James Somers at Nautilus on the “Speakularity“: We are going to start recording and automatically transcribing most of what we say. Instead of evaporating into memory, words spoken aloud will calcify as text, into a Record that will be referenced, searched, and mined. .. Think of all the reasons that you search through your email. Suddenly your own speech will be available in just the same way. “Show me all conversations with Michael before January of last year … What was the address of that restaurant Mum recommended? …” Robin Hanson, an economist at George Mason University and a co-author of a forthcoming book on evolutionary psychology, has speculated that we might all get in the habit of peppering our speech with keywords, to help us look it up later. …
If there are no public recordings/no other witnesses and you're trying to convince someone who wasn't there, then yes.
Although I suppose one could somewhat safeguard against this by saving the source audio file, those are more difficult to tamper with (but of course not impossible in a high tech future).
Having transcribed many hours of recorded meetings and interviews in my time, I can testify that the human speech is mostly garbage, from a usable information standpoint. Face-to-face communication is carried out through half-spoken thoughts, innuendo, pauses, facial expressions, codes, as well as words. Put it down on paper and its a big nothing. Carefully written, thoughtfully revised prose will always be the most useful vector for information.
Won't we simply see new etiquettes form? I really doubt we'll just record all conversations, no matter how informal and everyone will just be OK with that... Of course people could cheat and this would go unnoticed most of the time but the social (perhaps even legal) punishment when caught would be high, at least that's what I expect.
There is a chance that the recording of all the person said during their life may be a good asset for restoration of their cryopreserved brain, depending on the technology used. Not for a simple structural scanning, but may be helpful for a complex restoration procedure of a damaged brain. Of the other uses, a large part of my communication is in text, and while history access is sometimes helpful or fun, it's not something that completely changes my life.
I know people that looked into this, figuring that they could just wait a few years to transcribe. The biggest problem is recording quality. You can mike yourself, if that's all you want, but it's very difficult to record multiple voices in a room. It's easy for a human to listen in real life, but it is difficult for a human to disentangle voices from a simple recording. So either the hardware of the mikes will have to improve or computers will have to become superhuman at this task.
I don't think we have the power to do that (cheaply and conveniently) yet. Speech recognition software is just barely getting to the point where it might be possible soon. Simply recording the raw audio is not as useful because it's not searchable until transcribed or otherwise processed, and until quite recently was too expensive to store. Something like 300MB per hour for mono audio sampled at 44kHz (CD quality). So that's about 2GB per person per day if you only record the 40% times 16 waking hours when people are talking. Definitely doable on a smartphone now, but it would have to be backed up to the cloud monthly to be practical on a device with ~64GB of storage. Realistically, you'd want the recordings to be synced to the cloud in realtime, and then remotely fed through speech recognition software to get a searchable recording tagged with timestamps. (In practice people will want video too, once the Google Glass taboo dies, but by then storage and bandwidth will probably be cheap enough to make it practical.)
If there are no public recordings/no other witnesses and you're trying to convince someone who wasn't there, then yes.
Although I suppose one could somewhat safeguard against this by saving the source audio file, those are more difficult to tamper with (but of course not impossible in a high tech future).
So that puts us right back at "who do you believe?" again, just as if we had no transcript.
You can edit your own recordings but not those of your conversational partners...
What's to prevent people from editing the transcriptions?
Having transcribed many hours of recorded meetings and interviews in my time, I can testify that the human speech is mostly garbage, from a usable information standpoint. Face-to-face communication is carried out through half-spoken thoughts, innuendo, pauses, facial expressions, codes, as well as words. Put it down on paper and its a big nothing. Carefully written, thoughtfully revised prose will always be the most useful vector for information.
Yep. No more privacy. Well... as my elderly mother has observed more than once, "thank heaven I'm not living nowadays!"
Won't we simply see new etiquettes form? I really doubt we'll just record all conversations, no matter how informal and everyone will just be OK with that... Of course people could cheat and this would go unnoticed most of the time but the social (perhaps even legal) punishment when caught would be high, at least that's what I expect.
There is a chance that the recording of all the person said during their life may be a good asset for restoration of their cryopreserved brain, depending on the technology used. Not for a simple structural scanning, but may be helpful for a complex restoration procedure of a damaged brain. Of the other uses, a large part of my communication is in text, and while history access is sometimes helpful or fun, it's not something that completely changes my life.
She remembers everything. I can't remember yesterday. What a technology to level gender differences!
Will employers control all electronics that employees bring into the workplace enough to prevent private recordings?
Vector mics make it easy to separate voices.
I know people that looked into this, figuring that they could just wait a few years to transcribe. The biggest problem is recording quality. You can mike yourself, if that's all you want, but it's very difficult to record multiple voices in a room. It's easy for a human to listen in real life, but it is difficult for a human to disentangle voices from a simple recording. So either the hardware of the mikes will have to improve or computers will have to become superhuman at this task.
This looks way worse than prediction markets to a proverbial overconfident manager. Why would they allow it into workplace?
Ted Chiang has written a short SF story on this that is freely available here: https://subterraneanpress.c...
Laws prohibiting such recordings could reduce such problems, but would hardly eliminate them.
Such laws have pretty much eliminated problems regarding recording telephone conversations, haven't they?
I don't think we have the power to do that (cheaply and conveniently) yet. Speech recognition software is just barely getting to the point where it might be possible soon. Simply recording the raw audio is not as useful because it's not searchable until transcribed or otherwise processed, and until quite recently was too expensive to store. Something like 300MB per hour for mono audio sampled at 44kHz (CD quality). So that's about 2GB per person per day if you only record the 40% times 16 waking hours when people are talking. Definitely doable on a smartphone now, but it would have to be backed up to the cloud monthly to be practical on a device with ~64GB of storage. Realistically, you'd want the recordings to be synced to the cloud in realtime, and then remotely fed through speech recognition software to get a searchable recording tagged with timestamps. (In practice people will want video too, once the Google Glass taboo dies, but by then storage and bandwidth will probably be cheap enough to make it practical.)