How Meta and Google held back their tech to recognize unknown people's faces due to privacy worries, opening the door for startups like Clearview AI and PimEyes
New York TimesKashmir Hill
Context & Ripple Effects
Meta and Google’s restraint on identifying unknown people created a market opening that smaller facial-search providers could pursue. That opening sits alongside Clearview AI’s earlier funding despite legal and ethical concerns and its reported expansion of a web-sourced image database to more than 10 billion images.
Clearview AI and PimEyes gain room to market facial-search tools in an area Meta and Google chose not to broadly offer.
Meta and Google avoid directly deploying a particularly privacy-sensitive consumer capability, but relinquish a potential product category to more specialized providers.
Second-order effects
The competitive advantage shifts toward firms willing to assemble and operate large face-image indexes; Clearview had already said it had collected more than 10 billion images from the web.
Privacy, legal, and reputational scrutiny concentrates on the specialist vendors filling the gap, rather than disappearing with Big Tech’s restraint.
Third-order effects
If this division persists, facial recognition could develop as a separated market: major consumer platforms provide general AI infrastructure while narrower vendors commercialize identity-search use cases.
The case tests the scalability of ever-larger facial-photo databases and strengthens the case for governance focused on collection and permitted uses of public images, not only on which companies deploy the models.
The trend: This is one instance of privacy-sensitive AI capabilities migrating from consumer platforms to specialized vendors with different risk tolerances.
So tech companies make major tech advance but bury the tech. But this article never explains why? The most we get is vague “threats” to privacy, which obviously... Companies may suppress innovation, but there's always a reason, e.g. strategic or legal. https://www.nytimes.com/...
I saw demos of reliably identifying people in crowd photos at least 10 years ago, if not more. The technology works well and is very mature now. It's one of the rare cases where I've seen 2 tech companies say “Nah, it's a little creepy” and not release it.
Dark times indeed, as @kashhill documents the latest chapter: glasses that identify strangers to you as you look at them. (Of course, more likely you're the person being looked at.) https://www.nytimes.com/...
The absurd hat-phone, a particularly uncool version of the future, contained a secret tool known only to a small group of employees. What it could do was remarkable. @kashhill https://www.nytimes.com/...
We share a planet. “It didn't frighten me, though I knew it should. It was clear people who own a tool like this will inevitably have power over those who don't. But there was a certain thrill in seeing it work, like a magic trick successfully performed” https://www.nytimes.com/.…
In early 2017, a Facebook engineer sat in a conference room in Menlo Park with a smartphone standing on the brim of his baseball cap, attached w/rubber bands. Everyone was laughing at how absurd it looked. Then the phone started speaking their names aloud. https://www.nytimes.com…