Google has revamped its chatbot, Bard, after a disastrous initial launch earlier this year. The promotional video for Bard showed the chatbot answering a question incorrectly, resulting in Alphabet losing $100 billion in valuation and facing criticism for its inability to challenge OpenAI-Microsoft’s chatbot ChatGPT. However, since then, Google has taken a measured approach to developing Bard, internally testing the chatbot before rolling it out publicly for select users.
Under the guidance of CEO Sundar Pichai, Google sent a detailed list of do’s and don’ts to employees to follow while testing Bard. The company’s cautious approach highlights its fear of getting it wrong again and jeopardizing future success. Google has now updated Bard to include PaLM 2, a new, more powerful language model that has improved its search capabilities.
In addition to PaLM 2, Google has integrated other Google products and third-party plugins into Bard to enhance its capabilities. The company intends to add Adobe Firefly and image prompting soon, giving Bard the ability to generate images on demand. It has also enriched its responses with images and provided precise location support to offer better search results.
Google has addressed Bard’s weaknesses in mathematical tasks, coding questions, and string manipulation with the help of implicit code execution. Bard can detect computational prompts and run code in the background to improve its utility. Google is positioning Bard not just as a chatbot but as a search engine too.
By slowly improving Bard’s functionality over time, Google has avoided any further missteps and delivered a chatbot that is now challenging ChatGPT and Bing. Although Bard may not be perfect, it shows that Google has learned from its initial blunder and has taken a patient approach to development. As a result, Bard has now evolved into a more advanced chatbot that meets the expectations of its users.
