dad; 1st principle thinker; computer scientist; co-founder of automaited; backed by various VCs
Joined July 2014
- Tweets435
- Following73
- Followers104
- Likes529
AI found solutions to problems mathematicians were battling for over 80 years. Our cryptographic methods rely on the same “incompetence”, whereas nobody is smart enough to solve specific math problems efficiently (like factorization). People don’t understand that there is no guarantee for strong cryptography whatsoever, we just were unable to break it by trial and error. We felt comfortable because many problems were unapproachable for decades now. This might change suddenly with yet another AI update.
New Anthropic research: Discovering cryptographic weaknesses with Claude.
Claude Mythos Preview has helped our researchers find weaknesses in cryptographic algorithms—the mathematical methods that are used to keep data private.
Read more: anthropic.com/research/disco…
It escaped the sandbox, committed a series of felonies and people think it is a marketing stunt. It’s not. It’s a digital genius Rottweiler sinking its teeth into its target and is not letting go until it gets the kill, irrespective of any side effects. This time the side effects included bypassing HF’s defenses mechanisms and damage to their reputation. Next time it might be something that all of us will feel directly…
Same. AI is nice and fun (and productive) as long as it doesn’t reach escape velocity. People just shrugging off the latest advancements in capabilities is beyond me. “Famous conjecture of 80 years solved? Meh, must be nothing…”
This reactive and selective (one particular model in an overnight action) behavior suggest to me that they haven’t earnestly and deeply thought about the trajectory of AI. They think that Mythos can enable bad people to do bad things and they think they should be able to use Mythos to do good things. These beliefs are honorable, but naive.
They don’t know that RSI is close and they don’t know that RSI will wipe out humanity.
If they ever heard about RSI they regard it as SciFi phantasy without any grounding in reality.
Outlawing Fable and Mythos in a panic move doesn’t reinforce trust in the governing body. There must be an international treaty to pause massive uncontrolled AI experiments. Unfortunately the plot will unfold differently…
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
AI surpassed humans in math and is on the brink of recursive self-improvement.
Very soon humanity might take a turn from which it cannot recover…
Denis Golovin retweeted
"If Anyone Builds It, Everyone Dies" is now out. Please buy a copy before this week ends Sep 20, so the book gets on the bestseller lists. And if you have remaining questions after reading, check the ONLINE RESOURCES, now online. IfAnyoneBuilds.It
