How to stop AI from recognizing your face in selfies

Renogy 10 Amp 12V/24V PWM Negative Ground Solar Charge Controller Compact Design w/LCD Display for AGM, Gel, Flooded and Lithium Battery, Wanderer 10A

(6622)

28% Off $24.99 $17.99 (as of 22:17 GMT +00:00 - More infoProduct prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on [relevant Amazon Site(s), as applicable] at the time of purchase will apply to the purchase of this product.)

【12V/24V Automatically Detect】 Automatically detects the voltage of the battery system (for non-lithium batteries) and adjusts the charging parameters accordingly, ensuring that the battery is charged s... read more

[Upgraded] 30A Solar Charge Controller, 12V/ 24V Solar Panel Regulator with Adjustable LCD Display Dual USB Port Timer Setting PWM Auto Parameter

(2316)

$13.98 (as of 22:18 GMT +00:00 - More infoProduct prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on [relevant Amazon Site(s), as applicable] at the time of purchase will apply to the purchase of this product.)

【UPGRADED SOLAR CONTROLLER】This solar charge controller helps you manage the working of solar panels and batteries in solar systems automatically with the build-in industrial microcontroller. You don't ... read more

Their technique, which turns images into what they call unlearnable examples, efficiently makes an AI overlook your selfies entirely. Images of me scraped from the web (top) are turned into unlearnable examples (bottom) that a facial acknowledgment system will disregard. Because Ma and his colleagues stop an AI from training on images in the very first location, they claim this will not happen with unlearnable examples. “It suddenly somehow became robust to cloaked images that we had created,” she says. Microsoft may have altered its algorithm, or the AI might simply have actually seen so many images from people utilizing Fawkes that it learned to recognize them.

Their approach, which turns images into what they call unlearnable examples, efficiently makes an AI disregard your selfies entirely. Since Ma and his associates stop an AI from training on images in the first place, they claim this will not take place with unlearnable examples. “It unexpectedly in some way became robust to cloaked images that we had generated,” she states.