AI Lacks the Human Touch Needed in Historical Analysis for Now
=========================================================================================
In the realm of historical research, the quest for factual accuracy is paramount. A dedicated researcher has been delving into the movie-watching habits of US presidents since 2012, uncovering fascinating details that have been hidden for decades. This endeavour has been aided by the rise of artificial intelligence (AI) tools, which have proven to be both a boon and a challenge.
CEO Sam Altman's ambitious promise for GPT-5, the AI model that powers ChatGPT, was to offer "a legitimate PhD-level expert in anything, any area you need, on demand." However, as the researcher's findings demonstrate, the reality is somewhat different. While GPT-5 can provide a wealth of information, it sometimes falls short on accuracy.
For instance, ChatGPT claimed that The Birth of a Nation was the first movie ever screened at the White House. In reality, this isn't true - Woodrow Wilson watched The Crisis (1916) on March 6, 1917, a film that was ordered from the Library of Congress and uploaded to YouTube and the Internet Archive because it wasn't previously available.
The researcher's findings also debunked several other inaccuracies. President George H.W. Bush did not watch Batman on August 8, 1989, as claimed by Copilot's Quick Response. Instead, he watched the 1942 World War II classic Mrs. Miniver on that day. Similarly, Ronald Reagan watched The Lion in Winter at Camp David on June 15, 1985, not Rambo: First Blood Part II, as suggested by Copilot.
However, it's not just ChatGPT that struggles with accuracy. Other major AI chatbots like Google Gemini, Microsoft Copilot, Perplexity, and xAI's Grok have also shown flaws. For example, Gemini claimed Clinton watched The 13th Warrior at Camp David, but the source was an email forward about weekend box office receipts from the Christian Science Monitor.
Grok, on the other hand, managed to deliver the correct answer after clicking "think harder," but its sourcing was opaque, making it difficult for anyone wanting to find a reputable citation. Copilot's Deep Research, while delivering the correct answer for George W. Bush watching Twin Towers on September 10, 2003, also claimed DC 9/11: Time of Crisis was a possible option, which is incorrect.
Despite these shortcomings, the researcher acknowledges the value these tools bring to their work. They have used AI to verify military historical events, such as the activities of the 2nd Battalion of the Canadian Army on August 8, 1941, and to enrich architectural and cultural details for historical fiction writing, like visual and descriptive details about a 19th-century French farmhouse.
As the field of AI continues to evolve, it's likely that these tools will become more accurate and reliable. Until then, researchers must approach them with a critical eye, verifying their findings and cross-referencing their information to ensure the historical record remains untainted.
- The researcher has utilized social media platforms to share their findings about the inaccuracies in AI-generated information regarding the movie-watching habits of US presidents.
- Despite the occasional inaccuracies, AI tools like artificial intelligence (AI) have been instrumental in the researcher's exploration of health-and-wellness practices among past presidents.
- In a recent interview, the researcher emphasized the importance of cultural sensitivity when using AI, mentioning that the lack of awareness in this area contributed to some of the inaccuracies found in their research.
- The researcher is not only analyzing the movie-watching habits of US presidents but also delving into their dietary preferences to provide insight into the evolution of health-and-mental-health practices in the White House.
- As the researcher continues to explore the limits and possibilities of AI, they are actively participating in discussions about its ethical implications for social media, culture, and the future of scientific research.
