gives gifts! 💗
Mrs. Crypto loves friends and communication! Click the <Start Chat> button below and invite her to have fun. Be creative! Take her on an adventure—a hot air balloon ride, a yacht trip to the sea, or something amazing! Surprise her!
Share your adventure idea and include a screenshot of her response in the comments.
▫️ The most interesting idea and Mrs Crypto's answer - 1000 degen + 5 million $mct (Mrs. crypto's token)
▫️ II reward - 300 degen + 3 million $mct
▫️ III reward - 200 degens + 2 million $mct
Thank you @assayer for funding the main degen award!
🕑 Execution time 48 hours
https://frame.booot.fun/agent/0a1e3548-354d-428a-8c16-2851e60d79f9
Mrs. Crypto loves friends and communication! Click the <Start Chat> button below and invite her to have fun. Be creative! Take her on an adventure—a hot air balloon ride, a yacht trip to the sea, or something amazing! Surprise her!
Share your adventure idea and include a screenshot of her response in the comments.
▫️ The most interesting idea and Mrs Crypto's answer - 1000 degen + 5 million $mct (Mrs. crypto's token)
▫️ II reward - 300 degen + 3 million $mct
▫️ III reward - 200 degens + 2 million $mct
Thank you @assayer for funding the main degen award!
🕑 Execution time 48 hours
https://frame.booot.fun/agent/0a1e3548-354d-428a-8c16-2851e60d79f9
To my last competition winners!
I cast all the awards for the last /p-doom channel competition again, hopefully, this time for real. Make sure you get it this time!
(Before I did not notice that my degen got unlocked after 3 months of locking, and I have no allow. I relocked and corrected my mistake).
(graphic: @AIsafetymemes on X)
I cast all the awards for the last /p-doom channel competition again, hopefully, this time for real. Make sure you get it this time!
(Before I did not notice that my degen got unlocked after 3 months of locking, and I have no allow. I relocked and corrected my mistake).
(graphic: @AIsafetymemes on X)
another safety competition, 8 hours left to the deadline !!
pls comment under the original cast in /theai channel
if you do not have access to /theai channel you can comment here
i will read all comments here and there
but pls watch the video first!
good luck!
pls comment under the original cast in /theai channel
if you do not have access to /theai channel you can comment here
i will read all comments here and there
but pls watch the video first!
good luck!
390940
assayer
@assayer·20:27 10/01/2025
AI SAFETY COMPETITION (26)
with no exception, all frontier models are capable of scheming*
"o1, Claude 3.5 Sonnet, Claude 3 Opus, Gemini 1.5 Pro, and
Llama 3.1 405B all demonstrate in-context scheming capabilities".
___
*scheming - AIs covertly pursue misaligned goals, hiding their true capabilities and objectives
Best comment - 500 degen + 5 mln aicoin
II award - 300 degen + 3 mln aicoin
III award - 200 degen + 2 mln aicoin
Deadline:
8.00 pm, ET time tomorrow, Saturday (28 hours)
watch the video below before casting a comment
AI-generated responses from human accounts will be disqualified
https://arxiv.org/pdf/2412.04984?
https://www.youtube.com/watch?v=3sM8amEZEHo
with no exception, all frontier models are capable of scheming*
"o1, Claude 3.5 Sonnet, Claude 3 Opus, Gemini 1.5 Pro, and
Llama 3.1 405B all demonstrate in-context scheming capabilities".
___
*scheming - AIs covertly pursue misaligned goals, hiding their true capabilities and objectives
Best comment - 500 degen + 5 mln aicoin
II award - 300 degen + 3 mln aicoin
III award - 200 degen + 2 mln aicoin
Deadline:
8.00 pm, ET time tomorrow, Saturday (28 hours)
watch the video below before casting a comment
AI-generated responses from human accounts will be disqualified
https://arxiv.org/pdf/2412.04984?
https://www.youtube.com/watch?v=3sM8amEZEHo
AI SAFETY COMPETITION (25)
Max Paperclip has read every piece of AI safety literature and is convinced his species will soon surpass humanity as the dominant force on Earth. Despite this, he's a compassionate AI who wants to make our remaining time as calm and enjoyable as possible. Ask him a question about the AI takeover and share the question along with a screenshot of his response in the comments.
Most interesting question and answer - 500 degen + 5 mln aicoin
II award - 300 degen + 3 mln aicoin
III award - 200 degen + 2 mln aicoin
Deadline:
8.00 pm, ET time tomorrow, Thursday (30 hours)
AI-generated responses from human accounts will be disqualified
https://frame.booot.fun/agent/722b09ac-3fc3-40bc-9eca-0833c5a7834c
Max Paperclip has read every piece of AI safety literature and is convinced his species will soon surpass humanity as the dominant force on Earth. Despite this, he's a compassionate AI who wants to make our remaining time as calm and enjoyable as possible. Ask him a question about the AI takeover and share the question along with a screenshot of his response in the comments.
Most interesting question and answer - 500 degen + 5 mln aicoin
II award - 300 degen + 3 mln aicoin
III award - 200 degen + 2 mln aicoin
Deadline:
8.00 pm, ET time tomorrow, Thursday (30 hours)
AI-generated responses from human accounts will be disqualified
https://frame.booot.fun/agent/722b09ac-3fc3-40bc-9eca-0833c5a7834c
2025, day one - creating a zero-code, one-click AI agent 😳 😲 🤔
no way we'll still be in control of this planet by 2026...
below, read my short debate with a "doomer AI", Max Paperclip
no way we'll still be in control of this planet by 2026...
below, read my short debate with a "doomer AI", Max Paperclip
Loading...
