With Bard, Google joins the race against ChatGPT to answer our queries

Google unveiled 'Bard' - an AI conversation service to take on ChatGPT.

Google CEO Sundar Pichai announced the release of Google Bard, a conversational AI service on February 6. It will be available to “trusted testers” before the technology company allows public access to the product. Bard is a technology product built on Google’s Language Model for Dialogue Applications (LaMDA), a project Google started two years ago. It will use information already learned to respond to queries from users and then continuously learn from user responses to refine its capabilities.

Think about how Gmail or WhatsApp suggests quick responses to an email or a chat on the notification bar of your mobile phone. Google Bard’s capabilities would likely go well beyond chat responses and potentially draft whole poems, essays, or opinion pieces based on user prompts, similar to ChatGPT, powered by Microsoft-backed OpenAI released last November.

The Search Engine Killer

Google’s Bard launch comes two months after Microsoft released ChatGPT. It was important for Google to launch quickly because ChatGPT threatens Google’s search dominance. A Google search on a topic provides all relevant information that users must then sift through to find answers to their query. In contrast, ChatGPT sifts through its learnings and summarises the result for the user.

For example, a user would have to first search for ‘Modern Indian History’ on Google, open each link, collect information, and then write it out. On ChatGPT, the user can simply ask the software to write a 500-word essay on the same topic and get the same, if not a better researched output.

Under Regulatory Fire

It gets more complicated for Google in India. The company’s search engine commands 94 per cent of the online search market in India, according to information provided to the Competition Commission of India (CCI). Google made several changes to its business in India, after the CCI imposed a penalty of Rs 1,337.36 crore  for abusing its dominant position in multiple markets.

One such change is that Google will now allow Android users to choose their default search engine when they set up their device. Earlier, devices came preloaded with Google Search, but now users may choose another browser, including Microsoft’s Bing.

The post-CCI penalty changes may erode Google Search’s market share in India. Further, the Indian government announced a panel to review existing antitrust laws and submit a draft Digital Competition Bill in the next three months. It may require Google to make further changes to its business model in India.

Concerns About AI

The initial criticisms of ChatGPT, that Bard may also face are twofold. First, there are concerns that creativity and labour may become indistinguishable to readers. For example, a teacher may find it challenging to go through essays submitted by a tech-savvy younger generation and distinguish between AI-generated works and originals. To address this, OpenAI launched an AI Classifier that distinguishes between AI-text and human text last week. Bard may also need to launch a similar software.

Second, AI is prone to bias because they are trained on existing data and the data reflects societal inequalities of the times. AI generated artwork on portraits of couples at weddings in different parts of India produced biased outputs that social media users criticised for highlighting stereotypes. Both OpenAI and Google acknowledge the concern, and are investing resources to address the issue, but neither has found a solution.

In India, digital adoption does not mean digital literacy, and there are several constraints to digital access. Students may have access to ChatGPT or Bard, but teachers may not have access to or the skills to understand the AI Classifier solution. Second, AI generated content amplifies bias and readers may consume biased information as truth. Bard will need to address these concerns before it rolls out in India to win the trust of Indian users and the government.

Varun Ramdas is a technology policy researcher working at the intersection of emerging technology, public policy and society. Views are personal and do not represent the stand of this publication.