Google's Gemini AI beats OpenAI's ChatGPT in this test

Google Gemini has outperformed OpenAI's ChatGPT in a series if comprehensive multi-disciple tests, as per a new report. Gemini took the lead by 3 percent.

Published: Feb 19, 2024, 09:05 PM IST

Google has introduced Gemini 1.5, the successor to the Gemini 1.0.

Story Highlights

Google's Gemini AI has outperformed ChatGPT in this test.
Google Gemini AI took lead by 3 percent in a series of multi-discipline tests.
As per a report, the Gemini Ultra beat ChatGPT in four key fields.

Google’s newly updated AI chatbot Gemini has outperformed OpenAI’s ChatGPT, which is owned by Sam Altman. The AI tool was able to take the lead in a series of comprehensive multi-discipline tests.

5 Creative Things You Didn’t Know ChatGPT Could Do For You

According to report, Gemini outranked ChatGPT by three percent in four key fields including technology.

Oracle Layoffs: Over 100 Cloud Jobs Cut Amid Rising AI Costs

Gemini’s Ultra outranked the GPT-4 in the following subjects — Art and Design (47.1), Health and Medicine (67.3), Human and Social Sciences (78.3), and Technology and Engineering (47.1).

“Powered by Google’s custom-designed tensor processing units, its robust infrastructure enables efficient training and deployment, driving advancements in real-time applications,” said Stocklytics Financial analyst Edith Reads.

“The technical innovations powering Gemini extend beyond language processing, with remarkable performance in processing images, videos, and audio,” he added.

One area where GPT-4 was able to perform better than Gemini was in Business and Science. That’s a small lead but overall Gemini comes out to be better here.

Google’s Gemini AI system has three versions — Gemini Ultra, Gemini Pro, and Gemini Nano. The Ultra is the most advanced tier that offers GPT-4 a run for its money, as seen in the aforementioned test.

In other news about Gemini, Google confirmed that it will launch more Gemini models to customers with new updates and expand its availability on its Vertex AI platform.

Meanwhile, OpenAI recently launched its text-to-video AI model called Sora. The revolutionary tool lets you create anything with just some prompts. The initial videos shared by OpenAI came out to be far more realistic than any other existing text-to-video AI tool.

While it sounds like a great tool, you might be worried about its use cases. However, OpenAI has restricted the use case of certain prompts for safety. For instance, you cannot create a nude video of people or create deep fakes of a celebrities.

Sora is currently not available to the public but it may launch for everyone soon. For now, it’s access has been given to select individuals such as visual artists, designers, and filmmakers.