Meet Gemini, a Sophisticated Google Search Replacement Using AI Tech – 2 hours ago

Jakarta, CNBC Indonesia – Google officially released an artificial intelligence (AI) model called Gemini. Meanwhile, the AI ​​available to Google Cloud users can analyze graphics and write computer programming code.

Gemini is predicted to have better performance and capabilities than GPT-3.5, the AI ​​model that underlies OpenAI’s ChatGPT service. This model was released in three categories which are Gemini Ultra, Gemini Pro, and Gemini Nano.

Gemini Ultra is the model with the largest capacity, Gemini Pro is a model that can be developed for specific tasks, while Gemini Nano is used for specific things and can run on mobile devices such as cellphones.

Currently, Gemini can be used by companies using Google Cloud in their respective applications. Starting December 13, software developers and enterprise customers can access Gemini Pro using APIs in Google AI Studio or Google Cloud Vertex AI. Android developers can also start developing applications with Gemini Nano.

Gemini will also be the engine behind Google services, such as the Bard chatbot and Search Generative Experience, which is Google’s search model that answers questions in conversational form.

Search Generative Experience is said to replace the role of Google Search, which provides answers to internet search results in the form of a list of pages on the internet. This product is currently still in the development stage.

Based on reports CNBC International, Gemini can be used by companies to provide advanced customer service via chatbots. Another use is providing product recommendations to customers and identifying trends to design promotional campaigns.

Gemini can also be used to create marketing content to blogs or as a productivity supporter because it can summarize meeting results and write computer programming code.

Google, at the Gemini launch event, showed off Gemini’s ability to take screenshots of a graph, then refine the graph after analyzing hundreds of pages of reports. Gemini was also shown to be able to photograph math homework sheets, then identify the correct and incorrect answers.

Gemini Ultra is claimed to be the first AI model that can beat humans in MMLU (language understanding in large numbers of tasks). In the MMLU test, Gemini Ultra is given the task of understanding 57 fields of science including mathematics, physics, history, law and medicine. According to Google, the Gemini Ultra is capable of understanding nuances in complex subjects.

[Gambas:Video CNBC]

Next Article

6 New Google Bard Features Complete with How to Use Them, Check Out!

(Verda Nano/hsy)