Google's New Chatbot, GP3 vs BART: A Comprehensive Comparison

This article is sponsored by Brilliant

In the field of artificial intelligence, chatbots have become increasingly popular due to their ability to solve problems, code, and answer complex questions. Recently, Google released its own chatbot, GP3, to compete with other chatbots like OpenAI's GPT and Microsoft's DPT. In this article, we will compare GP3 with BART, a similarly designed chatbot, and see how they perform in terms of reasoning, creativity, instruction, understanding, and speed.

Coding and Scripting

To test their coding and scripting abilities, simple coding questions and scripting tasks were assigned to both GP3 and BART. While both were able to solve the coding questions and generate scripts without any errors, the responses from BART were often too long and not concise enough. In contrast, responses from GP3 were more refined and to the point.

However, one area where BART does excel is speed. BART proved to be faster than GP3 in generating the responses. But the tradeoff in speed is that sometimes BART would refuse to generate the script asked, while GP3 was able to provide the response required.

Instructional Understandings

When it comes to understanding instructions, it was found that BART had some difficulty in handling specific roles that allowed for customization of chatbot behavior. GP3, on the other hand, was able to adapt to such changes seamlessly.

For instance, when asked to act as an interviewer for a software engineering position, GP3 was able to start an interview right away while BART could only provide a script of a job interview. Furthermore, when asked to act as a language partner in Spanish, BART was unable to support other languages apart from English.

Creativity and Responses

To test creativity and the ability to generate responses, both GP3 and BART were given tasks that required them to create articles, emails, and blog posts. Although both chatbots performed reasonably well, GP3 responses were better structured, creative, and used fan examples, making them more engaging.

Furthermore, when BART was given a similar task, the post generated was straightforward with no fan examples, whereas GP3 used analogies that made the post more interesting and appealing.

Reasoning Capabilities

Finally, the reasoning abilities of both GP3 and BART were tested through math problems. When assigned simple math problems, both were able to provide correct answers without any issues. But when given more challenging math problems, GP3 performed better than BART, which gave the wrong answer.

Conclusion

Based on the tests, GP3 emerges as the clear winner. While BART can be faster at generating responses, its responses lack structure, and at times it refuses to provide an answer to the questions asked. In contrast, GP3 provides concise and relevant responses while also being more creative and adaptable to changing circumstances.

However, to secure a career in the age of AI, one needs to invest in critical thinking skills. That's where Brilliant comes in. With Brilliant, you can learn foundational and advanced math, computer science, AI, and much more interactively. Sign up for a free 30-day trial of Brilliant's annual premium subscription and develop your analytical thinking skills, which are essential for survival in the age of AI.

In conclusion, while BART was created and released with the competition in mind, it is still inferior to GP3 at present. Whether or not Google's decision to release BART was the right move remains to be seen in the future.