TECH NEWS – Google announced that Gemini 1.0 AI will be available in Nano, Pro and Ultra sizes.
Google presented Gemini 1.0 AI at Google I/O 2023. And today, the company finally launched it on the market in three sizes. Distribution to Bard begins. The company has shared plenty of details about how the new model will work. They shared some really fascinating details.
Google claims that Gemini 1.0 happens to be the most capable and general model available. The model is actually able to progress and form separate components for different modalities. You can combine them later.
It’s worth noting that Google mentioned that the model will deal with anything that requires more conceptual and complex reasoning.
Regarding Gemini, Google “pre-trained different modalities from the start” with TPU 4 and TPU v5e. Google also announced TPU v5p today. It claims to be the “most powerful, efficient and scalable”.
To show just how powerful Gemini really is, Google actually demoed Gemini processing 200,000 scientific research articles. He filters out the relevant ones and then summarizes the data in an hour. Coding is another thing that Google has paid attention to. They mentioned that the model will be able to understand, explain, and generate high-quality code in multiple programming languages such as Python, Java, C++, and Go.
Gemini 1.0 AI is available at the time of writing in three different sizes, which will range from data centres to phones:
- Gemini Ultra – the largest and most capable model for extremely complex tasks.
- Gemini Pro – the best model for a wide range of tasks.
- Gemini Nano – the most efficient model for in-device tasks.
AI is actually powerful in terms of multimodality as well. The Gemini Ultra will compete against the ChatGPT-4V in the image, video and audio tests. Google DeepMind even shared a technical report that provides more information.
Google has also stated that Gemini is one of the safest AI models to date, so users who are concerned about the issue need not worry about it in the first place.
Users can experience Gemini Pro with Google Bard. This is a tool I’m rolling out right now, and if you’re interested, this feature will offer advanced thinking, design, and writing. It will also help you understand and summarize the content. Google said that the new model will be able to perform better than ChatGPT 3.5.
Bard with Gemini Pro is now available in 170 countries, for now only in English. The European and British regions will receive the function shortly. Gemini Pro will initially work with text-based prompts. But other modalities will soon become available.
Source: Google Blog