4 ms·

Like what? All I can think of is tracking GPU purchases but that won't be possible when AMD and NV have viable international competitors.

mdhb 2y ago
There’s a famous saying in cryptography that says “anyone is capable of building encryption algorithm that they can’t break” which I am absolutely positively sure applies here also. In a world full of sensors where everything is logged in some way or another I think that it would actually be not a straightforward activity at all to build a clandestine AI lab at any scale. In the professional intel community they have been talking about this as a general problem for at least a decade now.
- jsty 2y ago
  > In the professional intel community they have been talking about this as a general problem for at least a decade now. As in they've been discussing detecting clandestine AI labs? Or just how almost no activity is now in principle undetectable?
  - mdhb 2y ago
    I’m referring to the wider issue of what’s referred to by the Americans as “ubiquitous technical surveillance” where they came to the kind of upsetting conclusion for them that they had a long time ago lost the ability to even operate in London without the Brits knowing. I don't think there’s a good public understanding of just how much things have changed in that space in the last decade but a huge percentage of all existing tradecraft had to be completely scrapped because not only does it not work anymore but it will put you on the enemy’s radar very early on and is actively dangerous. It’s also why I think a lot of advice I see targeted towards activist types I think is straight up a bad idea in 2025. It just typically involves a lot of things that aren’t really consistent with any kind of credible innocuous explanation and are very unusual which make you stand out from a crowd.
    - snickerbockers 2y ago
      But does that apply to other countries that are operating within their own territory? China is generally the go-to 'boogeyman' when people are talking about the dangers of AI; they are intelligent and extremely industrialized, and have a history of antagonistic relationships with 'the west'. I don't think it's unreasonable to assume that they will eventually have the capability to design and produce their own GPUs capable of competing with the best of NV and AMD; how will the rest of the world know if China is producing a new AI that violates a hypothetical 'AI non-proliferation treaty'? Interesting semi-irrelevant tangent: the Cooley/Tukey 'Fast Fourier Transform' algorithm was initially created because they were negotiating arms control treaties with the Russians, but in order for that to be enforceable they needed a way to detect nuclear weapons testing; the solution was to use seismograms to detect the tremors caused by an underground nuclear detonation, and the FFT was invented in the process because they were using computers to filter for the types of tremors created by a nuclear weapon.
      - mdhb 2y ago
        I’m actually in agreement with you here. I think it’s probably reasonable to assume that through some kind of combination of home grown talent and their prolific IP theft programs that they are going to end up with that capability at some point the only thing in debate here is the timeline. As I understand things (I’m not actually a professional here) the current thinking has up to this point been something akin to a containment strategy largely based on lessons learned from years of nuclear non-proliferation work. But things are developing at such a crazy pace and there are some major differences between this and nuclear technology that it’s not really a straightforward copy and paste strategy at all. For example this time around a huge amount of the research comes from the commercial sector completely independently of defense and is also open source. Also thanks for that anecdote I hadn’t heard of that before. This is a bit of a long shot but maybe you might know, I was trying to think of some research that came out maybe 2-3 years ago that basically had the ability to remotely detect if anything in a room had been moved (I might be misremembering this slightly) and it was said to be potentially a big breakthrough for nuclear arms control. I can’t remember what the hell it was called or anything else about it, do you happen to know?
