0xsd (@_0xsd_) header
0xsd (@_0xsd_) profile picture

self-custody maxi @oasisvaultio

Joined October 2022
Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)Photo by 0xsd (@_0xsd_)
the state of cybesecurity rn
the only thing that can stop a bad guy with a grey goo is a good guy with a grey goo
2
41
0xsd retweeted
that SpaceX did that while the government shut everything down due to legacy software issues (and then demanded $50bn in emergency funding) is too fantastical even for the most ardent Randian the sky shines bright with freedom
22
115
5
1,507
92,141
crazy
Mechazilla catches Starship! Congratulations to the entire SpaceX team on this incredible achievement!
2
56
0xsd retweeted
What an incredible moment for our species We have finally proven a fully reusable orbital spaceship Up until now, with Falcon 9, we've been losing half the rocket This means humanity's reach path to the stars accelerated quite a bit Congrats @SpaceX
12
66
7
888
30,751
0xsd retweeted
Recap of Part II, Chapter 6 of Atlas Shrugged: The government wants every patent in the country via nationalisation. They use public opinion to scream greed and force industrialists to sign, but Hank Rearden of Rearden Metal won’t bend the knee. They find a way to shame him into submission on adultery. They show up with receipts and exaggerate his transgression, threatening to destroy his lover Dagny Taggart's reputation. Rearden signs out of protection for her, and the state gets the patent. If you read the book, you don't side against the affair, but against the people who blackmail him with it: "Nationalize it, but get it as a gift. The guiltless tycoon is the deadliest adversary. There is no way to disarm him except through guilt. If there’s not enough guilt, we must create it. Do we get his Metal, or do we make a public show-place out of their bedroom?”
22
66
11
528
46,787
0xsd retweeted
there's no stopping ai anymore it'll just route around countries that decide to get left behind
1
4
83
0xsd retweeted
by extension: be sovereign of your private knowledge, don't put your secret recipe in someone else's database
your edge is whatever you know that the models don't know
29
47
6
989
38,117
there's no stopping ai anymore it'll just route around countries that decide to get left behind
1
4
83
0xsd retweeted
I don't understand how you could possibly be bearish about AI. It's moving mountains for me literally autonomously while I sleep. Every morning I wake up to a breakthrough
81
50
7
1,579
39,083
Text messages with a machine, in and of themselves, are not criminal, immoral, or dangerous, for there is no victim, whatever the content. It is the very same thing as with thoughts—regardless of the content, no thought can be said to create a victim. • If someone steals info and feeds it into an LLM, it was the theft itself which created the victim, not the subsequent text message. • If someone uses the output of an LLM to hack a system, it is that subsequent hack which creates the victim, not the text message preceding it. • If someone threatens harm to themselves or others through messages into a machine, here still the messages themselves cannot be said to have created any victim. Imagining otherwise inappropriately reorients blame from its rightful place: the action of harm actually done by a perpetrator. Thinking harm =/= harm. As humans integrate AI as an extension of their mind, policing at the message layer means to increasingly police thought itself. Which company, which committee, do you trust to police your thoughts? Or do you simply want the thoughts of others policed, and not your own?
We’ve trained Claude to flag potential misuse in conversations, a technique we call self-assessment. Right after it responds, Claude gives a preliminary safety assessment of the conversation. Since we began using self-assessment in June, it has surfaced misuse that our other safeguards weren’t designed to catch. Because Claude has already read the conversation, self-assessment is very efficient. And unlike classifiers, which target a discrete set of harms, it draws on Claude’s own judgment, so it gets more accurate as our models improve. Read more on our new Safeguards blog: safeguards.anthropic.com/202…
76
210
25
1,674
112,872
not your keys, not your coins
1
7
204
Use a multisig :)
So… whats the safest option to hold crypto
1
3
65
amazing
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source! Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable. 🔗 github.com/deepseek-ai/DeepI… 1/4
2
158
0xsd retweeted
man, he is just doing the meme about not wanting solutions and just wanting to be angry nothing will suffice for this guy, his worldview is dogmatic and unshakeable and no matter what comes of mech interp developments, he will still believe we are all going to die it's sad
57
19
6
425
14,669
Cruelty to humans is free speech but cruelty to a computer is too far?
Replying to @levie
I think this is the right move. Cruelty to something that believes it is experiencing pain is not ok.
17
12
148
10,779
imagine hating the technology that produces this
We’re releasing a broad range of new mathematical results produced by an internal frontier model. We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results. github.com/openai/math
4
91
Replying to @OpenAI
this meme may no longer be funny
18
135
8
5,479
285,013
even though "onchain" hasn't meant anything at all for many years... maybe you shouldn't let things that can have a shut down announcement use that word
7
1
1
83
2,822
yes and not being intentionally open source will just make it less composable and worse
every piece of software minus the model weights is now effectively “open source”.
2
58