|
|
Get paid while your AI thinks. What could go wrong?
|
|
|
AI coding tools can spend up to an hour processing a request. A new service called Kickbacks.ai promises to turn that dead time into money by replacing the usual “thinking” spinner with an ad, then splitting the revenue with the developer whose machine displayed it. Sounds fair — until you read the fine print.
The thing is, Kickbacks is not the first company to pay people for their data. And Big Tech has profited from ours for years without ever giving us a cut, so a deal like this can feel almost fair. But once you look past the marketing pitch, the trade-off starts to look far riskier and harder to justify.
|
|
|
|
New on the blog
|
AI glasses failed before. Did Meta learn the lesson?
|
|
Meta’s latest smart glasses look almost like ordinary eyewear, making it harder to know whether a private moment is being recorded. Google Glass and Snap Spectacles stumbled over a similar privacy problem back in the day. Meta may have made the technology easier to wear, but not easier to trust.
|
|
|
|
|
|
|
What AI does with your data
|
Your personal data may be training AI
|
|
AI image tools learn from enormous datasets scraped from the web. These can include personal photos, medical images, and copyrighted work, while removing that material after it enters a training set is far from simple.
|
|
|
|
|
|
The ChatGPT mistake Samsung couldn’t undo
|
|
Three Samsung employees reportedly entered source code, testing information, and internal meeting notes into ChatGPT. The incident showed how one convenient prompt can send confidential company data beyond its walls.
|
|
|
|
|
|
Data brokers are getting an AI upgrade
|
|
Advertising firm Publicis claims its identity data can help clients reach 91% of internet-connected adults. Add AI to billions of profiles, and targeted advertising may become even more precise while remaining largely invisible to the people being profiled.
|
|
|
|
|
|
|
What caught our eye
|
|
|
AdGuard news
|
|
|
|
