From election deepfakes to fake images of real-world conflicts, AI misinformation is spreading. If it wasn’t obvious before, now everybody…

View in browser 

AI'm Informed

29.11.2023

From election deepfakes to fake images of real-world conflicts, AI misinformation is spreading. If it wasn’t obvious before, now everybody should be aware that we can no longer trust everything we see online.  

    • Argentina might be the biggest example of AI-influenced elections yet, where it was used both for standard promotional materials, as well as creating controversial deepfakes and disinformation content. - NY Times

     

    • The Sports Illustrated website allegedly used AI-generated authors and content, creating fake profiles with fake headshots and biographies, without disclosing their use of AI – the content was removed after inquiry, and the company cut ties with third-party company AdVon Commerce that was allegedly responsible for it. - Futurism

     

    • AI-generated images of Israel-Gaza and Ukraine-Russia conflicts on Adobe Stock and other sites – they were not clearly labeled until recently, and now Adobe announced policy changes to crack down on AI-generated images of real, newsworthy events, whereas some providers like Getty Images have long banned AI-generated images from their library. - Washington Post

     

    • Many on X (Twitter) have been impressed with a promo of new generative video service, Pika – there’s a waitlist for their web app but you can use it through Discord right now, and when I checked it out it honestly wasn’t that impressive, and definitely not as good as their promo would make it seem. - X(Twitter), Pika

     

    • A 32-bit microcontroller that can handle heavy machine learning workloads on battery-powered devices – with security features that make them suitable for edge AI-enabled IoT devices, it’s a look at the future of scalable AI solutions without the need for cloud processing. - AAC

     

    • The US, UK, and 12 other major countries signed an agreement on AI safety – it’s non-binding and basically just a set of recommendations for companies to design and build systems with protection of the public interest in mind, but it’s still a step towards better AI regulations. - Reuters

     

    • New open-source model detects methane plumes in data from hyperspectral satellites, with an 81% + accuracy – it could help identify “super emitters” of methane and reduce greenhouse gas emissions. - Oxford News

     

    • I’ve mentioned tldraw before, but they keep shipping new AI features – the new AI drawing feature, Draw Fast, is just a blast to play around with (and this time you don’t need your OpenAI API key, it just works). - X(Twitter), tldraw

     

    • Another cool example of livening up drawings with AI – Livesketch is a method to add motion to a simple sketch with a prompt that describes the motion you want, outputting a short, easily editable animation in vector representation. - Livesketch

     

    Thanks for reading!

    I'd like to reach more people with this newsletter, because I believe it's important for all of us to stay informed on where AI is headed. If you'd like to support my mission, share this newsletter on LinkedIn.

    kuba filipowski
    Kuba Filipowski
    CEO and Co-founder at Netguru
    LinkedIn
    netguru-sign-rgb-01

    Netguru S.A., Małe Garbary 9, Poznań, Polska 61-740, Poland

    Unsubscribe Manage preferences