25 – Breaking News & Latest Updates 2026
Skip to main content

Emilia David

Emilia David

Former Reporter

Former Reporter

    More From Emilia David

    Emilia David
    Emilia David
    Tesla finally releases (sort of) its neural network Full Self-Driving feature.

    Electrek reports that Tesla is slowly releasing FSD v12 on beta to a very small group of testers after CEO Elon Musk promised its release last year. If the demo was anything to go by, delaying the release to 2024 was a good idea, especially after almost running a red light.

    FSD v12 is supposed to feature what Musk calls “end-to-end neural nets,” meaning that instead of relying on a mix of cameras and sensors like radar, FSD v12 will mainly use AI and cameras.

    Emilia David
    Emilia David
    AI can mimic handwriting because what else hasn’t it tried yet?

    Researchers from Mohamed bin Zayed University of Artificial Intelligence in Abu Dhabi said they successfully developed an AI that imitates handwriting. The model learns from looking at a few paragraphs of handwritten words. The researchers hope to use the technology to help those who can no longer pick up a pen write again but did acknowledge its potential for forgeries.

    Other AI developers have been working on similar technologies. Meta introduced its text replacement model, TextStyleBrush, in 2021, which changes words in images in the same style.

    Emilia David
    Emilia David
    Let people make their own AI chatbots, and they’ll find a way to make AI girlfriends.

    OpenAI’s GPT Store explicitly disallows GPTs “dedicated to fostering romantic companionship,” but a search of the store yields many custom chatbots doing just that.

    First reported by Quartz, GPTs like “Virtual Sweetheart” (with 100+ chats so far), “Your Ex-Girlfriend Jessica” (200+ chats), and “Judy” (with more than 1,000 chats) are racking up a lot of users. Of course, this is nothing new. Both Replika and Character.ai encourage AI companionship through chatbots.

    Emilia David
    Emilia David
    OpenAI is softening its stance on military use.

    The ChatGPT maker already didn’t have an outspoken public stance on the issue, but The Intercept reports that it’s removed a terms of service ban for “military and warfare” use of its tech — though it still bans things like using it for weapons development.

    The US military has long-standing and sometimes fraught dealings with tech giants, and it’s been exploring how to use generative AI. As reported by Bloomberg, the military ran an exercise to test how to use large language models to plan responses to global crises.