Clearview CEO doubles down, claims biz has now scraped over ten billion social media selfies for surveillance Clearview AI says it has scraped more than 10 billion photographs from people’s public social media accounts for its controversial facial-recognition tool. The startup’s CEO Hoan Ton-That also told Wired his engineers were working on new features to make blurry images sharper and to make it possible to recognize people even if …
This very much depends on the definintion of "abused" they are using. If the system is used as it was designed it's not being "abused", the fact that the system itself could be termed "abusive" means it is working exactly as designed.
Also 10b selfies, only operating in the USA (pop 300m) so assuming 25% of that 300m post selfies that's means on average they must have posted over 100 each which seems unlikely so I'm suggesting they lied about the "only USA" bit.
It seems like an extremely valuable data set for China and other actors. If you could train AI to recognise US or western citizens only, it could be used from simple identification to training autonomous drones to "neutralise" targets without affecting native population.
> Also 10b selfies, only operating in the USA (pop 300m)
I guess that has a simple explanation in that there are several "social media", and/or they scraped all pictures posted in those accounts (not just account portraits), in which case 10 billion is rather modest.
That been said, the "USA only" bit doesn't sound very credible, the big money is elsewhere. Not that the USA doesn't like to keep a tab on the peons, but other countries have more pressing needs and pay better. Besides they admitted they were working in the UK which AFAIK isn't in the USA, so their geography seems quite fuzzy.
"Besides they admitted they were working in the UK which AFAIK isn't in the USA, so their geography seems quite fuzzy."
Yes, but they did then go on to specify they don't have any EU customers. UK isn't in the EU so, of course, specifying anything "not-USA" after saying they operate "only in the USA" is either superfluous or contradictory.
That means they only have business in the USA, not that they scraped only images of US citizens. That is useful to shield them from a GDPR fine that would probably come as soon as they try to have a foot in EU.
UK is outside EU now, so it's up to ICO to do something - and looking at how the UK government is going to neuter it, ClearView could soon come to a camera next to you, Britons....
> US Bill of Rights for humans against AI
This is definitely a nice thing to hear! It means there is a general realization that "computer says" doesn't work like in the old SciFi shows, computers aren't infallible, quite the contrary, and since they increasingly govern peoples' lives something has to be done to prevent the worst.
Now obviously this being America the opposition will be hell-bent to avoid this from becoming reality. And since quick and sloppy work is way cheaper to do than a meticulous, bug-free one, AI companies will certainly help them scuttle this... *sigh*
Ah but Star Trek TOS has already explored that subject, and did a rather good job of it.
Good luck getting my ugly mug, my Failbook profile has a chimp scratching it's head as my profile pic. Which is usually what I'm doing when a user comes up to me with one of their "why would you do that" issues they have caused.
I only have the account for keeping in touch with my sisters family overseas.
Aye, there's the rub.
The US Constitution, the UK unwritten but allegedly valid constitution, and other Bills, Declarations, Offerings and Scripts all offer one thing:
Defence against the Government.
But Google, Facebook, Amazon, Alibaba, Twitter, and their likenesses are not Government but are trans-national corporate tax-minimizing sub-national "actors" with seriously risk-laden power. And only law, not unbreakable "Constitutions" can limit their power.
Good luck with that.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
