Check out my cool new protocol that looks just like I am loading a webpage about cat facts, which is actually a hidden VPN that I use to secretly look at webpages about cat facts.
There is actually a technique called steganography, that does exactly that. It is used to hide arbitrary binary info inside images, while still fooling your eyes into thinking there is nothing sketchy there.
I didn’t say that it can’t be detected. I said it fools your eyes.
Besides that, stop using ML for everything. My guess is that you need insane amounts of processing power for ML to detect hidden messages inside terabytes of live internet traffic.
In fact, the algorithm for steganography is standard. It’s probably trivial to detect it, unless you add encryption and padding to the mix.
Even if they go that route, and frankly I think they would get lynched before we got to that point, they can’t monitor every single connection. That just way too much traffic.
That’s why China has a firewall, because that’s the best option they can come up with because monitoring every Chinese persons data is an impossible task. Their only option would be to go North Korea route, and just close the internet but that would basically end their economy.
Good news, we closed that pesky loophole by banning encryption without backdoors.
If they can’t decode it, you better be ready to explain exactly what those bytes were!
Check out my cool new protocol that looks just like I am loading a webpage about cat facts, which is actually a hidden VPN that I use to secretly look at webpages about cat facts.
You get me. You’re my kind of person! ᓚᘏᗢ
There is actually a technique called steganography, that does exactly that. It is used to hide arbitrary binary info inside images, while still fooling your eyes into thinking there is nothing sketchy there.
Can’t it be detected? I imagine ML could be used to automate to some extent.
I didn’t say that it can’t be detected. I said it fools your eyes.
Besides that, stop using ML for everything. My guess is that you need insane amounts of processing power for ML to detect hidden messages inside terabytes of live internet traffic.
In fact, the algorithm for steganography is standard. It’s probably trivial to detect it, unless you add encryption and padding to the mix.
“Stop using ML for everything”
I see no other way to drink from the firehose. We’re talking nationstate level resources.
I know! Nothing about all this is new.
The only new thing is that the UK government is about to learn about those things.
Even if they go that route, and frankly I think they would get lynched before we got to that point, they can’t monitor every single connection. That just way too much traffic.
That’s why China has a firewall, because that’s the best option they can come up with because monitoring every Chinese persons data is an impossible task. Their only option would be to go North Korea route, and just close the internet but that would basically end their economy.
Why do you think the telecoms got away with stalling upgrades and fiber roll outs for so long?