Context:
Google has finally decided to answer the challenge and threat posed by Microsoft-backed OpenAI and its AI chatbot, ChatGPT. The search giant confirmed it will soon start public testing for a new AI chatbot of its own called Bard, based on the company’s Language Model for Dialogue Application or LaMDA.
Relevance:
GS III: Science and Technology
Dimensions of the Article:
- What is Bard?
- Key Difference between ChatGPT and Google’s Bard:
- Concerns with AI-based Generative Chatbots
What is Bard?
- Bard is a conversational AI chatbot based on LaMDA and Google’s technology
- Currently, it’s in an experimental stage and only accessible to trusted testers, with plans to make it more widely available in the coming weeks
- Bard provides in-depth, conversational, and essay-style answers using information from the web
- It’s a lightweight version of LaMDA that requires less computing power and is designed to scale to more users
- The model is in a testing phase and Google is gathering feedback to ensure its responses meet high standards for quality, safety, and realism.
Key Difference between ChatGPT and Google’s Bard:
Information Access:
- Bard has the ability to draw information from the web to provide fresh, high-quality responses.
- ChatGPT cannot access real-time information from the Internet. Its language model was trained on a vast dataset to generate text based on input and only includes information until 2021.
Generated Responses:
- Bard synthesises responses that reflect differing opinions for questions where there might not be a clear-cut answer.
- ChatGPT responds to complex queries with varying degrees of accuracy but is limited by the information available in its training dataset.
Concerns with AI-based Generative Chatbots
- Text generation software from Google and OpenAI can be prone to inaccuracies
- The ability to access the Internet in real-time can lead to the generation of biased and stereotypical content, including hate speech
- Google has taken a more cautious approach for Bard, possibly because of the risks involved
- A 2020 draft research paper by Google AI researchers flagged the need to proceed carefully with text generation technology
- The cautious approach resulted in two prominent researchers, Timnit Gebru and Margaret Mitchell, being fired.
-Source: Indian Express