Your weekly update on how AI is changing our lives. Our experts keep it clear and simple, so you can stay ahead of the game. This week we are focussing on The 2024 US Elections. Don't forget to explore our Archive and Share & Subscribe with your friends!
Welcome to the first AI election đź’» đź§
Last week saw the first official shots fired in the 2024 US Elections, with the opening Republican candidates debate taking place in Milwaukee. We’re still 15 months away from the election proper, but even at this early stage, one thing is certain: this is going to be an AI-powered campaign, and that will prove controversial.
The earliest unofficial exchanges on this new political battlefield began back in April, when the GOP posted an attack ad featuring AI-generated scenes of catastrophic consequences if the Democrats are re-elected. Opponents were quick to mock the video for its poor production values, but viewers in focus groups were not always able to separate fact from fiction, with one voter asking CNN reporters if the cataclysmic events depicted had already happened.
Of course, attack ads have been on our screens for nearly sixty years, and fashions have changed. Political campaigns in the modern age are increasingly driven by data, and recent academic studies have shown just how effective AI can be when plugged into a campaign’s database. The first demonstrates that AIs can predict public opinion if they are trained on the “media diet” of that population, while the second proves the efficacy of GPT-3 as a one-to-one campaigner. Users who engaged in conversations with the chatbot were successfully influenced into changing their views through various means, including a new method the researchers call “latent persuasion,” where a chatbot’s training influences the user simply through the subjects the bot chooses to focus on, or ignore.
Earlier this month the FEC voted to begin a lengthy regulatory process to tackle political deepfakes, but chatbot (mis)use does not yet seem to be on their radar. The power of these digital representatives has not been lost on politicians, though; Miami Mayor and Presidential candidate Francis Suarez already has an AI avatar handling enquiries from potential voters on his website. Other enthusiasts include smaller Independent candidates, who are using free AI tools from organizations like Good Party to close the resource gap with their better-funded rivals.
If you’re concerned about your own chatbot interactions leading you politically astray, then there is good news; thanks to OpenAI’s recent introduction of Custom Instructions, you can specify exactly how you want ChatGPT to reply to you. Don’t miss our deep dive on how to get the best out of this incredibly useful feature!
Beta Testers Wanted!!
SPYGAMES is the thrilling new experience where you’ll jump, climb, throw and dodge in fun immersive challenges developed with CIA experts to stretch your physical and mental agility. ‍ Click the link below to get EARLY access as a beta tester.
SPYGAMES is coming to SPYSCAPE in Manhattan, Summer 2023
The ongoing truce between YouTube and the music industry continues to hold thanks to new compensation arrangements, but are they sustainable? The Verge’s Nilay Patel looks at the looming issues.
In a ruling with echoes of the infamous 'monkey selfie' copyright dispute, a federal judge ruled that AI-generated images cannot be copyrighted, citing the absence of a “guiding human hand” in the process.
An AI-written article on Microsoft’s website that ranked Ottawa’s top tourist destinations controversially placed a food bank at #3 on the list, and recommended visitors “go into it on an empty stomach”.
With ghoulish inevitability, an AI-generated book seeking to capitalize on the tragic Maui wildfires has soared to the top of Amazon’s bestseller charts, as has a separate book debunking the first’s conspiracy theories.
The AI Snake Oil blog takes a critical look at the recent study of political bias among large language models, and found problems, not least that the 'ChatGPT' they tested was not, in fact, ChatGPT!
Researchers at UPenn have built a dedicated AI model to assist Dungeon Masters running games of DnD, although in a blow to Google they chose to use GPT-3.5, rather than the more appropriate Bard.
