Re: Re[2]: [draft] Defeating Botnets and Trojans: Episode 1
The purpose of my life is spitting out ideas, the one less probable then the other, and people who own the network, and my computer, use (or steal ymmv) what can be used. Normally I don't write to mailing lists, I'm just a lurker, but this had to be said. Normally I just write my ideas in the title bar of file explorer and I don't save them. My schizophrenic and naive mind does compare to a 12 year old ... but I love it still having imagination and creativity. Basic assumptions are wild imho, you easily overlook something. Op vr 10 jun. 2022 om 18:15 schreef enlight <enlight@riseup.net>:
you sound like a 12 year old, with naive imaginations. If you have studied computer science in college, generally in any computer system, there are some basic assumptions and we build our foundations on that.
------ Original Message ------ From "Nico Verrijdt" <nicoverrijdt@gmail.com> To "Karl Semich" <0xloem@gmail.com>; cypherpunks@lists.cpunks.org Date 07/06/2022 13:53:38 Subject Re: [draft] Defeating Botnets and Trojans: Episode 1
Imho the operating systems we use aren't adequate.
No single user is 100% accountable for what he/she does on the network.
I'm not an expert in this field at all, but I thought about this quite a lot figuring things out.
Here it goes: Suppose you implement a blockchain for identities as early in the boot process as possible, then you find the public key of every user in the blockchain and can communicate with every user while that user knows you are connected to him/her. Now you have a blockchain and can connect to virtually anyone. But the operating system is still vulnerable to hijacking and you should verify correctness somehow. Suppose, I'm really not sure if this effectively works, you implement a hashing function as early as possible in the compiler, then you can hash the code of the compiler and verify if the compiler's code changed. Then up to the operating system which you compile with that compiler and hash the code in its development stages. Now you're able to connect other users in the network and compare hashes of comparable computer systems. Maybe there are better alternatives. What happens in memory is a black box to me, somehow I like to hash and compare that too, and I understand that our computers become quite a lot slower.
Privacy issues do arise with this system, but it's nothing more than in real life where you meet someone and say 'Hi, I'm ...', politeness, naivety, in this world now where the internet is totally different, not moral indeed. And also, the internet is free, with such a system an internet with different speeds might be more easily made as every user is known within a certain time span.
I'm currently working on a blockchain of identities ( https://github.com/nvrrdt/onzecurrency) and I'd like to continue developing this, so I hope the concept has a chance to succeed in its endeavor.
Op di 7 jun. 2022 om 10:20 schreef Nico Verrijdt <nicoverrijdt@gmail.com>:
Imho the operating systems we use aren't adequate.
No single user is 100% accountable for what he/she does on the network.
I'm not an expert in this field at all, but I thought about this quite a lot figuring things out.
Here it goes: Suppose you implement a blockchain for identities as early in the boot process as possible, then you find the public key of every user in the blockchain and can communicate with every user while that user knows you are connected to him/her. Now you have a blockchain and can connect to virtually anyone. But the operating system is still vulnerable to hijacking and you should verify correctness somehow. Suppose, I'm really not sure if this effectively works, you implement a hashing function as early as possible in the compiler, then you can hash the code of the compiler and verify if the compiler's code changed. Then up to the operating system which you compile with that compiler and hash the code in its development stages. Now you're able to connect other users in the network and compare hashes of comparable computer systems. Maybe there are better alternatives. What happens in memory is a black box to me, somehow I like to hash and compare that too, and I understand that our computers become quite a lot slower.
Privacy issues do arise with this system, but it's nothing more than in real life where you meet someone and say 'Hi, I'm ...', politeness, naivety, in this world now where the internet is totally different, not moral indeed. And also, the internet is free, with such a system an internet with different speeds might be more easily made as every user is known within a certain time span.
I'm currently working on a blockchain of identities ( https://github.com/nvrrdt/onzecurrency) and I'd like to continue developing this, so I hope the concept has a chance to succeed in its endeavor.
