The generative artificial intelligence called Bard changes its name and becomes Google Gemini, inspired by the most advanced model available at Big G. At the same time, CEO Sundar Pichai announced apps for Android and iOS.
This brief background summarizes the telluric movements that will soon shake the world of generative artificial intelligence as we know it today and led by ChatpGPT by OpenAI and Copilot by Microsoft.
Google Gemini is an evolution of Google Bard and promises to deliver sparks although, for the moment at least, it is better to proceed by pulling the handbrake of expectations.
What Google Gemini promises
To fully understand what Google Gemini can do, it is necessary to start from the Pro version, which can be used in over 40 languages, including Italian, and which is to be considered as the “light” tool which, being free, only gives a taste of what is in the strings of Advanced versionfor a fee, but more capable and fully usable even on mobile devices.
Gemini Advanced leverages the Ultra 1.0 language model which, according to Google, is capable of increasing the complexity of the tasks that generative AI can solve, because it is equipped with a more refined and snappy “logical reasoning”.
All this reverberates on the creation of content, on the ability to write code (portions of software) and on the possibility of integrating Gemini into Gmail and into documents created with the Google Workspace suite, Big G’s alternative to the well-known parallel Microsoft Office suite.
The most interesting thing is that Gemini supports longer and more complex prompts, i.e. accepts instructions more detailed and this opens the doors to a more fierce challenge to direct competitors, especially to ChatGPT which, after a period crackling between the end of 2022 and the beginning of 2023, it attracted fewer and fewer users and showed even worrying limits on which the news often focused, as in the case of the New York lawyer who used it to write a speech discovering then that the Chatbot had invented previous sentences and legal references or, perhaps even worse, as in the case of complaints for copyright infringement.
The pros and cons for users
Starting with the advantages, Google Gemini Advanced has a cost of 21.99 euros per month (it can be tried for free for two months) and also includes 2 TB of space on the Cloud. The cost is similar to that of Copilot Pro (22 euros per month), the premium version from Microsoft, which promises to be able to surpass in effectiveness and productivity.
In short, more capacity (and 2 Terabytes of Cloud storage) at the same price, a sign of a challenge that also unfolds on the commercial.
The disadvantages, which are only hypothetical for the moment, may become apparent as time passes. In fact, at the end of 2023, Google showed off Gemini’s capabilities, but committed the error of having mystified reality a bit: it is true – as Google itself demonstrates – that Gemini has given results of a certain level, but distorting the comparison with ChatGPT a bit.
To explain with an example what happened, Google asked the same questions to Gemini and ChatGPT, obtaining better and faster answers from its product, without underlining however that – to use an image that is not entirely punctual but useful for the explanation – it played at Guess who? giving ChatGPT fewer attempts to arrive at the correct solution.
However, there are the ingredients to see an evolution of generative AI that increasingly enters our homes and in the performance of our professional tasks. Much of the work has already been done, the rest is still to be invented.