-Bard will initially be introduced to a select external and internal groups for testing before being fully released
Google has introduced Bard, an experimental conversational AI service that is powered by LaMDA – Language Model for Dialogue Applications which is the engine behind its next-generation language and conversation capabilities. LaMBA was released two years ago.
The company announced the introduction of Bard through a blog post by none other than the CEO of Google and Alphabet Sundar Pichai.
“We’ve been working on an experimental conversational AI service, powered by LaMDA, that we’re calling Bard. And today, we’re taking another step forward by opening it up to trusted testers ahead of making it more widely available to the public in the coming weeks.” Sundar said.
He goes ahead to say that Bard will combine the power of the world’s knowledge with the power of their large language models. This basically means that like ChatGPT, Bard will utilize the wealth of information available in the interwebs to give users instant responses to their questions.
Although the blog post by Mr. Sundar Pichai did not make it explicitly clear whether Bard will be able to give answers based on real-time information or most recent happenings, there was a hint to this effect in the blog. He alluded to the fact that Bard can give answers such as “the best strikers in football right now”. The hint is in “right now”. If Bard can give such information, then it may beat ChatGPT along the lines of data freshness. As those who have used ChatGPT may know, the answers it gives are based on information that is from 2021 and backwards. This means that ChatGPT may, for instance, not give you an answer that is backed by stats from 2022 or 2023. This has been tricky, and if Bard is capable of doing it, then it’s highly likely that it will give ChatGPT a good run and eat a good chunk of the millions of users it has gained so far.
A combined feedback from Google’s internal testing teams as well as a select external testers will be used to fine-tune Bard and ensure it meets the company’s sworn high bar for quality, safety and groundedness in real-world information.
When will Bard be available to the public? No specific date has been given, although Mr. Picha said it could happen in the coming weeks when it will be more “widely available to the public”. Again reading into the tiny details, it could appear that the availability may not be 100% public and this could explain the use of the word widely.
Many critics say that Google has been too cautious in this front, approaching it as if they want to release a perfect product – which is against the rule of technology. There is a general feeling among the curious lot that instead of waiting to get things perfect, Google should just release whatever products they have to the public and deal with the consequences as they come.
But there are those who have also come to Google’s defense, saying that Google is no longer the 1990s startup which had nothing to lose and everything to gain by being quick to release products. Google is now a giant whose actions have a huge bearing in the tech scene. And this then explains the company’s seemingly cautious speed towards the quickly evolving AI scene. Even Microsoft, which has been hailed as doing a great job by backing OpenAI which is the company behind ChatGPT, is not yet at the forefront of directly releasing fragile AI products. So we can therefore say that Google could be a hand in as far as putting a name behind a product is concerned.
AI driven search results?
Google is also gearing out to role AI features in search with the aim of making the search experience more insightful. It means for example, that you can go to the Google search bar, type a complex question that would normally take a lot of time to understand, and Google will unearth useful answers for you.
This could potentially point to an exciting future, where AI features will be capable of distilling information to help the user see the big picture, Says Sundar.
Starting March 2023, Google will invite developers, creators and enterprises to ry the company’s Generative Language API. This will initially be powered by LaMDA and more models will follow afterwards. This is really good news, and it even gets better as the CEO says that in future the company will create and provide a suite of APIs and tools to hep others build exciting applications with AI. Google Cloud is already entering into partnerships with various startps to accelerate the building of their AI systems. Most recent partnerships include Cohere and Anthropic,
