Home Artificial Intelligence A Practical Comparison of ChatGPT’s and Bard’s Free Versions Test 1: Brainstorm lists of molecules for an academic tool

A Practical Comparison of ChatGPT’s and Bard’s Free Versions Test 1: Brainstorm lists of molecules for an academic tool

0
A Practical Comparison of ChatGPT’s and Bard’s Free Versions
Test 1: Brainstorm lists of molecules for an academic tool

Results from tests devised to match the facility and limitations of those language models which can be just an URL away

Towards Data Science
Photo by Google DeepMind on Unsplash

No wonder I’ve integrated ChatGPT into my day by day work and activities lots. Really lots. It is amazingly helpful to brainstorm, correct or improve texts, do translations, write code and find errors in it, etc. Because of its web nature, ChatGPT is at all times only a URL away, free, with no installs or downloads needed. That feature of OpenAI’s model, key for mass adoption, is simply paralleled by Google’s Bard; nevertheless, because it rolled out later, it’s arguably less popular, at the least around my collaborators and friends. Why not give it a probability?

Here I present a comparison of those two models through tests that I devised based on the extensive experiments that Microsoft presented earlier this 12 months in a really provocative preprint that I discussed recently:

Here’s how this study actually unrolled.

Just a few days back, I needed to brainstorm with ChatGPT concerning the content I’m preparing for our (upcoming, stay tuned!) website for chemistry education using WebXR-based virtual reality, and to my horror I discovered that it was not operational. Well, I said to myself, it is a probability to check out Bard, which can be only a URL away.

After a couple of hours, ChatGPT got here back online. Then, I took the chance to probe it with the very same, or as close as similar questions as possible. I compared free versions only, and devised tests near how I might actually use the tools in my on a regular basis work.

Read on to see my results. Some findings are expected, others trivial, and others quite interesting. And skim the conclusion section for my “verdict.”

To be exact, the issue that gave place to this type of “competition” I ran…

LEAVE A REPLY

Please enter your comment!
Please enter your name here