The other election night winner: Perplexity

On a pivotal election night, two AI startups, xAI and Perplexity, sought to showcase the capabilities of their AI chatbots in real-time election reporting amidst one of the most significant political events in the United States. Led by Elon Musk, xAI’s Grok faced immediate criticism for providing incorrect information about race outcomes even before the polls closed. In stark contrast, Perplexity emerged as a reliable source, providing real-time insights, detailed maps, and links to reputable resources throughout the night.

Perplexity’s proactive approach paid off after it introduced a dedicated election information hub just before the elections. This hub featured real-time election maps populated with data from trusted sources like Democracy Works and the Associated Press, mirroring Google’s election mapping strategy, which was a notable departure from many AI chatbots that opted to refrain from answering election-related queries due to concerns over misinformation.

The reluctance of most AI companies to engage with the election reflected a cautious strategy born from past experiences with erroneous “hallucinations” in AI outputs. For instance, OpenAI’s ChatGPT Search, which recently launched, was unable to provide dependable answers regarding the election, with the company directing users to Vote.org instead. This underlined an awareness that its AI model, still in an experimental phase, was not yet ready for the scrutiny of a live election scenario.

On the other hand, Perplexity had been testing its Google competitor since December 2022 and believed it had gathered sufficient data to handle the demands of election reporting effectively. Its performance on election night showcased how the startup actively competed with traditional media outlets for audience attention during a moment of high stakes.

Although Perplexity had established agreements with organizations such as Democracy Works and the Associated Press to power its election-related features, it also sourced live coverage from other media outlets like CBS, CNN, and BBC without any clear revenue-sharing arrangements. This sparked concerns about whether Perplexity was in direct competition with these media companies, especially since it garnered significant traffic on election night.

Regarding the technical aspects, Perplexity’s election hub featured visually appealing charts and maps, elements that users found accessible and easy to navigate. The platform displayed a continuously updated electoral map that indicated the status of key races, albeit with occasional bugs, which the company addressed promptly in response to user reports. This real-time updating was essential for maintaining user engagement and trust.

While the charts were rooted in traditional electoral reporting techniques, Perplexity also integrated AI-driven features to supplement its offerings. Users could ask real-time questions about the presidential race, and though the AI’s answers weren’t as polished or insightful as commentators on mainstream media platforms, they generally provided accurate and relevant information with few inaccuracies. Responses rarely missed the mark, showcasing Perplexity’s commitment to maintaining timeliness and correctness.

For example, when users inquired about updates in “Blue Wall” states or the ballot counts in swing states, Perplexity occasionally provided incorrect references, yet overall, it succeeded in giving timely replies, something many competitors struggled with during such a critical event.

As the 2024 election season marked a unique intersection of technology and democracy, it became evident that AI chatbots would play an increasingly significant role in providing election information. While Perplexity’s successful navigation of election night has set it ahead of competitors in the race for reliable election reporting, the ongoing challenge for AI companies will be ensuring accuracy against the backdrop of complex, fast-evolving democratic processes.

Ultimately, Perplexity’s efforts on election night exemplified the potential of AI to offer accurate, real-time updates, suggesting a bright future within the realm of AI-assisted information dissemination during elections. The race is on for AI startups to refine their technologies, learn from experiences such as this, and provide trustworthy services that align with the critical interests of the public and democratic processes.