- mcphage 2y ago
  > There’s a famous saying in cryptography that says “anyone is capable of building encryption algorithm that they can’t break” That’s a new one on me (not being in cryptography), but I really like it. Thanks!
  - snickerbockers 2y ago
    It reminds me of all the idiot politicians who want to 'regulate' cryptography, as if the best encryption algorithms in the world don't already have open-source implementations that anyone can download for free.
  - daedrdev 2y ago
    I think the better cryptography lesson is that you should not build your own cryptography system because you will mess up and include a security flaw that will allow the data to be read.
    - deadbabe 2y ago
      That’s why you get AI to build it instead.
    - mdhb 2y ago
      That’s literally the same underlying lesson just stated differently.
    - mcphage 2y ago
      Isn't that the joke? “Anyone is capable of building encryption algorithm that they can’t break” doesn't say that nobody can break it, and it's probably easy for people to assume that "well I can't break it, then it's probably unbreakable" only to discover that in fact it is not unbreakable.
      - antifa 2y ago
        Yes, but it's written more clearly for the people who need to hear the message.
- fennecfoxy 2y ago
  A world full of sensors purchased from, developed and managed by a bunch of inept contractor corporate BS companies who usually have no idea what they're doing and use lawyers and MBAs with fancy college connections to keep their niche carved out.
mywittyname 2y ago
Electricity usage, network traffic patterns, etc. If a "data center" is consuming a ton of power but doesn't seem to have an alternate purpose, then it's probably training AI. And maybe it will be like detecting nuclear enrichment. Instead of hacking the firmware in a Siemens device, it's done on server hardware. Israel demonstrated absurd competence at this caliber of spycraft. Sometimes you take low-tech approaches to high tech problems. I.e., get an insider at a shipping facility to swap the labels on two pallets of GPUs, one is authentic originals from the factory and the other are hacked firmware variants of exactly the same models.
- hn_throwaway_99 2y ago
  None of these techniques are actionable. So what, someone is training AI, it's not like anyone is proposing restricting that. People are trying to make a distinction between "bad AI" and "good AI", like that is a possibility, and that's what the argument basically is, that it's impossible to differentiate or detect the difference between those, and signing declarations pretending you can is worse than useless.
  - thorum 2y ago
    Isn’t that moving the goalposts? The claim was made that it’s impossible to detect AI training runs and investigate what’s going on or take regulatory action. In fact, it is very possible.
    - hn_throwaway_99 2y ago
      2 points: 1. I was just granting the GPs point to make the broader point that, for the purposes of this original discussion about these "safety declarations", this is immaterial. These safety declarations are completely unenforceable even if you could detect that someone was training AI. 2. Now, to your point about moving the goalposts, even though I say "if you could detect that someone was training AI", I don't actually even think that is possible. There are far too many normal uses of data centers to determine if one particular use is "training an AI" vs. some other data intensive use. I mean, there have long been supercomputer centers that do stuff like weather analysis and prediction, drug discovery analysis, astronomy tools, etc. that all look pretty indistinguishable from "training an AI" from the outside.

