Google launches Gemini 1.0 model with Bard AI in different sizes including Ultra

We know that the world is now dependent on large language models or LLMs in order to power their chatbots as well as make use of it in different applications related to artificial intelligence. We know that ChatGPT was launched with GPT-3 and later GPT-4 was released and we will see the launch of GPT-5 which is the underlying algorithm that powers everything that ChatGPT or other bots based on it. Now, after teasing the model at Google I/O 2023, Google has announced the launch of Gemini 1.0 which will be launched with its Bard AI. As of now, Gemini 1.0 is included with Bard AI wherever Bard has been launched so you can clearly see the different between Bard’s responses earlier and now.

It is mentioned that Gemini is “Google’s “most capable and general model,” Gemini can “understand, operate across, and combine” text, code, audio, images, and video. Being “natively multimodal” allows for better understanding, reasoning, and coding capabilities. The current approach to creating multimodal models involves “training separate components for different modalities and then stitching them together.” While good at certain tasks, Google says these models “struggle with more conceptual and complex reasoning.” For Gemini, Google “pre-trained from the start on different modalities” using TPU 4 and TPU v5e. Google also announced the TPU v5p (shown below) today as its “most powerful, efficient, and scalable” AI accelerator, especially for advanced models.

9to5Google adds in its report that “To show off its “sophisticated reasoning” capabilities, Google demoed Gemini digesting 200,000 scientific research papers, filtering out the relevant ones, and then summarizing the data in an hour or so. Coding is another tentpole, with Gemini able to “understand, explain and generate high-quality code” in Python, Java, C++, and Go. There are currently three sizes in which Gemini 1.0 is available which are Gemini Ultra, Gemini Pro, and Gemini Nano with Ultra being the largest and Nano being the smallest. As for benchmarks, Gemini Ultra is the “first model to outperform human experts on MMLU (massive multitask language understanding)” at 90.0%. That benchmark “uses a combination of 57 subjects such as math, physics, history, law, medicine, and ethics for testing both world knowledge and problem-solving abilities,” with OpenAI’s offering scoring 86.4%.