News for hackers. Stuff that matters.
Re:Snowden eviscerates OpenAI's decision to put a former NSA... by racional on Monday June 17, 2024 @03:26PM (#40709969)
Sounds like you're the kind of person who likes to build on on hypotheticals.

I'm not that kind of person. That said you can go on thinking that people who don't share your appetite for rank speculation are blithering idiots, as suits your fancy.

I'm fully aware of what he might have done. I just don't find it useful to pretend that I therefore know what he did.

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by mdhb on Monday June 17, 2024 @05:20PM (#40711206)
This isn’t a hypothetical. This is stated as a fact even in the absence of public evidence in intelligence circles that there is just zero chance in hell that he would find himself in that situation where he had zero good moves left that he wouldn’t be actively exploited for everything he knew before he would be offered any kind of asylum.

Russia actually had a strong incentive to turn him over if he wasn’t already an asset as not only did they already have all the file he had leaked but it was put to them at the time that this would effectively destroy the intelligence relationship between the two countries and that if he was actually acting independently that they didn’t stand to benefit by protecting him by offering him asylum.

Something happened to make that cost worth it to them. Again, this is in no way a controversial statement in professional circles.

Edit: I want to tighten up my position here for clarity’s sake. Was Snowden an operation is not taken as gospel in the IC but is a credible theory without compelling evidence against it but if it wasn’t and he effectively was an independent walk in that he would be expected to act as a consultant of sorts in exchange for ongoing protection. Both of those are really shitty outcomes however you look at it.

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by racional on Tuesday June 18, 2024 @03:45PM (#40721440)
It's certainly a credible thesis, what you're offering here, and yours is probably the best formulation I've heard of it.

Still, from my own brushings with folks in "professional circles" -- they tend to be pretty consistent about packaging assessments such as these with probabilistic labels (e.g. "likely", "highly likely" and so on). My main concern here is quick toggling to "known" or "near certain" status (or insinuations that "everyone in the IC believes this", which I am noting is not your intention here, but it did sound like it at first).

It's also just a general policy I like to follow with events out in the real, messy world. No matter good our "model" may sound (to our own ears) about what someone did or why, it's always entirely possible we're missing something basic -- not so much a piece of bombshell evidence, but about the basic human context of what's going on; and that our analysis simply flawed.

There's also significant tail risk that comes with putting too much stake in our models, especially when they get coupled together, and motivated reasoning inevitably starts to kick in.

On a finer-grained note (and to explain why I can't assign this narrative too high of a degree of likelihood): It's not automatically clear that the Russians could have traded Snowden in by now -- they may be asking too high of a price (since he's apparently way more valuable than anyone the US has in their trading deck right now), and since 2014 they may have been coupling it with their hopes that the US will decide to start being "reasonable" about the situation in Ukraine, and so forth.

I also don't see how we know that "they already had all the files he had leaked", or that anything about his case has the potential to destroy the intelligence relationship (which has always found a way of surviving, despite all the bile and vitriol we see on the surface).

(This is all in regard to the "weak" version of this model, that Snowden acted independently but had to have broken once got suck in the Motherland. The "strong" version in the sibling post -- which asserts that he must have been cooperating with his current hosts from the very start -- has just way too many plotholes, and is way too top-heavy in my view. Not to be dismissive, but I have to limit my time here, so that's all I can offer on it at the moment).

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by mdhb on Tuesday June 18, 2024 @05:43PM (#40722431)
You should really watch that video I posted when you have an opportunity. I think it would give you some helpful context here but I can’t accept some of your premises here unfortunately.

I don’t have any reason to think you have any experience whatsoever with that world or even a basic understanding of US intel history but they are major missing holes in your theory.

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by racional on Tuesday June 18, 2024 @11:12PM (#40724389)
I'm not offering a "theory"; but you did, and a plainly untenable one.

If the only response you can think of is to speculate about the background of the person pointing this out to you -- that's a pretty common psychological defense, and if it offers some form of comfort to you, you're more than entitled to it.

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by mdhb on Wednesday June 19, 2024 @02:02AM (#40725295)
You are missing my point entirely here. This is not some “psychological comfort” thing or whatever nonsense you’re offering.

I’m saying that we are currently talking about a topic that I happen to know a lot about and you happened to start thinking about when the conversation started.

So when you say that you aren’t sure if things sound kind of plausible it doesn’t actually hold any weight to the conversation because again it’s clear you don’t have any of the required background to have an informed opinion to give here.

It’s not a personal attack or anything, but it’s also nothing to do with trying to make myself feel better.

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by racional on Wednesday June 19, 2024 @02:36AM (#40725479)
It’s not a personal attack or anything,

But of course that's precisely what is. You're not providing any insight into the subject at hand here. You're simply talking up your vastly superior knowledge, and making weird projections about the person you're talking to.

What I notice about people who are genuinely knowledgeable and confident about a subject is that in general their deportment and tone are exact opposite of what's on display here.

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by mdhb on Wednesday June 19, 2024 @03:22AM (#40725745)
This doesn’t even make sense. Surely you must realise you don’t actually know anything about this topic. Like you aren’t secretly thinking you’re an expert in this in your mind.

It’s cool if you you don’t think I do or not; that’s not at all what I’m trying to convince you of. I know what I know, I’m not looking for your validation on that topic here because again you aren’t in a position to give it. Not that I don’t like you or any personal quality because I don’t know you.

Re:Snowden eviscerates OpenAI's decision to put a former NSA... by racional on Wednesday June 19, 2024 @12:01PM (#40729537)
Like you aren’t secretly thinking you’re an expert in this in your mind.

If you were genuinely confident of your knowledge, you wouldn't feel the need to repeatedly respond in a purely ad hominem manner such as this.