Recently, Chatgpt has been the talk of the town, with its uniqueness, accuracy, and incredible speed it was impressing upon the technological world but sending shockwaves in the Google headquarters because google is the current leader in the search engine market and makes most of its revenue through ads and since the introduction Chatgpt, many people have reported their declining use of google for the solution of their problems. Therefore Google as a response to Chatgpt introduced its google bard which is trained on Lambda, however, failed to impress people during its first presentation. Finally, Google released a beta version of bard, right after GPT-4 came out, therefore it is fair for us to compare the two platforms since GPT has already proved its mettle and google has all the data in the world, unlimited computing power, and the best talent to produce a chatbot which can be much more powerful and accurate. Our comparison is divided into 4 phases :
CODE
We prompted both GPT-4 and google BARD to program a todo list (or app, which is a very simple app that is usually programmed by beginners or asked to be coded in coding interviews mainly – it was intended to line up our daily tasks ) for us:
Bard amazingly got the entire job done in a second while GPT 4 took some time. The only drawback was that BARD’s response was fast but inaccurate.
RESPONSE – BARD
Problems in the code
- It was trying to add the event listener to an element that does not exist.
On prompting bard to correct the error, bard fixes the error using a conditional statement however the todo app still doesn’t work because the element still doesn’t exist.
- In trying to prompt it in the right direction, bard continually fails to solve such a basic problem.
RESPONSE – GPT 4.0
On the other hand, GPT’s response was not only pinpoint accurate, but it was also very well documented, it was clean, and easy to understand.
Round 1 goes to GPT 4
IDEA GENERATION (CREATIVE ASPECTS)
Now we asked BARD and GPT to write us an idea for a romance novel to check how creative both can get and how they express their artistic endeavors.
RESPONSE – BARD
Bard’s response was again very fast but it was very simple and cliche – we did not prompt it further to refine it because we are only looking for the first responders from both platforms.
RESPONSE – GPT 4
GPT’s response was extremely creative and at a first glance you would know that it clearly wrote a better story.
Round 2 would also go to GPT – 4, its mesmerizing creativity speaks for itself.
POETRY
Now we prompted GPT AND BARD to write a poem for us and not just any poem, but a poem about coding and programmers in the style of Dr. Suess!
RESPONSE – GPT-4 and Google Bard
Both of them did a very impressive job, lines like
“I do not like this soul destroyer;
I do not like it, sawyer-the-employer!”
Showcase the creative aspect inside bard and on further prompts, it can be modified and improved further!
GPT’s response however still stands out! , its poetry is artful work, no one can tell if a machine can write this!
Lines like
“In the land of code a lot, where the bugs do dwell,
The programmer would strive to make his code excel,”
Just raise the bar high for other AI bots and platforms, round 3 could have been a tie but there is just something special about GPT 4!
Facts and textbook knowledge
Now that we were done with creative aspects and technical knowledge, we tested both bots’ information accuracy and prompted them to tell us “how do brainwaves work”
RESPONSE-Google Bard and GPT-4
Both GPT-4 and google BARD gave us accurate responses, there were absolutely no factual errors in both of them, however, GPT’s response was slightly better and readable, it sounded more interesting and made the information easy to understand
ENDNOTE
It is very clear that GPT 4 stands out in most of the departments and has impressed upon its users once again, however, it is important to note that GPT is charging a premium price for its facility while google bard is completely free for now, also BARD is currently in its beta form, it is in an experimental phase, but still, a product from GOOGLE – who has all the data in the world, unlimited computing power and best talent to produce a chatbot which can be much more powerful and accurate failing against a relatively newer company OPENAI’s GPT 4 was quite shocking.
We are all very optimistic that Google will definitely come up with something better in the coming days and give tougher competition to GPT and its later versions!!
