Google today announced Gemini, a large multimodal language model, and has confirmed that the AI model will be added to Search and Ads in the coming months.
Twins in search. Gemini will be available in Search, Ads, and other products in the coming months. But Google Search is already testing Gemini, which has reduced Search Generative Experience response latency by 40% for English in the US and brought unspecified quality improvements.
What is Google Gemini?. Officially known as Gemini 1.0, it is a great language model that can work with all kinds of information: text, images, audio, video or code. Gemini is meant to compete with OpenAI’s GPT. Three “sizes” were announced:
Ultra, for very complex tasks. Pro, Google’s “highest performance model” for a wide range of tasks. Nano, for on-device tasks.
Twins in Bard. Google Bard now works with a “tuned version” of Gemini Pro. Bard now has more advanced reasoning, planning and understanding, Google said. Calling this Bard the “biggest update” since launch, it’s now available in English in over 170 countries, with plans to expand even further in the “near future.”
A new version, called Bard Advanced, will debut early next year, said Google. It’s unclear how much Bard Advanced will cost, if at all. It will be powered by Gemini Ultra.
Google DeepMind has also shared a report which found that Gemini outperformed GPT-3.5 for six of eight industry benchmarks. But OpenAI now offers GPT-4, released in March.
Why we care Although Gemini is only in testing in Search, Google has confirmed that it will be part of the Google Search experience in the near future. Also, if you’ve been using ChatGPT for any of your daily tasks, you might want to check out the updated Bard to see if Bard really beats the free version of ChatGPT, powered by GPT-3.5.
[ad_2]
Source link
