Unveiling the Truth: A Journey Comparing GPT Platforms

Fact-Checking the Machines: My Experience Comparing GPTs Over...

Over the past few months, I have been delving into the realm of artificial intelligence, focusing on prompt engineering, no-code programming, and formal coursework in AI governance, strategy, and shadow AI.

My attention has recently turned to evaluating prominent GPT platforms such as OpenAI's ChatGPT, Anthropic's Claude, Google's Gemini, and Perplexity AI. OpenAI's Task (Beta) feature initially caught my eye, allowing customized prompts for daily newspapers covering various topics like San Francisco Giants baseball, Privacy News, Technology Developments, and U.S. Events. I have been enjoying these updates as personalized podcasts via audio playback throughout the day.

However, during this evaluation, I have encountered significant factual inaccuracies, especially in sports updates. These inaccuracies include outdated player details, fabricated highlights, and errors in standings and performance summaries. For instance, my daughter pointed out a discrepancy between Google Search and Google Gemini regarding Bruce Bochy's age, emphasizing the importance of caution and thorough fact-checking.

Comparison Across Platforms

In response to these discrepancies, I have started running the same prompts across different platforms and have found Perplexity to provide the most reliable factual accuracy.

This experience has been enlightening, shedding light on the strengths, limitations, and future considerations of generative AI technologies.

Additionally, I have explored creating a futuristic cartoon image of the San Francisco Giants game statistics versus the Texas Rangers on 4/26/2025 using the ChatGPT prompt. The visualization was a fascinating exercise in AI creativity and imagination.

As I continue my journey in the AI landscape, I look forward to further exploration and innovation in this dynamic field.