ChatGPT vs Google Bard: What’s the difference and who comes out on top?

In the world of artificial intelligence, two prominent language models have been making waves: ChatGPT and Google Bard. There can be no disputing that these state-of-the-art AI systems have revolutionised the way we interact with technology, and have become a cornerstone in industries spanning construction and manufacturing through education, retail, healthcare, and more.

But when it comes to choosing the best, who truly comes out on top? Vapour CEO, Tim Mercer, explores…

Natural language understanding

ChatGPT, powered by OpenAI, boasts an impressive natural language understanding (NLU) model. As well as being able to comprehend a wide range of questions and prompts, it’s able to respond too — providing detailed and contextually relevant answers. Its vast training data ensures a rich understanding of human language, making it an ideal companion to streamline typically time-intensive manual tasks like research.

Google Bard, a product of Google's extensive research powered by LaMDA, also exhibits strong NLU capabilities. It can engage in creative and dynamic conversations, adapting to user input, and the tool’s training data emphasises storytelling too, allowing it to weave captivating narratives.

Because Google Bard can access the internet in real time, compared to ChatGPT which is trained on previous data sets, you’d presume that the results produced are more timely, and therefore accurate. However, with the tool having faced problems with generating incorrect answers in its unveiling, we can’t yet assume that’s necessarily true.

Context and consistency

Able to retain information from previous prompts, ChatGPT allows for more coherent conversations with its users. This feature is particularly useful for tasks requiring extended dialogue or multi-turn interactions. And generally, the tool is able to mimic human interaction to provide full pieces of content, rather than simple, bite-sized chunks of information.

While Google Bard has made significant strides in context preservation, it may struggle to maintain consistency throughout extended conversations. Its storytelling focus sometimes leads to unexpected twists or changes in narrative direction. While this can be entertaining, it might not be the most suitable choice for users seeking precise and predictable responses, in a business environment.

Training data and bias

Both Chat GPT and Google Bard rely heavily on extensive training data to learn and generate responses. However, the sources and methodologies of data collection differ between the two models, leading to varying levels of potential bias.

OpenAI has made strides in addressing this within Chat GPT. While biases can still arise due to the data it was trained on, efforts have been made to reduce and mitigate them. OpenAI also encourages user feedback to improve the system's performance and address any similar concerns.

Google Bard equally undergoes rigorous testing and evaluation to minimise biases — even implementing strict guidelines and policies to ensure responsible AI usage. However, as with any AI model, potential biases can still exist, and continuous monitoring and improvement is necessary.

Platform features

While both platforms offer a similar service — in which users enter a prompt in an empty search box and receive an AI-generated response, with the ability to access and edit previous chats — there are a few minor differences in feature sets.

• Voice prompts: Google Bard has a unique advantage in that it allows users to give prompts through voice input. This hands-free experience adds to the convenience, making it easy to interact with the AI system without typing. In contrast, ChatGPT currently lacks this specific voice prompt feature, relying primarily on text-based interactions.
• In-line research: A standout feature of Google Bard is the "Google it" button. This nifty option provides users with in-line links to continue their research outside of Bard. It allows seamless access to additional information and resources, enabling us to explore and delve deeper into specific topics. ChatGPT, unfortunately, does not offer this integrated research functionality.
• Drafts for variation: Google Bard introduces the concept of ‘drafts,’ providing users with different variations of responses for each prompt entered. This feature proves valuable when refining the desired output, allowing users to compare and select the most suitable response from the available drafts. In contrast, ChatGPT does not offer this specific draft feature, potentially requiring users to rely on generating multiple prompts to explore variations.
• Prompt history and management: Bard presents a convenient way to view and manage previous prompts through its ‘Bard activity’ feature. Users can easily access a log of their prompts, view details, and delete them if necessary. Additionally, users can choose to disable the storage of their activity, if they wish. ChatGPT, on the other hand, organises prompts in conversations displayed in the sidebar. While users can rename or delete conversations, it may require scrolling back to view specific prompt responses, which can be cumbersome.
• Dark mode and conversation clearing: ChatGPT offers a visually appealing dark mode option, enhancing the user experience by reducing eye strain and improving readability, especially in low-light environments. Additionally, it provides the convenience of clearing all conversations at once, allowing users to start afresh or declutter their chat history. These features are not currently offered by Google Bard.

So, who comes out on top?

It's an exciting time in the realm of AI language models, but in the battle between Chat GPT and Google Bard, there is no clear winner just yet. Both models possess impressive natural language understanding capabilities — with ChatGPT shining in maintaining context and consistency, making it a reliable companion for various tasks. Google Bard, on the other hand, excels in creative storytelling and engaging conversations.

The choice between Chat GPT and Google Bard ultimately depends on your specific needs and preferences. If you require accurate and contextually-rich information, Chat GPT might be your go-to AI. Conversely, if you seek imaginative storytelling or engaging dialogue, Google Bard might be the perfect fit.

Google Bard impresses with its voice prompt capability, integrated in-line research functionality, drafts for response variation, and easy prompt management. On the other hand, ChatGPT provides a focused conversation-based organisation, dark mode for visual comfort, and the ability to clear all conversations conveniently.

What’s your tool of choice? Share your thoughts with us on LinkedIn or Twitter.