Bard arrives in Italy: Google’s generative AI responds (and speaks) in Italian

Bard arrives in Italy: Google's generative AI responds (and speaks) in Italian

[ad_1]

It took less time than expected: in the middle of last May, in a short chat with Italian TechSundar Pichai, current CEO of Google, told us that “we want to be in Europe with Bard” but that “we want to be there respecting all the rules”.

Said done, apparently: from this morning, exactly two months after that meeting, if you connect to bard.google.com, Google’s generative AI is not only accessible but also responds in Italian. In writing and orally, too.

The case

ChatGPT slows down the race: in June the first drop in users

by Emanuele Capone, Bruno Ruffilli


How does Bard deal with privacy

This is the news for us, but the news not only this: again from today, Bard is available throughout the European Union and also in Brazil, for a total of 59 new countries and territories (yes add to the previous 180) and 40 languages ​​overall, including Arabic, Chinese, Hindi, and Spanish. And Italian, indeed.

Jack Krawczykhead of Product Management at Mountain View, spoke of Google’s intention to continue having an “ambitious and responsible” approach to AI and obviously confirmed that there was “a long confrontation with the authorities that deal with the protection of privacy” and that this confrontation “will continue as new people use this tool” and eventually face unexpected or unforeseen problems.

When interacting with Bard, “Google collects data such as conversations, location, comments, and other usage information, with the goal of developing, improving, and delivering our products, services, and AI-powered technologies: From what we understand, Bard can be accessed from EU and comply with the GDPR thanks to some changes to the interface, new warnings to users that will appear during interaction and above all thanks to a new panel that allows you to select (or deselect) what kind of information you want (or don’t want) Bard to collect and the ability to also remove entire conversations later had with Google’s artificial intelligence.

In detail, Google “stores the activities on Bard in the account (of the user, ed) for a maximum of 18 months, but it is possible to change this duration to 3 or 36 months” and in fact it is also possible to “completely deactivate this function and easily delete activity on Bard”; We also “allow users to request removal of content in accordance with our policies or applicable laws and to export their information.”

Future

Once upon a time there was AI: if fairy tales are told by artificial intelligence

by Francesco Marino



Bard in Italy, the new functions

Bard, that is “accessible to all without a waiting list”, as Krawczyk confirmed to us, is supposed to be “a tool to unlock the greatest computer of all, which is the human brain”. For this reason, Google speaks of “augmented imagination” rather than a chatbot or a conversational AI: the idea is that Bard gives answers but above all offers ideas from which people can start to face what they have to face. Whether it’s writing an email or any text, preparing for an interview, fixing the sink, figuring out how to get to Mars or solve the 3-body problem (which is this).

All the answers, Krawczyk wanted to point out, they will be “grounded”, i.e. supported by what is Google’s core business, what the company has been good at for decades: online searches. For now Bard will operate alone, on its own site, but will still be linked to search results and will report links to the sources of various information, especially in the case in which copyrighted materials are mentioned.

Bard arrives in Italy, and in the rest of the European Union, but is also enriched with functions such as integration with Lens, the Google tool that allows you to search starting from an image: for now only in English, you can use (for example) a photo as a prompt and get help from Bard to start a conversation or a text starting from the subject shown . Initially only in English (but “will soon be extended to other languages”) also the possibility of changing the tone and style of the answers, so similar to the Magic Compose feature of Pixel smartphones and here according to 5 different options: simple, long, short, professional or informal.

More: it will be possible resume the conversation from a previous and recent answer, export code Python in Replit and in Google Colab, Bard can be used in Gmail to reply to an e-mail message, he will be able to speak and (for example) tell a storyinvented or already known, and also there will be a button Share for sharing both individual prompts and the responses they generated. So that Bard’s insights can serve as insights for other people.

@capoema

[ad_2]

Source link