mdhb 2y ago

There’s a famous saying in cryptography that says “anyone is capable of building encryption algorithm that they can’t break” which I am absolutely positively sure applies here also. In a world full of sensors where everything is logged in some way or another I think that it would actually be not a straightforward activity at all to build a clandestine AI lab at any scale. In the professional intel community they have been talking about this as a general problem for at least a decade now.

jsty 2y ago
> In the professional intel community they have been talking about this as a general problem for at least a decade now. As in they've been discussing detecting clandestine AI labs? Or just how almost no activity is now in principle undetectable?
- mdhb 2y ago
  I’m referring to the wider issue of what’s referred to by the Americans as “ubiquitous technical surveillance” where they came to the kind of upsetting conclusion for them that they had a long time ago lost the ability to even operate in London without the Brits knowing. I don't think there’s a good public understanding of just how much things have changed in that space in the last decade but a huge percentage of all existing tradecraft had to be completely scrapped because not only does it not work anymore but it will put you on the enemy’s radar very early on and is actively dangerous. It’s also why I think a lot of advice I see targeted towards activist types I think is straight up a bad idea in 2025. It just typically involves a lot of things that aren’t really consistent with any kind of credible innocuous explanation and are very unusual which make you stand out from a crowd.
  - snickerbockers 2y ago
    But does that apply to other countries that are operating within their own territory? China is generally the go-to 'boogeyman' when people are talking about the dangers of AI; they are intelligent and extremely industrialized, and have a history of antagonistic relationships with 'the west'. I don't think it's unreasonable to assume that they will eventually have the capability to design and produce their own GPUs capable of competing with the best of NV and AMD; how will the rest of the world know if China is producing a new AI that violates a hypothetical 'AI non-proliferation treaty'? Interesting semi-irrelevant tangent: the Cooley/Tukey 'Fast Fourier Transform' algorithm was initially created because they were negotiating arms control treaties with the Russians, but in order for that to be enforceable they needed a way to detect nuclear weapons testing; the solution was to use seismograms to detect the tremors caused by an underground nuclear detonation, and the FFT was invented in the process because they were using computers to filter for the types of tremors created by a nuclear weapon.
    - mdhb 2y ago
      I’m actually in agreement with you here. I think it’s probably reasonable to assume that through some kind of combination of home grown talent and their prolific IP theft programs that they are going to end up with that capability at some point the only thing in debate here is the timeline. As I understand things (I’m not actually a professional here) the current thinking has up to this point been something akin to a containment strategy largely based on lessons learned from years of nuclear non-proliferation work. But things are developing at such a crazy pace and there are some major differences between this and nuclear technology that it’s not really a straightforward copy and paste strategy at all. For example this time around a huge amount of the research comes from the commercial sector completely independently of defense and is also open source. Also thanks for that anecdote I hadn’t heard of that before. This is a bit of a long shot but maybe you might know, I was trying to think of some research that came out maybe 2-3 years ago that basically had the ability to remotely detect if anything in a room had been moved (I might be misremembering this slightly) and it was said to be potentially a big breakthrough for nuclear arms control. I can’t remember what the hell it was called or anything else about it, do you happen to know?
mcphage 2y ago
> There’s a famous saying in cryptography that says “anyone is capable of building encryption algorithm that they can’t break” That’s a new one on me (not being in cryptography), but I really like it. Thanks!
- snickerbockers 2y ago
  It reminds me of all the idiot politicians who want to 'regulate' cryptography, as if the best encryption algorithms in the world don't already have open-source implementations that anyone can download for free.
- daedrdev 2y ago
  I think the better cryptography lesson is that you should not build your own cryptography system because you will mess up and include a security flaw that will allow the data to be read.
  - deadbabe 2y ago
    That’s why you get AI to build it instead.
  - mdhb 2y ago
    That’s literally the same underlying lesson just stated differently.
  - mcphage 2y ago
    Isn't that the joke? “Anyone is capable of building encryption algorithm that they can’t break” doesn't say that nobody can break it, and it's probably easy for people to assume that "well I can't break it, then it's probably unbreakable" only to discover that in fact it is not unbreakable.
    - antifa 2y ago
      Yes, but it's written more clearly for the people who need to hear the message.
fennecfoxy 2y ago
A world full of sensors purchased from, developed and managed by a bunch of inept contractor corporate BS companies who usually have no idea what they're doing and use lawyers and MBAs with fancy college connections to keep their niche carved out.

mywittyname 2y ago

Electricity usage, network traffic patterns, etc. If a "data center" is consuming a ton of power but doesn't seem to have an alternate purpose, then it's probably training AI. And maybe it will be like detecting nuclear enrichment. Instead of hacking the firmware in a Siemens device, it's done on server hardware. Israel demonstrated absurd competence at this caliber of spycraft. Sometimes you take low-tech approaches to high tech problems. I.e., get an insider at a shipping facility to swap the labels on two pallets of GPUs, one is authentic originals from the factory and the other are hacked firmware variants of exactly the same models.

hn_throwaway_99 2y ago
None of these techniques are actionable. So what, someone is training AI, it's not like anyone is proposing restricting that. People are trying to make a distinction between "bad AI" and "good AI", like that is a possibility, and that's what the argument basically is, that it's impossible to differentiate or detect the difference between those, and signing declarations pretending you can is worse than useless.
- thorum 2y ago
  Isn’t that moving the goalposts? The claim was made that it’s impossible to detect AI training runs and investigate what’s going on or take regulatory action. In fact, it is very possible.
  - hn_throwaway_99 2y ago
    2 points: 1. I was just granting the GPs point to make the broader point that, for the purposes of this original discussion about these "safety declarations", this is immaterial. These safety declarations are completely unenforceable even if you could detect that someone was training AI. 2. Now, to your point about moving the goalposts, even though I say "if you could detect that someone was training AI", I don't actually even think that is possible. There are far too many normal uses of data centers to determine if one particular use is "training an AI" vs. some other data intensive use. I mean, there have long been supercomputer centers that do stuff like weather analysis and prediction, drug discovery analysis, astronomy tools, etc. that all look pretty indistinguishable from "training an AI" from the outside.