3
AI-generated election disinformation will likely be in all places
If latest elections are something to go by, AI-generated election disinformation and deepfakes are going to be an enormous downside as a file variety of folks march to the polls in 2024. We’re already seeing politicians weaponizing these instruments. In Argentina, two presidential candidates created AI-generated photographs and movies of their opponents to assault them. In Slovakia, deepfakes of a liberal pro-European get together chief threatening to boost the value of beer and making jokes about youngster pornography unfold like wildfire through the nation’s elections. And in the US, Donald Trump has cheered on a gaggle that makes use of AI to generate memes with racist and sexist tropes.
While it’s arduous to say how a lot these examples have influenced the outcomes of elections, their proliferation is a worrying pattern. It will grow to be more durable than ever to acknowledge what’s actual on-line. In an already infected and polarized political local weather, this might have extreme penalties.
Just a couple of years in the past making a deepfake would have required superior technical abilities, however generative AI has made it stupidly simple and accessible, and the outputs are trying more and more practical. Even respected sources is likely to be fooled by AI-generated content material. For instance, users-submitted AI-generated photographs purporting to depict the Israel-Gaza disaster have flooded inventory picture marketplaces like Adobe’s.
The coming 12 months will likely be pivotal for these preventing in opposition to the proliferation of such content material. Techniques to trace and mitigate it content material are nonetheless in early days of improvement. Watermarks, comparable to Google DeepMind’s SynthID, are nonetheless principally voluntary and never utterly foolproof. And social media platforms are notoriously sluggish in taking down misinformation. Get prepared for a large real-time experiment in busting AI-generated pretend information.
—Melissa Heikkilä
STEPHANIE ARNETT/MITTR | ISTOCK, ENVATO
4
Robots that multitask
Inspired by a number of the core strategies behind generative AI’s present increase, roboticists are beginning to construct extra general-purpose robots that may do a wider vary of duties.
The previous few years in AI have seen a shift away from utilizing a number of small fashions, every educated to do completely different duties—figuring out photographs, drawing them, captioning them—towards single, monolithic fashions educated to do all these items and extra. By exhibiting OpenAI’s GPT-3 a couple of further examples (often called fine-tuning), researchers can practice it to unravel coding issues, write film scripts, go highschool biology exams, and so forth. Multimodal fashions, like GPT-4 and Google DeepMind’s Gemini, can clear up visible duties in addition to linguistic ones.
The identical strategy can work for robots, so it wouldn’t be crucial to coach one to flip pancakes and one other to open doorways: a one-size-fits-all mannequin may give robots the flexibility to multitask. Several examples of labor in this space emerged in 2023.