Can AI Be Trusted in real time? The 2024 US Elections Put Elon Musk's Grok to the Test – and It Flopped

On the night of the US election results, two AI startups Grok and Perplexity tried to present their chatbots as credible and timely information providers using practically the highest stakes situation – the presidential election. However, Elon Musk’s highly anticipated Grok, built by xAI, failed spectacularly and produced wrong results concerning elections before the poll stations were shut.

If Grok's failure was merely a bug, then the incident revealed a deep-seated flaw in xAI’s product: its inability to respond to high-stakes, live updates thus eradicating any confidence users could have had in its true efficacy.

While the other AI startup Perplexity, owned by Aravind Srinivas , was able to provide reasonably precise information with relevant links as well as context, Grok made a pretty pathetic showing, which only served to fuel Skepticism regarding Musk’s chatbot plans. This incident made one thing painfully clear: While Musk’s assertions sound like reasonable visions of the future of AI, Grok is still a far cry from being reliable for handling important or time-sensitive data, which casts doubts on what xAI is building.

Grok's Missteps: Inaccurate Election Results

Grok, built by Elon Musk’s company, xAI, was supposed to offer live midterm election coverage by analyzing content from the social networking platform X (previously Twitter). However, the errors were observed at the initial stages, and Grok was, in one case, providing the wrong winners before the poll ended.

For instance, Grok, when asked about the election results of Ohio, made the wrong impression of Donald Trump as having won the election even when people were still voting. This early mistake is a clear indication of the hazard of incorporating AI systems that may produce misleading or incorrect real-time data.

The discrepancies were by no means random events. For example, some of Grok’s responses would show the topics that are more popular on X, which had a certain bias in favour of specific narratives, potentially because of the users of the platform. This reliance of President Trump on social media trends without substantiation resulted in the spread of misinformation, which resulted in questioning the reliability of the said platform during significant events.

Perplexity AI's Success: Accurate Results

Perplexity AI launched a special feature called the "Election Information Hub” that allows users to have clean and up-to-date information regarding the elections. The hub is comprised of interactive maps, brief summaries of what each candidate offers, and real-time election results.

To ensure that the information that was collected was accurate, Perplexity sourced theirs from reputable sources such as The Associated Press and Democracy Works.

However, confusion analysis was not limited to data display only at Perplexity. The platform offered historical perspective and information on the voting and elections in America, as well as information on the voting maps of the US, the UIScreen, votes by state, and pollsters' forecast and reality, helping in explaining the voting trends, the strategic states, and the pitfalls of polling. 

This full-service approach not only educated users but also enabled them to critically analyze the real-time election results. This was not just presenting the data to the users. There was also an educational component where people could get more context about voting: how it works, major states, the limitations of polls, etc. In addition to this, this balanced and effective approach did not only educate the users but enabled them to analyze the election results as they occurred.

Implications for AI in Real-Time Information Dissemination

The different performances of Grok and Perplexity during the election underscore several key considerations for AI applications in real-time information dissemination:

• Data Source Reliability: AI systems cannot afford to base their data on unadulterated sources since faulty information may be disastrous during an event.
• Contextual Understanding: Imputing historical nuggets and educational value to information makes it valuable in order to make good decisions.
• Avoiding Overreliance on Social Media Trends: A particular aspect of social media is its ability to support real-time data collection; however, the discoverable data is often skewed and unreliable. To avoid such risks, AI platforms should encompass sound verification systems.
• Transparency and Accountability: Each AI application developer must maintain an open description of their source of data and the method they use in producing their results so that users of the application can gauge the reliability of the data presented.

Looking Ahead: The Role of AI in Future Elections

The experiences of Grok and Perplexity during the 2024 election provide valuable lessons for the future integration of AI in electoral processes:

• Enhanced Verification Protocols: Extreme measures in verification help to minimize the spreading of early and wrong information.
• User Education: This outlook can be encouraged by giving the users more information about how AI works and what it is capable of, besides alerting them once in a while that it just can’t do everything.
• Collaborative Efforts: Joint efforts of AI developers, the agencies responsible for electoral processes, and media outlets will create new opportunities for the enhancement of AI-based information services.

I have no doubt that as AI develops further, so will its impact on changing discourse and information dissemination in public. The 2024 U.S. presidential election is used as a case analysis to examine the strengths and weaknesses of AI as a tool in real-time, highly charged settings. In these cases, it is possible for stakeholders to devise proper ways and means for the acquisition of drastic AI without compromising the sort of important data in sensitive events.