Help us prepare for our appointment
Leslie
Ever faster and ever more relevant, the new artificial intelligence models continue to impress us. Faced with increasingly fierce competition, Google is not waiting for its rivals to make the first move. The web giant is already planning to invest heavily in AI to develop Gemini 2.0. Discover the experimental models and platform updates.
Gemini 2.0 is evolving with two new experimental versions.
On the one hand, Flash Thinking Experimental aims to improve the model’s reasoning capabilities. How? By breaking down queries further. Rather than responding in one go with an imprecise (and sometimes irrelevant) result, Gemini will follow a series of steps that it will present to the user. More logic, more details, for answers that are closer to the user’s expectations. And this, regardless of the complexity of the query. With this new version, Google intends to promote interaction with various applications in its ecosystem, such as YouTube, Maps and Search.
On the other hand, Gemini Pro Experimental is developing its performance for ultra-complex tasks, such as computer programming and mathematics. Fewer errors, greater precision, and a context window of up to 2 million tokens.
Good to know: these versions are currently available via Google AI Studio and Vertex AI (or for Advanced users).
For both versions, the objective is to improve model performance (in terms of both reasoning and speed) while reducing costs.
With the launch of DeepSeek and its £5.6 million price tag, AI leaders are becoming aware of a new financial issue: how to reduce the costs of producing new AI models.
To address this, Google is launching Gemini 2.0 Flash Lite. This is a lighter version of its model. In terms of performance, Flash Lite is comparable to version 1.5 Flash, but with better quality benchmarks.
This development is clearly a response to competition, both from DeepSeek and Open AI and its 03-mini model.
The Gemini 2.0 Flash model has also benefited from several updates. Among the new features, you will find:
All these improvements are necessary in the face of increased competition.
To finance all these AI developments, Google plans to invest £60 billion. Compared to 2023, this is more than double (£24 billion at the time).
Once again, this increase in investment reflects intensifying competition. In order to maintain its position and meet growing demand, Google must continue its research into artificial intelligence (to develop new features or improve computing capabilities). Not to mention strengthening its existing infrastructure.
Google is investing so heavily in Gemini 2.0 because of its financial potential. With this model, the web giant will continue to do what it does best: online advertising. Gemini could thus become a new-generation advertising platform. With the potential to reach billions of monthly users, artificial intelligence could offer advertisers even more targeted leads.
At present, we do not know the release date for the integration of advertising on Gemini. Nevertheless, Google CEO Sundar Pichai has already mentioned native advertising concepts. So let us be prepared.
Please do not hesitate to contact our SEO agency to optimise your advertising campaigns.
