After Google expanded the availability of its AI chatbot Bard this week, the AI arms race has become more of a comparison game than ever before.
Google's opening access to the generative AI chatbot comes a week after OpenAI released the latest version of its large language model, GPT-4, which is now available on ChatGPT-Plus and as an API.
As Bard is a direct competitor to ChatGPT, many users have no choice but to compare Bard to ChatGPT.
While it still in its early phase, Bard does not quite stand up to ChatGPT, many observers have already noted.
"It is definitely a product still in development," Forrester Research analyst Rowan Curran said.
Rowan CurranAnalyst, Forrester
Curran signed up for the waitlist and gained access to Bard.
Similar to Bing Chat (Microsoft's search engine with ChatGPT functionality), Bard attempts to provide different and variable responses to the users, according to Curran.
However, unlike Bing Chat, Bard's responses are inconsistent. For example, while Bard can respond to queries about a specific Instagram account, such as the handle and where it got its responses from, it may not know who the person linked to the account is, Curran said.
"Bard is what I might have expected to see when Bing Chat was still in internal beta," he added. "There seem to be some inconsistencies in the way that it performs. I don't think it speaks to underlying lag of the capabilities of the platform."
Instead, these inconsistencies speak to how the technology is still in its infancy.
With OpenAI's popularity and Microsoft's financial and computing power support of the AI research and development firm, the AI race is currently geared favorably towards Microsoft and OpenAI.
Therefore, Google has no choice but to respond by opening Bard to a more extensive test audience, said Daniel Newman, CEO and analyst at Futurum Research.
"Google needs to make sure this moment doesn't pass and they're sort of seen as laggard," Newman said. "They had a lot of this technology in flight, and this was a pivotal moment for them to shine."
Despite Bard's early hiccups, Google could still come out on top of the AI race.
"These large language models and the applications it runs on gets better as they get used more," Newman said.
He added that he anticipates an aggressive push in the next six months to make Bard more accessible, usable and desirable.
"Google still does have the lion's share of search, so in that particular area of utilization, they do stand to have an opportunity to vet it somehow and then commercialize it," he added. "There's going to be some sort of a technological crawl chart, you could say, with the features being advanced across the portfolios of products."
Both Bard and Bing Chat may look and feel similar when they reach general availability, Curran said. Yet they each may provide compelling value in different ways.
"There are multiple ways to have a large language model-driven search experience that are both similarly compelling and simple," he said. "How each of these evolve ... may have implications for how people end up wanting to use it in the enterprise context."
Esther Ajao is a news writer covering artificial intelligence software and systems.