Google’s Bard seem to be doing better than ChatGPT at Coding

RealDaddy

Repeatedly violating rules
Joined
Jun 30, 2018
Messages
9,011
Reaction score
11,299
It looks like Google has struggled to master the English language side of their Bard model in the race against OpenAI, and has decided to sneak back into it via the coding route… Their flagship chatbot has been updated with code generation, debugging, and explanation tools in over 20 different languages. Given that coding is a huge use-case of ChatGPT, it’ll be a pretty decent move if they can try to capture this part of the AI market.

And from experiments and rumors doing the rounds, Bard is actually much better at coding as well, so they might just manage to succeed. Also, it shows open-source repos as the source when the code is too inspired by them, brownie points for that.

(https://www.theverge.com/2023/4/21/23692517/google-ai-bard-chatbot-code-support-functions-google-sheets)
 
I've found Bard to be no better, no worse than ChatGPT when given coding tasks. For anything else though, ChatGPT wipes the floor with Bard, unless you want to be quoted with Wikipedia excerpts constantly.
 
I haven't seen any evidence that it is better at coding than GPT4
 
I agree with @Steptoe ('Arold!!) - in some ways Bard is better - it is able to complete a script when tokens are exceeded, unlike ChatGPT which finds it very difficult to deal with long scripts. But both come over as scripting simpletons with terrible amnesia.
 
GPT-5 might be in progress as we speak
We never know
 
It looks like Google has struggled to master the English language side of their Bard model in the race against OpenAI, and has decided to sneak back into it via the coding route… Their flagship chatbot has been updated with code generation, debugging, and explanation tools in over 20 different languages. Given that coding is a huge use-case of ChatGPT, it’ll be a pretty decent move if they can try to capture this part of the AI market.

And from experiments and rumors doing the rounds, Bard is actually much better at coding as well, so they might just manage to succeed. Also, it shows open-source repos as the source when the code is too inspired by them, brownie points for that.

(https://www.theverge.com/2023/4/21/23692517/google-ai-bard-chatbot-code-support-functions-google-sheets)
Yes you are right
 
ChatGPT goes in circles while trying to debug code lately. Even BingAI is better in that front.
 
ChatGPT goes in circles while trying to debug code lately. Even BingAI is better in that front.
Is it really usable for programming?
 
As of now, ChatGPT looks good compared to Bard Google AI tool or any other tool. Majorly, ChatGPT is getting familiar with people.
 
Back
Top