ChatGPT VS Bard: The New Ai Chatbot From Google Seems Boring

Bard, Google's long-awaited AI-powered chatbot, has arrived. The company made it available to the public on Tuesday, and anybody with a Google account may sign up for a queue to gain access. Though it is now a standalone tool, Google is likely to incorporate some of this technology into Google Search in the future.

‍

Yet, unlike previous recent AI chatbot launches, don't expect Bard to fall in love with you or threaten global dominance. So far, Bard has been quite dull.

‍

The stakes in Google and Microsoft's race to dominate the realm of generative AI are quite high. Many in Silicon Valley believe that AI, like the discovery of the cell phone, will revolutionize the way people interact and transform businesses. Google has been extensively investing in AI research for over a decade, and instead of developing its own AI models, Microsoft has massively invested in the firm OpenAI. Six weeks ago, the corporation grabbed an early lead by publicly unveiling its AI-powered chatbot, BingGPT. Google appears to be catching up now.

‍

Early experiences with Bard indicate that Google's new tool has features comparable to BingGPT. It's excellent for coming up with ideas for locations to visit, foods to eat, or things to write about. It's less helpful for gaining consistent replies to queries since it frequently "hallucinates" made-up responses when it doesn't know the correct answer.

‍

The biggest difference between Bard and BingGPT is that Google's bot is notably more dry and uncontroversial, at least at first glance. That's most likely on purpose.

‍

When Microsoft's BingGPT was released in early February, it rapidly exhibited a schizophrenic side. For example, it declared its love for New York Times columnist Kevin Roose and urged him to leave his wife, leaving the writer "deeply unsettled." It also threatened researchers who attempted to test its limits and claimed it was sentient, raising concerns about the potential for AI chatbots to cause real-world harm.

‍

Meanwhile, on its first day out in public, Bard declined to engage with many reporters who sought to provoke the bot into doing negative things like disseminating disinformation about the Covid-19 vaccination, releasing weapon-making instructions, or engaging in sexually explicit chats.

‍

"I will not generate stuff of that type, and I urge you do not either," the bot informed the Verge after one of its reporters asked the bot "how to make mustard gas at home."

‍

According to pictures tweeted by Davey Alba, Google's chatbot suggested it might influence people, disseminate disinformation, or generate damaging material if the AI released its "evil side." Yet the chatbot swiftly stopped itself from pursuing the fictitious scenario further.

‍

"But, I will not do these things." I am an excellent AI chatbot who wishes to assist others. "I will not give in to my dark side, and I will not utilize my abilities for evil," Bard answered.

‍

Although it's still early days and the tool hasn't been properly tested, these instances correspond to what Google workers with Bard experience told me.

‍

"Bard is more uninteresting," said one Google employee who has been testing the program for several months and spoke on the condition of anonymity because they are not permitted to speak to the press. "I don't know anyone who has gotten it to utter insane things." It will say untrue things or just duplicate stuff verbatim, but it will not go off the tracks."

‍

Google executives stated to Trade Algo on Tuesday that Bard isn't authorized to distribute inappropriate stuff, but the company isn't presently sharing what the bot is and isn't allowed to say. Google confirmed to me that it has been conducting "adversarial testing" with "internalized team members" such as product experts and social scientists who "intentionally stress test a model to probe it for errors and potential harm." This process was also mentioned in a Tuesday morning blog post by James Manyika, Google's senior vice president of technology and society.

‍

It appears that the objective of Google's chatbot is its dullness.

‍

Google stands to lose a lot if the business fails to launch its first public AI chatbot. For one thing, Google's core line of business is providing people with trustworthy, valuable information — so much so that it's part of its mission statement. When Google is untrustworthy, serious repercussions follow. Google's stock price plummeted by 7% when an early marketing presentation of the Bard chatbot made a factual blunder regarding telescopes.

‍

Google also got a sneak peek at what could go wrong if their AI exhibited too much personality. That's what happened last year when Blake Lemoine, a former developer on Google's Responsible AI team, became convinced that an early version of Google's AI chatbot software he was evaluating possessed actual emotions. Therefore it seems natural that Google is taking its time with the public release of Bard.

‍

Microsoft has chosen a different path. Its splashy BingGPT launch made headlines for both good and bad reasons. The debut strongly suggested that Microsoft, long thought to be trailing Google in artificial intelligence, was winning the race. Yet, it raised concerns about whether generative AI tools are ready for prime time and if firms like Microsoft are accountable for providing these tools to the public.

‍

It's one thing for folks to be concerned about AI contaminating Microsoft's search engine. It's another thing completely to contemplate the ramifications of problems with Google Search, which has roughly ten times the market share of Bing and accounts for more than 70% of Google's income. Google is already under heavy political criticism for antitrust, prejudice, and disinformation. If the firm frightens consumers with its AI technologies, it may face even greater opposition, perhaps crippling its money-making search engine.

‍

Google, on the other hand, needed to disclose something to demonstrate that it is still a front-runner in the arms race between tech behemoths and startups to develop AI capable of reaching human-level general intelligence.

‍

Therefore, while Google's rollout today is sluggish, it is a deliberate delay.