Google held an event in Paris today where it was expected to speak in greater detail about Bard, its particular rival to ChatGPT. This conversational chatbot was presented this week with the aim of demonstrating that Mountain View is working to avoid the threat posed by Microsoft with the new Bing, but the truth is that the event has been conservative.
Bard wants to play it safe. Those responsible for Google hardly offered new details about the conversational chatbot. They confirmed what was already published on their blog a few days ago: this is an experimental chatbot based on LaMDA with which it is possible to propose a first competitor for ChatGPT.
This was what a conversation with Bart looked like in the short demo offered today by Google.
privileged access. This first iteration of Bard is intentionally lighter: that will allow them to build up as they test. Initial access, yes, will be limited to a group of “trusted testers” (“trusted testers” to later launch it globally. Google did not give specific deadlines or dates, and there were only a couple of small demonstrations of its capacity. They did speak that next month they will give limited access to their Generative Language API, with which developers will be able to integrate those functions in future platforms.
Questions without correct answer. In Google they talked about how the current search engine already gives direct answers to questions that have them (“what is a constellation?”). However, there are questions that do not have a single valid answer. They call them NORA (“No One Right Answer”) and the example would be, continuing the previous one, trying to answer the question “what constellation is the best to follow when you want to observe stars?”, or then asking “when is the best time of the year to contemplate them?”, something that will result in a series of suggestions.
Going slow doesn’t mean being perfect.. At Google they have been very cautious from the beginning with the adoption of this type of technology. The firm has had LaMDA available for some time, but chose not to remove it from the laboratory so as not to damage its reputation.
Bard is an experimental conversational AI service, powered by LaMDA. Built using our large language models and drawing on information from the web, it’s a launchpad for curiosity and can help simplify complex topics → https://t.co/fSp531xKy3 pic.twitter.com/JecHXVmt8l
— Google (@Google) February 6, 2023
Bard, based on that platform, is, as we said, a light and limited version that will try to strengthen its answers, but they have already been caught in a resignation. As they point out in Reuters, when announcing the appearance of Bard they put an example that actually shows how these engines are not precise.
It wasn’t the James Webb. In the example, Bard is asked “What new discoveries from the James Webb Space Telescope (JWST) can I tell my nine-year-old son about?” In Bard’s response he includes the suggestion that the JWST was used to take the first images of a planet outside the solar system (an exoplanet), but that is not true: NASA indicates how that image was taken by the Very Large Telescope (VLT). ) from the European Southern Observatory in 2004.
Microsoft risks (because it has nothing to lose). Google’s attitude is to some extent understandable if we take into account that its search engine absolutely dominates the search market. According to Statcounter GlobalStats, its share in January 2023 was 92.9%, which makes it normal for them to tread carefully when making revolutions in this segment.
Microsoft, for its part, has practically nothing to lose, and that is probably why it is more dynamic in this case. The new Bing with built-in ChatGPT will begin rolling out in the coming weeks, likely ahead of the adoption of this type of conversational chatbot in Google searches.
