Hacker News Reader

Comment by disadvantage | Hacker News Reader

Humans have went dark considerably over the years. I don't know the exact date when encrypted comms became the norm, but we rarely transmit plaintext out into the Universe anymore. We still leak plaintext, but the vast majority is encrypted now.

View on HN

Compression is also of interest. A well-compressed file should resemble noise: if you can predict the next bit, you could have avoided sending it. As we've moved to digital formats, there's an increasing opportunity to save bandwidth through compression, making the 'plaintext' much less readable. You could theoretically backwards-engineer the mp3 format by seeing enough examples, but it's a harder problem than, say, playing back a record that someone velcro'ed onto the side of a space probe.

Pretty sure that was meant to be an analogy, not to be taken literally.

TV? FM Radio?

Both switched to digital broadcast format years ago. Looks like FM might be simulcast. And I am definitely not an expert so I can’t speak to the contents but I would assume encoded but not encrypted.

Everything is encoded somehow. That's not terribly germane to whether it's encrypted.

Naively, I guess that non-digital signal decoding is trivial compared to digital signal decoding where the signal might be indistinguishable from encryption without a reference spec or unencoded sample.

We have people smart enough to crack all kinds of signal encodings and actual attempts at encryption, and we're probably not even that advanced of a civilization. Also, I'm not an expert in EM signalling or whatever the field is called, but I understand that most physical encodings actually have a lot of structure that distinguishes them from random. I know for sure that plaintext signals tend to have a LOT of structure unless they're compressed, and that structure will be reflected to some extent in the physical encoding unless you take measures to stop it.

Meanwhile the usual standard for encryption is "indistinguishable from random"... which kinda just doesn't make sense at the physical level. I'd be very interested to know if anyone has figured out how to do physical encoding/encryption that's indistinguishable from physical noise.

I agree with most of what you’re saying but my thought was that digital broadcast signals are compressed and multiplexed so they’re likely to be much more complicated than analog FM transmissions. I think they also broadcast to equivalent range at lower power, so I assume they have checksum bits in the stream so the decoder can reconstruct or reject a weak or intermittent signal. Then, if an entity trying to decode the signal had no framework for human communication the difficulty would be significantly higher because it would need to approach the data using a pure information-theoretical approach. Meanwhile I can (theoretically :)) build a crystal radio that “decodes“ analog FM in an afternoon.

Reply on news.ycombinator.com