Google has finally joined the AI race and launched a ChatGPT rival called Bard – an “experimental conversational AI service” earlier this year. Google Bard is an AI-powered chatbot that functions as a poet, a primitive mathematician, and even a decent conversationalist.
The chatbot is similar to ChatGPT in many ways, able to answer complex questions about the universe and give you the ability to dive deep into a range of topics in a conversational, casual way. However, the bot differs from its rival in that it is connected to the network and therefore according to Google gives “fresh, high-quality answers”.
Google Bard is powered by LaMDA (short for Language Model for Dialogue Applications). Like ChatGPT, it is a type of machine learning called “large language model” that has been trained on a huge dataset and is capable of understanding human language as it is written.
Google I/O 2023
Google I/O 2023, an event where the tech giant discusses upcoming features, gadgets and updates across its product line, has revealed a ton of new information about Bard, Google’s artificial intelligence (AI) chatbot.
While we initially found Bard to fall short in features and performance compared to its competitors, Google has announced new tools that will greatly improve the experience and make it much more competitive with other AI chatbots such as ChatGPT and Microsoft Bing.
The I/O event revealed several new features coming now and soon, including Google PaLM 2, improved citations, export capabilities to Google Docs and Gmail, an export tool that will allow users to export Python code using Replit, increased language support to 40 languages, multi-modal functionality, wide range of Adobe integrations and more.
ChatGPT and Microsoft Bing dominate the AI realm, but we hope Bard catches up and hopefully with these additions, Bard can really challenge the big dogs of the AI chatbot world. We hope that Bard’s logical capabilities will improve so that he can match ChatGPT and Bing in coding and logical reasoning – basically being able to arrive at answers or conclusions based on established facts or data. It would also help propel Bard into the big leagues should we see improved search engine integration.
So what exactly will Google’s Bard do for you, and how does it compare to ChatGPT, which Microsoft seems to be building into its own search engine, Bing? Here’s everything you need to know about it.
What is Google Bard?
Like ChatGPT, Bard is a chatbot built on deep learning algorithms called “large language models”, in this case called LaMDA.
To begin with, Bard was released on the “light model” version of LaMDA. Google says this allows the chatbot to scale to more people because this “much smaller model requires significantly less computing power.”
At the moment, Bard is no longer in the testing phase as he has officially left the waiting list. It’s now available in over 180 countries, and users can even join with a Workspace account instead of a personal Google account.
Google has combined external feedback with its own internal testing. This is because, for all their advantages, chatbots have also shown a propensity for negative traits that include everything from bias to powering cyberattacks.
Opening up chatbots for public testing comes with huge benefits that Google is “excited about,” but also risks that explain why the search giant was so cautious about releasing Bard into the wild. ChatGPT’s meteoric rise, however, apparently forced it into action and accelerated the public launch of Bard.
What will Google Bard allow you to do?
Until recently, Google was a bit vague about Bard’s capabilities, but in short, Bard is the next generation of Google search engine development that has the potential to change the way we use search engines and search for information on the web.
Google says Bard could be “an outlet for creativity” or “a launch pad for curiosity, helping explain new discoveries from NASA’s James Webb Space Telescope to a 9-year-old or learn more about the best strikers in football right now, then do the exercises, to develop your skills.”
Unlike traditional Google Search, Bard uses information from the web to help it answer more open-ended questions with impressive depth. For example, instead of standard questions like “how many keys does a piano have?” Bard will be able to give long answers to a more general question like “is it easier to learn piano or guitar”?
One of the biggest updates is PaLM 2, Barda’s language model and a direct response to OpenAI’s GPT-4 model that will be the foundation of Google’s AI products. It’s a huge general LLM with extended reasoning, language, and coding capabilities. Smaller LLMs can and have been developed to support more precise and specialized tools such as Med-PaLM 2 and Sec-PaLM.
There are also enhanced citations to tackle misinformation and plagiarism. Bard will annotate a line of code or text that needs to be quoted, then underline the quoted part and provide a link to the source material. Bard will also allow users to export Python code via Replit in the near future, and will increase language support from the current three to 40 languages.
A huge new feature coming soon is the ability Google Bard to create generative images from text. This feature, a collaboration between Google and Adobe, will be rolled out by the company Content Authenticity Initiativeopen-source Content Credentials technology that provides transparency to the images generated by this integration.
The whole project is made possible by Adobe Firefly, a family of creative generative AI models that will use Bard’s conversational AI service to power text-to-image capabilities. Users can then take these AI-generated images and further edit them in Adobe Express.
Google Bard vs ChatGPT: What’s the Difference?
Basically, the chatbot is based on technology similar to ChatGPT, with even more tools and features that will bridge the gap between Google Bard and ChatGPT.
Both Bard and ChatGPT are chatbots built on “big language models”, or machine learning algorithms, that have a wide range of talents, including text generation, translation, and prompt response based on the massive datasets they’ve been trained on.
Both chatbots, or “experimental conversational AI service” as Google calls Bard, are also being fine-tuned with human interactions to steer them towards desired responses. The potential difference between the two, however, is that ChatGPT is currently not connected to the internet, meaning it has very limited knowledge of facts or events beyond 2021.
Google says Bard, meanwhile, “gets information from the web” to provide up-to-date answers. This may not be the difference between Bard and ChatGPT for a very long time – a leaked preview recently showed Microsoft’s version of Bing with ChatGPT integration, suggesting it won’t be long until we see the latter combined with Microsoft’s search engine.
Does Google Bard only give text answers?
Until recently, Google’s Bard initially only responded to text prompts with its own written responses, much like ChatGPT. But one of the biggest changes to Bard is its multi-modality. This allows the chatbot to answer users’ questions and queries with both text and images.
Users can also do the same and Bard can work with Google Lens to upload images to Bard and Bard responds in text. Multimodal functionality is a feature that has been mentioned for both GPT-4 and Bing Chat and now Google Bard users can take advantage of it. And of course we also have Google Bard’s AI image generator based on Adobe which will be powered by Adobe Firefly.