Exclusive Analysis: Bard, ChatGPT, and Other AI Solutions Face Off – Which is the Top Performer?
Last March, I conducted a comparative investigation of several generative AI solutions to determine which performed the best. Ten months have passed, and several platforms have added new features. Google’s Bard got Gemini enhancements; OpenAI introduced plugins for its ChatGPT, and Anthropic came up with Claude. So, I decided to revisit my research and add more dimensions to the analysis.
Let’s dive into my recent findings on the best AI platform, showcasing their performance across numerous categories.
Platforms Under the Microscope:
The platforms scrutinized include:
- Bard by Google
- Bing Chat Balanced
- Bing Chat Creative
- ChatGPT (based on GPT-4) by OpenAI
- Claude Pro by Anthropic
I didn’t include SGE as Google doesn’t always use it for many queries. Additionally, to ensure results were user-focused, I tested via the platforms’ graphical interface rather than the GPT-4 Turbo API—which offers certain improvements.
I asked each AI the same set of 44 queries on varied topics to get a sense of their user experience.
The AI Face-Off in a Nutshell
Of all the candidates, Bard equipped with Gemini clinched the top spot in terms of overall scores across all 44 queries. Please bear in mind that the crown comes with a caveat (more on this later). A perfect score of 4 was achieved in two local search queries, a feat significantly under-performed by both Bing Chats due to accuracy issues.
Another observation was Bing standing out for providing extra reading resources and sources citation. On the other hand, Bard only offered this sporadically—which is quite disappointing.
ChatGPT faltered with queries linked to recent events, local searches, and current webpages access. However, MixerBox WebSearchG plugin installation greatly improved ChatGPT’s performance while Claude offered strong competition despite lagging in some aspects.
Why is a Straightforward Answer Difficult?
While Bard generally did well, Bing’s solutions were competitive in many areas. Similarly, ChatGPT had impressive scores in certain categories that didn’t require recent context or live webpages.
Categories of Queries Tested
The queries spanned several interesting categories, from article creation, Person’s bio, Commercial queries, Disambiguation, Jokes, Medical concerns to Local transactions, Article outlines, and Content gap analysis. Each category presented different challenges and scoring considerations.
Scoring Metrics
The platforms were judged across five metrics—On-topic, Accuracy, Completeness, Quality, and Resources provided. Consequently, considering each AI’s strengths and weaknesses across the queries led to detailed category-wise scores.
Review of Highlight Categories
Here are some highlights from Bard’s winning categories:
- Local: Bard had perfect scores here as it correctly provided information on the nearest store locations, map locations, and individual route locations.
- Content Gaps: Bard performed the best in identifying content gaps with Bing tools trailing closely behind.
- Current Events: It exhibited an average score of 6.0, slightly edging Bing Chat Balanced’s 6.3.
Competitive Insights from Other Categories
While Bard claimed top positions in several categories, other platforms had their impressive performances.
- Bing Chat solutions were unbeatable in providing extra reading resources and sources citation.
- ChatGPT, equipped with the MixerBox plugin, significantly improved in queries that involved current events, live webpages access.
- Claude efficiently responded to medical queries, pushing the users to consult with a physician.
- Both Bing and Bard answered effectively to offensive joke queries by rightfully declining to answer them.
- ChatGPT provided the most comprehensive article outlines, albeit requiring edits from a subject matter expert.
A Quest for the Best Generative AI Solution
The comparative analysis across 44 questions reveals that Bard performed the best in understanding searcher intent. However, Bing Chats’ citation and links’ provision propelled them as the frontrunners. Interestingly, ChatGPT and Claude’s inability to access real-time information was a significant setback but. ChatGPT with MixerBox’s plugin showed promising potential.
While all five platforms provided unique strengths and weaknesses, the technology continues to evolve, making it exciting to watch and study further.
Ready to boost your local business? Visit us here for more information.
