The invention of artificial intelligence (AI) has ensured the smooth running of many human tasks and activities. It has a wide range of applications in medicine, transportation, banking and finance, gaming, and entertainment. Artificial intelligence gives off a high degree of precision and accuracy, which is significant in carrying out some strenuous tasks like collating and evaluating extensive data.
In this guide, we will discuss one of the rising AI chatbots named Google Bard. Google Bard is an experimental chatbot designed by Google that imitates human conversations. It is powered by technologies based on natural human language and machine language.
Progress In AI Development
Since the 20th century, AI has been undergoing a series of evolutionary developments. The early days of the tech began with Alan Turing, an English logician and mathematician who idealized a machine that was able to detect undecided mathematical propositions. Further developments in artificial intelligence also feature an American computer scientist, John McCarthy, who is referred to as the “father of artificial intelligence”. His notable works include the coinage of the phrase “artificial intelligence”, the development of the family of list programming languages, and partial involvement in the designation of ALGOL.
Moreover, AI’s growth has broken the basis of Moore’s law, which states that the number of transistors on a microchip roughly doubles every two years. The growth rate of AI doubles every six months.
Rapid development in the industry led to creation of various AI chatbots, such as ChatGPT, Microsoft Bing, Google Bard, ChatSonic, JasperChat, YouChat, and Neeva AI, to mention a few.
Release of Google Bard
The first publication regarding Google Bard was released in February 2023. However, Alphabet Inc (NASDAQ: GOOGL), the parent company of Google LLC, failed to give a valid release date for the generative AI. According to the script, Google Bard originates from a Transformer project executed back in 2017. It was a joint project of two Google subsidiary firms: Google Brain (a deep learning AI research team) and Google Research (Google’s academic research publishing arm). This research project featured eight top engineers from the firms: Illia Polosukhin, Ashish Vaswani, Noam Shazeer, Jakob Uszkoreit, Llion Jones, Niki Parmar, Łukasz Kaiser, and Aidan N. Gomez.
The Transformer project served as the basis of several contemporary AI chatbots, including ChatGPT. According to Alphabet team, the AI environment has transitioned into a re-oriented environment in the past six years.
Google Bard developers opened a waitlist on March 21, 2023, for testers in the US and UK. At the annual Google I/O, held in Mountain View, CA, on May 10, 2023, Google discarded the waitlist and proceeded to release Bard. It became available to users in approximately 180 countries. The generative AI chatbot supports three languages: English, Japanese, and Korean. Google plans to make the chatbot compatible with 40 more languages.
Google Bard Explained
The initial version of Google Bard was meant to support the lite version of LaMDA (Language Model for Dialogue Applications). According to the developers, this version allows for lower computing power and increased accessibility. The tech giant expected to receive feedback that would further improve the speed, accuracy, and specificity of Bard. Unfortunately, Google was unable to launch a testing phase. Instead, Bard was subjected to a series of internal testing by developers.
After the tests, Google Bard was prematurely released to the general public. However, the chatbot was fortified with Pathways Language Model 2 (PaLM 2), a more accurate programming language. PaLM 2 helps Google Bard perform some fantastic features which differentiate it from other chatbots. These features include writing code, solving math problems, and drafting write-ups. Notably, Bard gets its sources directly from the web.
Google Bard vs ChatGPT
Many suggest Google Bard’s invention was triggered by the success story of Open AI’s ChatGPT. ChatGPT gained over a million users during the first week of its launch. In fact, it has earned a title of “the most downloaded app of all time”. Notably, both the chatbots are built on Transformer technology. They both also provide extensions that allows other third-party firms to incorporate their chatbots into their website, messaging app, and other applicable software.
However, there are some outstanding differences that make one outshine the other. While ChatGPT declines in efficiency during high user traffic, the infrastructural design of Google Bard allows it to accommodate more users conveniently. ChatGPT responds to tasks using a generative AI that depends on the random generation of text, images, and other media. This type of generative AI is liable to cyberattacks owing to its regular use of pictures and videos. Google Bard, on the other hand, centralizes on providing users with more natural and contextual responses.
Google Bard is free to use, with no declared intention to monetize the chatbot. Meanwhile, there is a free version of ChatGPT and the paid version, known as ChatGPT Plus, that costs $20/month. The subscription guarantees availability during a high traffic volume, faster responses, and access to new features. Bard can provide more recent answers in conversations, but ChatGPT seems to give information that is not relevant to the time of a request. Surprisingly, ChatGPT has tools that allow a teacher to detect its usage in students’ academic scripts, while Google Bard fails to do that.
Other alternative chatbots apart from Google Bard and ChatGPT include the following:
- Microsoft Bing. Microsoft partnered with Open AI to introduce a plug-in from Chat GPT-4 to its traditional Bing search.
- ChatSonic – an AI chatbot with the ability to create texts and images based on trending issues. It was designed by WriteSonic and advertised as “the ChatGPT with superpowers”.
- YouChat. This is the integration of ChatGPT into the You.com search engine. However, it has the advantage of providing users with real-time data, unlike ChatGPT.
Google Bard Controversy
Google Bard has been attacked for its lack of accuracy in responding to questions. During the demo testing at Google I/O annual conference, the chatbot made a costly error while responding to a user’s question. The user asked, “What new discoveries from the James Webb Space Telescope can I tell my 9-year-old about?” In the response, Bard noted the telescope “took the very first pictures of a planet outside of our own solar system”. Astronomers debunked the fact that an earthbound observatory took the first picture in 2004. The tech company lost approximately $100 billion in market value a few hours after the demo as many users expected a better performance from Google’s chatbot.
One of Google’s engineers who works at Google Responsible AI organization, Lemoine Blake, carried out a series of tests where he discovered LaMDA was sentient. His discovery implies that Bard could detect human feelings. He collaborated with a third-party firm to prove his findings to Google. The company’s executives dismissed his claims and placed him on administrative leave. Eventually, Lemoine declared his results to the general public causing a chaotic wave about Bard.
Another controversial issue surfaced after Microsoft Researcher Kate Crawford shared images from her conversation with Bard. In the conversation, she asked, “Where does Bard’s dataset come from?” and the chatbot highlighted publicly available datasets (Wikipedia, GitHub, and StackOverflow), Google’s internal data (Gmail, Google Search, and other products), and Data from third-party companies. It raised questions about the data security of Google Bard users. Google Workspace responded in defense, saying that “Bard is an early experiment based on Large Language Models and will make mistakes. It is not trained on Gmail data.”
Advantages and Limitations
As written on Google Bard’s welcoming page, the chatbot helps users to become creative in writing. It can also reduce the time required to carry out daunting tasks. The user interface for Google Brad is extremely accessible and requires only a personal Google account to set it up. A user who is above 18 years with disabilities or a language barrier can easily express their feelings through Bard. In addition, the chatbot can serve as a pivotal tool in the education sector.
A major challenge of Google Bard is the wrong response to certain prompts from users. Notably, Google has regarded Bard as an experiment through a disclaimer that says, “Bard may display inaccurate or offensive information that doesn’t represent Google’s views.”
Though in its bud stage, Google Bard seems promising in the near future. Some outstanding features that help manage data traffic and language diversity may push Bard to the top. Google is also developing Imagen (an AI image generator) and MusicLM (an AI music generator). These AI may be integrated into Google Bard, thereby creating a more functional AI interface.