Excerpt from Search Engine Land
Google has finally launched Bard. Here's a first look at indexing speeds, inconsistencies and search intent on Google's AI chatbot.
As of writing, Bard has been open to a small portion of the public for a full day. I’ve had the opportunity to play around with Google’s new AI chatbot, and I’m getting mixed results.
With Bard’s inconsistent approach to citation, the quality and authenticity of its answers become a potential cause for concern.
Google does not hide the fact the AI chatbot will make mistakes and could potentially share offensive responses.
“Accelerating people’s ideas with generative AI is truly exciting, but it’s still early days, and Bard is an experiment. While Bard has built-in safety controls and clear mechanisms for feedback in line with our AI Principles, be aware that it may display inaccurate information or offensive statements.”
– Bard FAQ, Google
Early findings on Bard
Bard has the potential to be an incredibly powerful and helpful tool.
However, I can't help but think it has been rushed into an early launch to compete with Bing Chat and ChatGPT.
Indexing speeds and inconsistent answers
When I received early access to the tool, I first asked, “Who is Billie Geena?”
It provided a short paragraph about me. It did not cite its sources, but after looking into it further, it appears to have combined information from my blog and LinkedIn profile.
When Bing Chat was released, we saw incredible indexing speeds that ensured its AI chatbot provided the most up-to-date information possible.
We can assume that IndexNow plays a part in this. Brodie Clark did a great experiment on how quickly Bing now finds and indexes information.
I wondered if Bard had something in place to match Bing Chat's impressive indexing speeds. I changed the data sources Bard used to answer "Who is Billie Geena?" and allowed a few hours for the information to update.
Upon searching a second time, I received the following answers.
Bard was no longer able to answer the question. I decided to specify my full name to see if more details would help it find an answer but to no avail.
I am curious about the consistency of Bard's answers and will continue to experiment with the same questions over a period of time.
Unlike ChatGPT, Bard cannot remember previous conversations. Thus, it's more likely that you will see significant changes in answers as each conversation will not have the context of anything prior.
Click here to read complete article at Search Engine Land.