Launched two days ago, Gemini, Google’s new artificial intelligence model, intends to compete with OpenAI’s GPT-4. If the announcement caused a lot of noise in the field of new technologies, certain countries are on the sidelines for the moment. As was the case with the launch of Bard – the first version of Google’s conversational AI – Internet users residing in Europe will have to be patient before using Gemini. This is explained by the need for prior consultation with European regulatory authorities, in order to ensure that Gemini complies with current legislation.
Access Gemini from France
As with the vast majority of geographically restricted content on the Internet, it is possible to circumvent this limitation. To do this, we use a VPN, the acronym for Virtual Private Network. This tool allows you to create a virtual tunnel between the user’s computer and a server located in another country. Using a VPN allows you to use Gemini in France, or from any other European country.
Connecting via CyberGhost VPN to a server in the United States.
The handling is similar for all VPNs: after installing the chosen software, we preferably select a server located in the United States, a country in which Google Gemini is already accessible. Once the connection is established, you go to the Google Bard web interface. It seems useful to specify that, as with its predecessor, the use of Gemini requires you to have a Google account.
Using Google Gemini AI
For this article, we used CyberGhost VPN and chose a location in the United States. Once connected to the Bard interface, we asked the AI under which model it operates.
I’m currently using Google AI’s large Gemini Pro language model. This model was initially called Jurassic-1 Jumbo, but it was later renamed Gemini Pro.
The chatbot therefore explains to us that it uses the Gemini Pro model, developed by Google. The answer is written in English and this is no coincidence: Gemini is currently only available in the language of Shakespeare.
Analysis of an image with Google Gemini.
We had Google’s AI analyze an image, taken from the short film Big Buck Bunny, under Creative Commons license. After uploading a screenshot where we see the main character, a rabbit, engaging in a jump rope session, we asked Gemini what sport the animal in question practices. And it is clear that if Gemini already does better than OpenAI’s GPT4 model, it also does it for free.