Google unveils Bard, its answer to ChatGPT: soon also in Search

Last December, interviewed about why Google had not yet prepared an answer to the famous ChatGPT chatbotSundar Pichai had highlighted how one of the main reasons for this choice was related to the fact that it is problematic to propose such a powerful tool when this is largely unreliable; people consider Google a real unshakeable point of reference with regard to online research, which is why it is not good to present yourself with a product that is not able to return correct answers to the questions of the general public.

Let’s take a leap forward from the declarations of December and we find ourselves today, in the first days of February, with Bard’s official announcement, one of two chatbots that Google is developing. The blog of the Mountain View house is very clear and confirms that the work begun on LaMDA and on all the other Google AI and DeepMind technologies is evolving at a very high speed, requiring double computing powers every 6 months (well beyond the computational growth rate established by Moore’s Law), and that now is the time to go further and test the full potential of LaMDA through a completely new tool, namely Bard.

Bard is none other than the Google direct response to ChatGPT, which from today will be available to a small group of testers chosen by Google. The aim is to expand the audience quite quickly, making Bard available to an increasing number of people over the next few weeks.

The main difference between the proposals of OpenAI and Google lies in the fact that the latter does not rely on a well-defined information database, but rather draws its resources directly from the webin order to provide answers always updated and in line with the latest news present. Google sees Bard as a useful tool to help people develop their knowledge and creativity; for example can explain the latest discoveries of the James Webb Space Telescope to a 9 year oldor can offer an analysis of the best football forwards right now, or offer advice on how to put your skills and abilities to good use.

At this early stage, Google specifies, Bard will use a simplified version of LaMDA, therefore going to require a lower computing power than the real potential of the system. This will allow Google to make it available to more people, so that they receive enough feedback to allow the company to improve Bard based on very high quality, safety and soundness criteria.

In fact, Google continues to consider it essential that its AI tools accessible to the general public – therefore starting with Bard – are able to be extremely precise and reliable, in order to prevent scenarios in which incorrect and misleading advice is provided by the chatbot. The poor reliability of a system like Bard, in fact, would contribute to making it an unattractive and even dangerous tool, for this reason Google insists on noting that its approach to new technologies related to artificial intelligence It will always be bold but also responsible.

The expansion of tools like Bard will also happen through the traditional Google Search, where some elements able to make it more effective in giving answers to complex questions and which include advice or opinions will soon be integrated. For example, the image above shows how Google Search will be able to provide an accurate answer to the question “Is it easier to learn piano or guitar and how much practice is needed for each?“, going to compose a text that has not been fished from one of the results indexed in Google, but it is a real original answer realized through the application of AI algorithms which will be integrated into the search box.

In short, the innovations thought up by Google in these fields are really many and also involve many of the other AI projects – such as PaLM, Imagen and MusicLM – which will soon be able to offer support to many people in very various fields. To learn more about the topic, we leave you with the article published by Google on its official blog; you can find it in Source.