Comments

French Mailman November 16, 2023 3:29 PM

Interesting initiative. This will surely become a much needed capability in the near future.

There is a mistake in the published timeline – the first date should read November 2023, not November 2024.

pup vas November 16, 2023 4:43 PM

How AI can detect diabetes with a 10-second voice sample
https://www.dw.com/en/how-ai-can-detect-diabetes-with-a-10-second-voice-sample/a-67400425

=Misuse cannot be ruled out

Critics and data protectionists have warned of the enormous risk of voice analysis software being used in bad faith, for example by employers or insurance call centers.

There is a risk voice analysis software could be used without explicit consent and that customers or employees could be disadvantaged on the basis of personal health information.

It would also be comparatively easy for sensitive medical information to be passed on, hacked, sold or otherwise misused. However, clear regulations and limits on voice analysis as a diagnostic tool cannot be set by scientists. This falls squarely within the purview of politics.=

Dragons Peek November 17, 2023 4:15 PM

This new kind of issue (voice cloning) is exactly why I decided against using a different personal (amateur) security technique:

For a while I had been considering recording my own voice (at professional high fidelity) with me saying a wide variety of things, from totally to serious to totally silly, from meaninful to ridiculous, from verbose lexical to just coughs and laugher or even me actually crying.

The initial purpose and goal was to create a large set of convincing audio and upload them to a public file sharing site as fully downloadable, public domain or copyleft (or creative commons) free use types of files. They likely would have been downloadable as monaural MP3 files for smartphone ringtone compatibility.

They would have been available for anybody in the world of www to download and set as a phone ringtone.
Ideally, I would have optimised each file for 100% playback compatibilty (backwards and forwards) and for intelligeability and average RMS levels.

I had to shelf the idea though.
It just seemed to risky in case any hostile people decided to use the project to camouflage themselves.
And then of course, as soon as voice cloning showed up, it seemed like a really bad idea for me to do such a project.

I did attempt to forewarn both the FBI and a portion of the US Military, however.
I wanted them to know that I was considering doing the project, so that they would’nt be surprised or confused.

My main 2 goals (would have been):

A) Audio Decoy: make it more likely that anybody using my voice as a ringtone would function as an ad hoc soundalike for me.

b) Reduce the accuracy and frequency of me being stalked by either algorithms or people.

If the project had ever been initiated, I was prepared to start uploading to foreign locations first, in high quantities, and with explanations. Then, secondly, eventually do the same for my home country.

It’s a shame that I couldn’t really do it though; I had to break some promises, and I really do have some oddball stalkers from time to time making my life miserable at times.

I even already told people what the name of the download was preplanned to be: “The passphrase is an entire bluray DVD ROM”.

Nevertheless, I never do any voice automated banking; I tried to convince my bank to quit marketing that kind of “service” (vulnerability).

Last but not least, there are amazing things possible with synthesized speech these days. Except for the classic “Stephen Hawking” voices, they all sound realistic and convincing. I’m not sure what comes next(!)

Leave a comment

Login

Allowed HTML <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre> Markdown Extra syntax via https://michelf.ca/projects/php-markdown/extra/

Sidebar photo of Bruce Schneier by Joe MacInnis.