With the arrival of ChatGPT, AI chatbots have exploded in popularity. While ChatGPT is getting a leg up by integrating with Microsoft’s Bing browser, other companies are creating their own competitor. Google is working on Google Bard, another AI chatbot powered by Google Language Model for Dialogue Applications (LaMDA). It is currently in beta testing, but uses a lightweight model of LaMDA that can be easily scaled to many users. In this guide, we'll be focusing on Google Sparrow vs Bard and comparing the difference between the two.
A Google-owned company, DeepMind, is also working on their own AI chatbot, Sparrow. Sparrow was introduced to the world in September 2022, demonstrating its ability to learn and draw information from Google’s searches. Both Bard and Sparrow are going to be Google’s entries into the AI chatbot field. Are there any differences in the two, and what sets them apart?
You'll find everything you need to know about Google Sparrow vs Bard in our detailed comparison below.
What does Google Bard have?
Google Bard is powered by the Langauge Model for Dialogue Applications technology, which is a conversational AI service. By asking Bard questions, you get answers which draw upon Google’s knowledge from the Internet. The LaMDA AI has considerable capabilities, to the point where it was deemed sentient by one of the developers. There isn’t much documentation around Bard or any details regarding how it has been refined. However, there is documentation on how LaMDA works, by looking at Transformer, the neural network architecture LaMDA is built with.
As it is currently in beta testing, the full extent of its features aren’t fully known. But it is capable of answering questions and having a conversation like ChatGPT. It will also be released on a lightweight LaMDA model which has less computing power. This eases the requirements necessary to run Google Bard, allowing more users to have access to it.
Until Bard moves past beta testing or is given to more users for testing, it’s difficult to know what Bard is capable of. It’s come a long way since its first presentation, and it must be able to compete with ChatGPT. But without any more information, the full capabilities of Bard remain unknown, along with its inner workings.
What are Google Sparrow’s capabilities?
Google Sparrow is designed to be more helpful, correct, and harmless compared to other language model AI chatbots. DeepMind released a whitepaper that goes over most of the features that Google Sparrow will contain. It gives a better perspective of the functions that Sparrow has, while Bard doesn’t have any documentation. The documentation does mention that Sparrow does have similarities to the LaMDA AI model, which means functionality is largely the same.
Sparrow will be learning from responses just like Bard, but it also has a number of tests behind the responses. Testing has also been done to try and “break” rules within Sparrow, trying to get Sparrow to do something considered unhelpful or adversarial. Sparrow will also reference its information, letting you know where the information comes from. This allows you to follow up on your response, and provide an avenue for further study if necessary.
Instead of just generating text, Sparrow can also perform language translation, summarise texts, and respond to questions. It can perform all of these tasks at once, which is something that ChatGPT is unable to do.
Apart from the testing and behind-the-scenes work that goes on behind Sparrow, it won’t do anything fundamentally different from Bard. As the whitepaper shows, there’s been significant testing to ensure Sparrow will return proper answers after some feedback from users.
Similar to Bard, Sparrow hasn’t been released to the public yet. It has gone through some tests and is considered a worthy competitor to ChatGPT. But apart from the announcement that it could come out later this year, there’s not much information about it. The whitepaper remains the best way to learn about Sparrow’s inner workings, how it gathers information, and how it responds to users.
Are there differences between Bard and Sparrow?
The biggest differences appear to be the way Sparrow collects data and learns compared to Bard. It’s premature to compare the two since there isn’t any documentation on Bard’s learning or widespread testing. Sparrow is also influenced by LaMDA, which is the same technology that Bard is using for its chatbot. The two chatbots are not the exact same, but even being designed by different companies doesn’t create a lot of differences.
Bard isn’t known for providing references to the information it presents to its users, while Sparrow does. But other than that, both chatbots will present data the same way and learn through feedback. As both chatbots gain information, they will be able to understand user responses and refine their responses accordingly.
Sparrow has been touted as the “ChatGPT killer” or “the more grownup version of ChatGPT” for its ability to reference sources. That’s something neither Bard or ChatGPT do, but Bard could include that in the future. As it is also being made by Google, it wouldn’t be strange for Bard to copy Sparrow’s referencing capabilities.
Bard is still undergoing beta testing, with more functions potentially coming as feedback comes in. While copying Sparrow entirely would not be the smartest idea, they are both chatbots being created by Google. Having similar capabilities with slight differences would not be unheard of, especially as their usage data starts coming in.
Sparrow also hasn’t been released yet, though its whitepaper gives a good insight into how it works. It’s too early to compare Bard and Sparrow given that both are still in their testing phases. While initial responses have been positive, the feedback from being released to the public remains to be seen. ChatGPT is also continuing to make advances, which both Google and DeepMind will track as they continue testing. It’s hard to say what the final product will look like, but a glimpse might be seen later this year.
The differences between Bard and Sparrow will become more apparent once both of them have been released. Proper comparisons between the two can be made, and other apps will begin integrating them for certain tasks.