The 2024 US presidential election has already begun, and the main contenders for the post are Kamala Harris and Donald Trump. The results of the vote will be announced later, but research using artificial intelligence (AI) has already made a prediction about who may become the next president.
Young businessmen, confident that their artificial intelligence chatbots could more accurately model human behavior, decided to test their model in the 2024 elections. The two students who founded Aaru are using census data and other demographic information to develop thousands of voter avatars using artificial intelligence.
Each avatar has hundreds of personality characteristics, allowing it to effectively imitate real people. The agents, which are bots, obtain data from news and other sources that voters might be interested in, and then ask them questions about preferences, such as who they plan to vote for.
Fink and Koch claim that their tests demonstrate the high accuracy of their system, which is significantly more cost-effective and faster than traditional survey methods. They created representative bots to recreate the electorate in key states, with agents reaching thousands in each, and conducted preference surveys.
The models had access to the news until midnight Sunday, allowing them to learn about rallies for Donald Trump in Pennsylvania and Kamala Harris in Michigan. In several simulations, they estimated Donald Trump’s chances of winning: 73.3% in Arizona, 62.1% in North Carolina and 61.8% in Georgia.
At the same time, Kamala Harris has a 63.3% chance of winning Michigan, 53.4% in Nevada, 52.4% in Pennsylvania and 50.9% in Wisconsin. If the election had been conducted entirely by bots, the voting would have been tight, with Kamala Harris winning by a slim margin.
Earlier, “Cursor” wrote that on November 5, elections will be held in the United States, at which the 47th president of the country will be elected. This race includes Democratic Representative Kamala Harris and Republican Donald Trump. Large media companies traditionally play a key role in covering elections in the United States, relying on their own calculations, exit polls, and various forecasts.
Interview Between Time.news Editor and AI Expert
Setting: A virtual meeting room where Time.news editor, Lisa Harper, and AI expert Dr. Jonathan Reyes are poised to discuss the implications of artificial intelligence in predicting electoral behavior for the upcoming 2024 U.S. presidential election.
Lisa Harper: Thank you for joining us today, Dr. Reyes. The 2024 US presidential election is shaping up to be quite a dynamic event, especially with figures like Kamala Harris and Donald Trump in the mix. It’s fascinating that AI is being utilized to predict voter behavior so early in the race. Can you tell us more about how the Aaru team is approaching this?
Dr. Jonathan Reyes: Absolutely, Lisa! The students behind Aaru are pioneering an innovative approach by constructing detailed voter avatars using census data and other demographic information. Each avatar represents a subset of the electorate, complete with hundreds of personality traits. This detailed modeling allows them to statistically simulate how different segments of the population might vote based on various factors.
Lisa Harper: Right, that sounds incredibly sophisticated. How exactly do these avatars work in practice? What kind of data are they using, and how does it inform their predictions?
Dr. Jonathan Reyes: The team is analyzing a wide range of data—from demographic information like age, gender, and location to social media behavior and historical voting patterns. By creating these avatars, they’re able to simulate interactions and predict voting outcomes based on nuanced personality combinations. Essentially, they’re using a multi-dimensional approach that goes beyond traditional polling methods.
Lisa Harper: That’s where AI really shines—quantifying complexities of human behavior. But given the unpredictability of elections, how reliable do you think these predictions will be?
Dr. Jonathan Reyes: Reliability is the key question. While AI models can significantly enhance our understanding of voter behavior, there are always unforeseen factors in politics—like significant events or debates close to the election. However, what makes this approach compelling is its ability to adapt and refine as new data comes in. The efficacy increases with more context and real-time input.
Lisa Harper: Interesting! So, if the model actually predicts a win for one candidate over another, how might that influence public perception or even voter turnout?
Dr. Jonathan Reyes: Great question! If AI models—especially if widely reported—suggest a clear frontrunner, it could either mobilize voters for the underdog or lead to complacency among the leading candidate’s supporters. Public perception of the “inevitability” of a candidate can drastically shape turnout and campaign strategies, making the predictive power of AI quite impactful—even if the predictions aren’t perfectly accurate.
Lisa Harper: So, it’s a double-edged sword! As AI becomes a more potent tool in politics, do you think we might see ethical concerns arise, perhaps regarding data privacy or manipulation?
Dr. Jonathan Reyes: Absolutely. As with any powerful technology, the potential for misuse is significant. The ethical use of data, ensuring transparency in how data is generated, and the need to avoid bias in these models are all critical concerns. Policymakers will need to establish clear guidelines on how these technologies are implemented in election processes to guard against misuse and maintain public trust.
Lisa Harper: It seems like as we move closer to the election, the interplay between technology and politics will only intensify. Thank you for shedding light on these issues, Dr. Reyes. It will be fascinating to see how AI impacts the 2024 election landscape!
Dr. Jonathan Reyes: Thank you, Lisa! I’m looking forward to watching how it all unfolds. It’s a pivotal moment for both technology and democracy.
[END OF INTERVIEW]