Google opened access to ChatGPT competitor

Google Corporation opened limited access to Bard, an artificial intelligence (AI) chatbot that could compete with ChatGPT.

Today we’re starting to open up access to Bard, our early experiment that lets you collaborate with generative AI. You can use Bard to boost your productivity, accelerate your ideas and fuel your curiosity. Learn more, including how to sign up ↓ https://t.co/4zDI5RD1fr

— Google (@Google) March 21, 2023

At the moment, the bot is available for testing only to residents of the US and the UK who have received invitations. It is planned that the developers will expand access to the bot over time, but the exact release dates are not indicated.

According to project leaders Sissy Xiao and Eli Collins, Bard will be more of an addition to, rather than a replacement for, traditional search. Like similar services, Bard contains a text field in which users can ask any question they are interested in. However, responses can be inaccurate and offensive, warns the bot. Moreover, Bard’s answer has nothing to do with the position of the developer company.

First reviews

Journalists from The Verge have already held the first “test drive” of the bot. They shared that they received a quick response from Bard to questions such as “how to encourage a child to go bowling” or “recommend movies about robberies.” When answering a question, the bot generates three options that are almost the same. Each answer is accompanied by a Google button, clicking on which opens the classic search.

However, like competitors, Bard was not good at working with facts. Journalists asked the AI ​​to talk about the latest briefing at the White House. As a result, the bot correctly identified the press secretary as Karine Jean-Pierre, but did not mark the other people present.

ChatGPT offered to invest in risky tokens

One of Bard’s main competitors, ChatGPT, has designated high-risk assets as low-risk and approved them for listing. The Coinbase crypto exchange team conducted a study in which the bot was asked to analyze assets for listing on the platform.

The obtained results were compared with the real results of the company’s employees. In 12 out of 20 cases, the opinion of the bot coincided with the opinion of Coinbase. However, five of the remaining eight assets were designated as low-risk by ChatGPT, which was a very poor result.

As a result of the experiment, Coinbase came to the conclusion that ChatGPT cannot yet be used for a full asset verification. However, the company is confident that its accuracy can be significantly improved. Now AI is planned to be used as an auxiliary tool in the analysis of assets.