Op di 7 jun. 2022 om 09:22 schreef Karl Semich <0xloem@gmail.com>:
Cypherpunks, Anarchists, and Security Professionals can agree on something. Something we have agreed on for a long time, but fought viciously over.
We need uses of computer networks to be moral.
Our communities have been attacked for quite some time now.
but this had to be said. Normally I just write my ideas in the title bar of file explorer and I don't save them.
i did this for a couple years i'm guessing it's a human-trafficking-surveillance thing, lets people communicate deniably with keyloggers
but this had to be said. Normally I just write my ideas in the title bar of file explorer and I don't save them.
i did this for a couple years
i'm guessing it's a human-trafficking-surveillance thing, lets people communicate deniably with keyloggers
I encountered quite a few coincidences in the last 15+ years. One such coincidence is when I once wrote 'that's two flies in one clap', right after writing this a few desks further someone clapped hard on his desk, that was in a quiet place. There were also times that the management seemed to react immediately to what I typed. Op za 11 jun. 2022 om 09:57 schreef Undiscussed Horrific Abuse, One Victim of Many <gmkarl@gmail.com>:
but this had to be said. Normally I just write my ideas in the title bar of file explorer and I don't save them.
i did this for a couple years
i'm guessing it's a human-trafficking-surveillance thing, lets people communicate deniably with keyloggers
I encountered quite a few coincidences in the last 15+ years. One such coincidence is when I once wrote 'that's two flies in one clap', right after writing this a few desks further someone clapped hard on his desk, that was in a quiet place.
This sounds like a coincidence to me, because the sets of words in sentences and nearby behaviors are huge, frequent, and happen adjacent to each other.
There were also times that the management seemed to react immediately to what I typed.
My understanding is that this is considered be corporate machine learning algorithms that have been overfit. The described scenario is that they don't know what you typed, but AIs trying to make money or politics (even at other businesses, in browser advertisements, for example) are constantly trying things to see if they work, and we learn to respond to the things they try (like suggestions). It kind of seems like human body language and intuition could have gotten a little entrained. I don't know how to find people who talk about that, but https://www.humanetech.com/ puts some effort into connecting people together who do. On this list, a big topic that isn't addressed directly very often, but is continuously addressed indirectly, is information security conflict. It's very normal for mainstream machines to be beset by systems (backdoors) that let other groups control them. It's a dangerous and relevant situation, extremely complicated by AI, where often it can seem that a bunch of entities appear invested in clouding discussion around it. Spy agencies. Governments. Military corporations. Political parties. Other people likely know more than me about these topical things. I understand there are some groups at DefCon and likely other Cons regarding AI security. This is a pretty important topic. I haven't been to one myself.
I encountered quite a few coincidences in the last 15+ years. One such coincidence is when I once wrote 'that's two flies in one clap', right after writing this a few desks further someone clapped hard on his desk, that was in a quiet place.
This sounds like a coincidence to me, because the sets of words in sentences and nearby behaviors are huge, frequent, and happen adjacent to each other.
I don't believe much in coincidence. Another such a coincidence is when I needed an eraser and somehow I typed that down while a short moment later an eraser flew from one side of the office to the other side, from one colleague to another. The probability of this happening is extremely low as an eraser was, even then, not much used anymore and one colleague in need for an eraser while other throwing one at the same time of me writing that, ..., I was never that fortunate when playing the lottery. I just like to think they snooped on the characters from my file explorer's title bar. Op za 11 jun. 2022 om 12:55 schreef Undiscussed Horrific Abuse, One Victim of Many <gmkarl@gmail.com>:
I encountered quite a few coincidences in the last 15+ years. One such coincidence is when I once wrote 'that's two flies in one clap', right after writing this a few desks further someone clapped hard on his desk, that was in a quiet place.
This sounds like a coincidence to me, because the sets of words in sentences and nearby behaviors are huge, frequent, and happen adjacent to each other.
There were also times that the management seemed to react immediately to what I typed.
My understanding is that this is considered be corporate machine learning algorithms that have been overfit. The described scenario is that they don't know what you typed, but AIs trying to make money or politics (even at other businesses, in browser advertisements, for example) are constantly trying things to see if they work, and we learn to respond to the things they try (like suggestions). It kind of seems like human body language and intuition could have gotten a little entrained.
