In this article, I will share my first impressions of Bard, the new chat AI from Google. Bard is a large language model(LLM) based on the Google LaMDA model.
I have criticised Microsoft for forcing me to use Microsoft Edge to use their chatbot. I experienced none of that nonsense with Bard, and to be fair to Google, I didn’t expect my browser choice to be an issue. You can read this criticism in my first impression of Bing chat.
Bard, why don’t you introduce yourself?
I started by asking Bard to introduce itself, and the chatbot provided three possible responses. I would imagine that Google and the AI model are using the selection of our preferred responses to refine Bard further.
The following two paragraphs were Bard’s response to that question, and I selected the one from a choice of three which I thought was the best response.
“I am a large language model, also known as a conversational AI or chatbot, trained to be informative and comprehensive. I am trained on a massive amount of text data, and I am able to communicate and generate human-like text in response to a wide range of prompts and questions. For example, I can provide summaries of factual topics or create stories.”
“I am still under development, and I am always learning new things. I am excited to see what the future holds for me, and I hope to be able to help people in many different ways.”
Bard, can you lay out a blog post about yourself?
Google, Microsoft, and Open AI have positioned their chatbots to help creatives. As a content creator and writer, I asked it to create a layout for this blog post.
I’ve decided to change its order, as it seems to me to share my experiences of using Bard before sharing my impressions. Still, other than that, I will at least use it as the blueprint for my post, just as I use any layouts I write. But the writing process might take me away from what was initially planned.
The Boolean Algebra challenge
To give us a base level to compare it with Open AI chat GPT and Bing chat, I asked Bard to explain to me what Boolean algebra is
It seemed to be a concise answer, and it all appears to be factually correct.
My first impression of Bard
Bard is nowhere near being a finished application and, in my mind, is little more than a proof of concept. However, after saying that, it is an impressive piece of concept technology that will likely develop very quickly.
Currently, you have to be wary of what Bard is telling you is the truth as, like all large language models Bard will make things up and tell you convincing lies due to it being good with language. This is why I think it is helpful that Google includes a link to the Google search that Bard used to help it generate its answers.
After saying all of this, I will keep using Bard, at least in the short term, to help me lay out my blog content, as it has come up with some ideas that I hadn’t even considered during testing, which I think will improve the quality of my content.
If you are interested in AI, you should look at my review of Manna, a science fiction book written to help us consider the world we want to live in and why I think it is a book all of us should read.