Bard VS Chat GPT? What do you think?

[Charank]

i've seen some folks who copy and paste from it, without going thru the common sense review or noting the redundancy.

Yup, I have used these tools to understand the context of how certain functions in a tool or process work. And rather than just copying and pasting them hoping that whatever it says is right, I was able to reproduce those issues/ results with guidance. But rather than just being at ChatGPT and Bard websites, the integrations with various platforms like Google Collab where Bard clearly serves its purpose, it clearly explains why a code doesn't work and what could be done to get the code to work teaches small nuances you might have missed. Which is a bigger learning curve than I have done in during my college days. There are Huge use cases for both these models, some are well-trained for certain purposes. Even Perplexity is good to the extent that I can get the source to validate the information from the prompt results. It all comes down to the individual and how they use them to get the work done.

 

[Helmuts]

Yup, I have used these tools to understand the context of how certain functions in a tool or process work. And rather than just copying and pasting them hoping that whatever it says is right, I was able to reproduce those issues/ results with guidance. But rather than just being at ChatGPT and Bard websites, the integrations with various platforms like Google Collab where Bard clearly servers its purpose, it clearly clearly explains why a code doesn't work and what could be done to get the code to work teaches small nuances you might have missed. Which is a bigger learning curve than I have done in during my college days. There are Huge use cases for both these models, some are well-trained for certain purposes. Even Perplexity is good to the extent that I can get the source to validate the information from the prompt results. It all comes down to the individual and how they use them to get the work done.

have to admit this is the first tie I've heard of collab.. checking it right now at https://colab.research.google.com/ .. do you use it yourself? .. and how?

 

[Biggie]

It all comes down to the individual and how they use them to get the work done.

Hi

so.... is AI creating more work for humans to do, than they'd have without it?

right now, it seems to be the big pacifier for the populace to suck on, as they feed the infant AI brain with queries, questions and what if's.

using one AI to check another AI and compare that with a third AI

i still haven't willingly given AI any of my thoughts yet.


imo...

 

[Charank]

have to admit this is the first tie I've heard of collab.. checking it right now at https://colab.research.google.com/ .. do you use it yourself? .. and how?

Anybody can use it, you or your team, right now I use it some of my teammates use it, but for more computation, we might need to pay, especially if GPU is involved. Check the Colab AI on the top, near connect if you are running a Python code. Any syntax errors or if there have been changes on libraries colab AI explains it well and asks you to use the updated one. Instead of going to Stackoverflow and opening 100 tabs, I find Colab AI easier to find a solution and not get stuck for hours.

 

[Charank]

so.... is AI creating more work for humans to do, than they'd have without it?

It depends on the use case, for me it shortens my time spent on different websites to find a solution, and keeps me more thorough. For example SCISPACE (typeset(dot)io), I can read so many research papers, Literature reviews, and even previous patents ask the right questions, and find the formulas used.

right now, it seems to be the big pacifier for the populace to suck on, as they feed the infant AI brain with queries, questions and what if's.

They Train an advanced AI model with an existing AI model. OpenAI exactly does use AI to train another AI. They just develop an AI for training other AI no wonder their releases are very frequent. Nvidia used AI to develop their design faster, so they were able to ship it faster these days.

 

[Helmuts]

so.... is AI creating more work for humans to do, than they'd have without it?

nice thought

 

[Charank]

Anybody can use it, you or your team, right I use it some of my teammates use it, but for more computation, we might need to pay, especially if GPU is involved. Check the Colab AI on the top, near connect if you are running a Python code. Any syntax errors or if there have been changes on libraries colab AI explains it well and asks you to use the updated one. Instead of going to Stackoverflow and opening 100 tabs, I find Colab AI easier to find a solution and not get stuck for hours.

Something like this

 

[Charank]

using one AI to check another AI and compare that with a third AI

I would encourage you to do that because it's easy for AI models to train that way instead of training from Scrapped data. There are several Competitions to give a suggestion based on output and pick a winner. Kaggle is one of them, they conduct competitions for DataScience and ML community to submit their models and Pay good prize money for a efficient model, but they want it to be open-sourced so Google can use them lol. AI models mostly use Data and it needs extensive training, be it text, image, Audio, or video. For an AI model, all these are Data. Here is just an example(https://chat.lmsys.org/) of how would you interpret the information, both are good AI models (GPT 4 and Claude 2.0 but I prefer Gpt 4 here).

 

[NIFS]

I live in Canada not the US so I'm not able to access it. Maybe with a VPN. What are your experiences with Bard VS Chat gpt?

Choosing between Bard and ChatGPT depends on your specific needs. Bard excels in generating creative content like stories and poems, while ChatGPT specializes in conversation and answering questions comprehensively. Consider the task at hand; if you seek imaginative writing, Bard might be preferred, whereas ChatGPT is excellent for interactive dialogue and information retrieva

 

[namegulf]

After all the controversies surrounding Gemini (previously Bard), today they've introduced and front gated with Human Reviewers. You'll notice this banner inside Gemini interface.



Due to market pressure, they rushed an AI product which is clearly not ready.

With traditional products (minus the bugs) we can with 100% accuracy say what it can and can't do, that cannot be said for AI products and Google is learning the hard way.