Google’s BARD: ChatGPT Copycat or another Revolution in the AI Language Model?

For this article about Google’s BARD, the assistance of ChatGPT was sought to find out where the said tool can be accessed to gain at least get a first-hand experience, and it’s response was: 

“The BARD program (Barrel Roll Detection) is a security tool developed by Google to help detect and prevent barrel roll attacks on websites. BARD is not publicly accessible and is only available to select partners and customers of Google. If you are interested in learning more about BARD or potentially utilizing it for your website, you can contact Google’s sales team for more information.”

For purposes of efficiency however, instead of contacting Google’s sales team, instead performed a simple Google search on how to access BARD. However, it appears that there is still no news on when BARD is going to be publicly accessible. 

For starters, and to those who have already been aware of OpenAI’s ChatGPT, it is the latest rival AI language model from Google to match ChatGPT. And similar to ChatGPT’s capabilities, it is meant to answer all of your questions and help you communicate more effectively by understanding and generating human-like text. Like ChatGPT, it is also based on state-of-the-art machine learning algorithms, which have been trained on a vast amount of text data from the internet. 

But doesn’t it sound like it’s just another ChatGPT rebranded in the form of Google? Well, not exactly.

We can’t promise to give you the full rundown of comparison just yet because access to BARD is still limited to a select group of trusted users, and Google has only provided a small number of public examples of what it can handle (such as helping you plan a friend’s baby shower, comparing Oscar-nominated movies, and suggesting lunch ideas based on what’s in your fridge). 

But what we can speculate for now is that BARD doesn’t seem to have all the features that ChatGPT does, like being able to write research papers or haikus. Since BARD’s primary goal is to improve Google Search, and since Google hasn’t yet disclosed everything it can and can’t do, it may not be ready for tasks that don’t involve improving search results just yet.

Also, the main distinctions between BARD and ChatGPT are the language model employed by the chatbots and the extent of access granted. GPT-3.5 stopped receiving training after the year 2021, as evidenced by the disclaimer “Limited Knowledge of World and Events After 2021” displayed in the chat.openai.com/chat interface, while LaMDA is constantly being trained and may provide more accurate information. 

However, this is no guarantee of LaMDA’s accuracy; on February 8, 2023, Google’s stock price fell by about $100 billion due to a public demonstration error by BARD. When asked “What new discoveries from the James Webb Space Telescope can I tell my 9-year-old about?,” BARD gave an incorrect answer, claiming that the telescope had taken the “very first pictures” of an exoplanet outside our solar system. Note that the Very Large Telescope at the European Southern Observatory captured the first image of an exoplanet in 2003. It wasn’t until last year that the first exoplanet photo taken by the James Webb Space Telescope was made public.

Furthermore, what we believe is also known is that Microsoft-backed OpenAI has a competitor in Google-backed Anthropic, which is also an artificial intelligence startup. Google is working to make it easy and safe for other companies to build upon their best models so that everyone can benefit from AI advancement. Google had recently onboarded developers, creators, and businesses so that they can try out the Generative Language API, which will initially be powered by LaMDA. Google is now watching Microsoft very closely, after the latter announced recently that ChatGPT would be integrated within its Bing search engine.

One thing that we are certain for now is that the introduction of Google BARD marks a pivotal turning point in the development of AI chatbots. Unfortunately, it may be quite some time before we get to witness BARD in action. However, the real showstoppers are the revolutionary strides that BARD and ChatGPT have made in AI. It has the potential to completely revolutionise the way in which we interact with machines if they can learn to understand human speech and produce speech that sounds human. The success of their development demonstrates the utility of deep learning as well as other approaches to machine learning. The significance of achieving this milestone, which paves the way for the creation of more sophisticated AI models in the years to come, cannot be overstated. The increasing interest in the field of AI can only stand to benefit from the healthy competition that is taking place between BARD and ChatGPT. Hang on tight for the exciting times that lie ahead as significant development in the AI ecosystem is anticipated in the coming years, or even quicker!