Google has unveiled the new Gemini artificial intelligence (AI) model, a system that will drive a number of the company’s products, including its Pixel phones and the Google Bard chatbot. “The most capable and general model we’ve ever built,” according to the business, Gemini will make AI “more helpful for everyone.” It might be accurate, based on the amazing video demo located further down this page.
There will be three’sizes’ available for Gemini: Ultra, Pro, and Nano, each intended for a distinct purpose. All of them will be multimodal, which means they can process a variety of inputs. Gemini, according to Google, can process suggestions in the form of text, code, audio, graphics, and video.
Gemini Nano will fit in your smartphone, while Gemini Ultra is made for very demanding use cases like in data centers. This means that the greatest Android smartphones might potentially gain a major AI advantage.
With all of this additional capability, Google maintains that it carried out “rigorous testing” in order to detect and stop negative outcomes from people using Gemini. Gemini’s multimodal feature makes it difficult to combine two seemingly innocent inputs, like text and a picture, to produce something insulting or harmful, the business claimed.
approaching all of your devices and services
Google has been under pressure to match ChatGPT from OpenAI and its cutting-edge AI features. In fact, rumors that Google had postponed its Gemini announcement until next year because of the device’s reportedly subpar multilingual capabilities surfaced just a few days ago.
Now, it appears that the reports were inaccurate or that Google is moving on with Gemini in spite of its purported flaws. It’s noteworthy that Gemini will initially only operate in English on this point.
For you, what does Gemini mean? Google claims that the Pixel 8 Pro phone can now run Gemini Nano, enabling you to carry all of its AI power in your pocket. Gemini may be found in two new Pixel 8 Pro capabilities, according to a Google blog post: Summarize in Recorder, which can summarize your recorded chats and presentations, and Smart Reply in Gboard, which offers message replies to you.
The business claims that this is “the biggest upgrade to Bard since it launched,” and the Google Bard chatbot has also been modified to run Gemini. Furthermore, Google states that in the upcoming months, “Gemini will be available in more of our products and services like Search, Ads, Chrome, and Duet AI.” Google says.
ChatGPT vs. Gemini
It’s no secret that over the past few months, OpenAI’s ChatGPT has been the most popular AI tool; Google hopes to change that with Gemini. The business has also asserted some very audacious promises regarding its capabilities.
For example, according to Google, Gemini Ultra performs better than state-of-the-art outcomes in “30 of the 32 widely-used academic benchmarks” that are utilized in the development and research of large language models (LLMs). Put differently, Google believes it surpasses GPT-4 in almost every aspect.
According to Google, Gemini performed better than the GPT-4 LLM, which runs ChatGPT, in seven of the eight text-based assessments. According to Google’s comparison, Gemini performed better in all 10 benchmarks for multimodal tests.
Does this imply that a new AI champion has emerged? We’ll have to wait for additional independent users to evaluate it in the real world before we can make that determination. Nonetheless, it’s evident that Google is fighting AI head-on. Now, the initiative is firmly in the hands of Microsoft and OpenAI.
1 Comment