Google, which declared a "code red" for the arrival of ChatGPT in December last year, has just introdotto its conversational artificial intelligence (AI). It's about google bard, a solution that we have already heard about e whose heart is LaMDA, a powerful experimental language model designed especially for dialog applications. The announcement was made by the CEO of Alphabet Sundar Pichai on the official blog above you will find the link. But let's go into detail.
Will Google Bard really be a viable alternative to ChatGPT? The war is now open and it will not be fought only between Big G and OpenAI
In a publication entitled "The next major step in our AI journey“, the Google executive acknowledged that artificial intelligence is the most complex technology they are working on, but which they consider essential to fulfill their mission of organizing the world's information and making it universally accessible. The research giant explains that for more than six years it has refocused its strategy towards AI. Since then he has worked on several solutions, one of which is TheMDA, that it was in the testing phase and limited to a very few people (including the engineer who was fired for suggesting he had a "conscience").
Google just renamed LaMDA (Language Model for Dialogue Applications) to Bard, a much friendlier name now that its general rollout is getting closer. The company says it will be available to a small group of "trusted testers" starting today. The company's movement, though limited to a few, is nonetheless significant. Google, which for years has benefited from being the dominant leader in the research market, has seen that position threatened by some changes in the industry. One of their concerns, it seems, is Microsoft's possible adoption of GPT-4 in Bing soon.
Read also: ChatGPT creators, OpenAI, launch a tool to detect AI-generated texts
We still don't know what capabilities Google Bard will have when it arrives to the general public, but Pichai has given us some very important clues. First of all, we are sure that will be connected to the internet, a substantial advantage over the current ChatGPT and ChatGPT Plus, the paid version of the bot, which limited knowledge, specifically, to sometime in 2021. Reportedly, “Bard takes information from the Internet and uses it to offer high-quality, current answers“. Furthermore, it will serve to “dampen curiosity or give free rein to creativity”. The idea is to ask questions in natural language and quickly receive elaborate and complete answers.
Google's conversational AI will not work independently of the search engine, but will be integrated as a function that seeks to add value to searches and attract those users who are looking for a direct explanation in depth and don't want to navigate through hundreds of results. We recall that one of the great challenges facing those involved in the launch of massive products powered by artificial intelligence is the veracity of the answers. This technology has a great capacity to invent facts, give inaccurate data and support (paradoxically) social problems such as racism and xenophobia. In fact, Google has tried to be very cautious to avoid damaging its reputation: in essence, there will be many limitations.
If all goes to plan, the massive launch of Google Bard shouldn't take too long. The Mountain View company says it will make it publicly available "in the coming weeks" after i tests they will have bothered to test it. Also, from next month, developers will have the opportunity to experiment with the APIs of this technology to power their applications. The race to excel in the world of artificial intelligence with different products is getting more and more interesting. OpenAI's move to release ChatGPT to the public last year seems to have woken up a sleeping giant, Google.
