The Future. The recent revolution in generative AI has made convincing deepfakes cheap and easy to make. That’s a serious problem for the upcoming presidential election, because more sophisticated misinformation tools could have a larger impact on voters.
Better means worseMisinformation has been used in the past two elections, but improved tech means more convincing lies.
The 2016 and 2020 elections saw misinformation spread over Facebook in the form of text- and image-based posts, but generative AI can create audio and even video with the advent of tools like OpenAI’s Sora.
This January, New Hampshire residents received a robocall that sounded like Joe Biden urging them not to vote in the presidential primary — and the call took only 20 minutes and $1 to make.
Such misinformation will likely run rampant on X now that Elon Musk has seriously curtailed the platform’s content moderation policies.
The wild westMajor lawsuits filed against OpenAI by big media players like The New York Times and Getty Images could shut the industry down completely one day. But such cases are notoriously slow, and in the meantime, improved deepfakes are already here.
Be careful — AI is now so good that you can’t believe your eyes.
The post The 2024 election has a deepfake problem appeared first on TheFutureParty.