I don't know how to find people who talk about that, but https://www.humanetech.com/ puts some effort into connecting people together who do.
On this list, a big topic that isn't addressed directly very often, but is continuously addressed indirectly, is information security conflict. It's very normal for mainstream machines to be beset by systems (backdoors) that let other groups control them. It's a dangerous and relevant situation, extremely complicated by AI, where often it can seem that a bunch of entities appear invested in clouding discussion around it.
Spy agencies. Governments. Military corporations. Political parties. Other people likely know more than me about these topical things.
I understand there are some groups at DefCon and likely other Cons regarding AI security. This is a pretty important topic. I haven't been to one myself.
On 6/11/22, Nico Verrijdt <nicoverrijdt@gmail.com> wrote:
I encountered quite a few coincidences in the last 15+ years. One such coincidence is when I once wrote 'that's two flies in one clap', right after writing this a few desks further someone clapped hard on his desk, that was in a quiet place.
This sounds like a coincidence to me, because the sets of words in sentences and nearby behaviors are huge, frequent, and happen adjacent to each other.
I don't believe much in coincidence. Another such a coincidence is when I needed an eraser and somehow I typed that down while a short moment later an eraser flew from one side of the office to the other side, from one colleague to another. The probability of this happening is extremely low as an eraser was, even then, not much used anymore and one colleague in need for an eraser while other throwing one at the same time of me writing that, ..., I was never that fortunate when playing the lottery. I just like to think they snooped on the characters from my file explorer's title bar.
There were actually two responses in my reply; in one I considered the experience a coincidence, in the other I considered it to be related to ongoing acts of surveillance, marketing, and influence. The things you describe are being described by a lot of people. They are also symptoms of heightened paranoia. The fact that these things are real, is heightening people's paranoia, and we get a lot of real things we would not think are true, and also a lot of worrisome things that can be just happenstance. It's hard to tease them apart. The speed of the event described implies to me that both you, and the other people, were primed with the concept of eraser somehow. I could be wrong. I can also write you a python script to demonstrate that coincidences exist: but really the fact that we notice them shows that they're not coincidences, as our mind is not a uniform random distribution. About a decade ago, a paper was shared on these lists, maybe this one, regarding tracking causality. I'm sorry I don't know more about it, but I imagine researchers far off in their math towers, trying to make the world make sense.
Op za 11 jun. 2022 om 12:55 schreef Undiscussed Horrific Abuse, One Victim of Many <gmkarl@gmail.com>:
I encountered quite a few coincidences in the last 15+ years. One such coincidence is when I once wrote 'that's two flies in one clap', right after writing this a few desks further someone clapped hard on his desk, that was in a quiet place.
This sounds like a coincidence to me, because the sets of words in sentences and nearby behaviors are huge, frequent, and happen adjacent to each other.
There were also times that the management seemed to react immediately to what I typed.
My understanding is that this is considered be corporate machine learning algorithms that have been overfit. The described scenario is that they don't know what you typed, but AIs trying to make money or politics (even at other businesses, in browser advertisements, for example) are constantly trying things to see if they work, and we learn to respond to the things they try (like suggestions). It kind of seems like human body language and intuition could have gotten a little entrained.
I don't know how to find people who talk about that, but https://www.humanetech.com/ puts some effort into connecting people together who do.
On this list, a big topic that isn't addressed directly very often, but is continuously addressed indirectly, is information security conflict. It's very normal for mainstream machines to be beset by systems (backdoors) that let other groups control them. It's a dangerous and relevant situation, extremely complicated by AI, where often it can seem that a bunch of entities appear invested in clouding discussion around it.
Spy agencies. Governments. Military corporations. Political parties. Other people likely know more than me about these topical things.
I understand there are some groups at DefCon and likely other Cons regarding AI security. This is a pretty important topic. I haven't been to one myself.
participants (2)
-
Nico Verrijdt -
Undiscussed Horrific Abuse, One Victim of Many