Google’s Gemini Pro vs. ChatGPT3.5 and ChatGPT4 — a quick test.

Google’s announcement of its sophisticated AI model, Gemini, has the tech world buzzing. Gemini’s multimodal abilities represent a leap in AI’s evolution, promising to empower developers with a tool that understands text, code, audio, and visual information seamlessly. How does this advancement fare in everyday AI applications? Let’s explore by examining a simple acronym generation task.

I challenged Google’s latest AI, along with OpenAI’s ChatGPT-3.5 and 4, to fill an acronym for a hypothetical app. A kind of wordsmithing task that Google Bard is mainly intended for. Bard’s integration of Gemini Pro began with the spot-on “Dalle ArtWork Generator” but then strayed from the task. ChatGPT models stayed on course, showing the nuances of how each AI tackles problem-solving. Overall, Gemini Pro seemed weaker than even GPT3.5, which it seems to position itself as an alternative to.

The experiment was simple yet revealing, demonstrating Gemini’s potential, reflected in its first suggestion that stood out for its appropriateness and directness. As Gemini gets integrated into Google’s suite of products, from Bard to the Pixel 8 Pro’s on-device tasks, it’ll be intriguing to see how its advanced reasoning capabilities enhance our digital interactions​​​​.

Google Bard’s initial suggestion using Gemini Pro starts strong, but a deviation from the task underscores the challenges and complexities of AI-guided creative tasks.

OpenAI’s ChatGPT 3.5 showcases its grasp of the acronym generation task, offering creative and contextually relevant suggestions that better align with the user’s request

ChatGPT 4 takes the lead with a nuanced and logical understanding of user instructions, delivering a wide array of acronym possibilities that stay true to the task’s parameters.

As Gemini begins to percolate through Google’s ecosystem, from Search to smartphone features, its impact on AI-enhanced products is keenly anticipated. This acronym test was a small but telling showcase of the sophisticated reasoning we can expect from Gemini and other AI systems.