internet giant Google, Today it took its biggest step against the OpenAI threat. Minutes ago big tongue model “Gemini” was officially introduced.
Google, Developed by OpenAI ChatGPTAfter the explosion of , he turned his entire focus to the productive artificial intelligence side. The company, which has been investing in the field of artificial intelligence for years, but is very uneasy about the success of ChatGPT, decided to create a productive artificial intelligence-based chat bot in order not to be left behind. We first encountered it with the chat bot Bard, which is also open to Turkey.. Some made by employees, “it was very rushedBard, who started to become a part of our lives under the criticism of ” He was the first pioneer of the new era for Google. Today, the company has finally announced the very advanced large language model that it has been working on for a long time, which it had previously announced. Gemini made the official announcement for. Gemini has three divisions within itself at this stage. leaving. Gemini Nano Built directly into Android phones, it is designed to work without an internet connection. Gemini Pro, is a more comprehensive model than Nano in terms of internet access and will power many artificial intelligence services, especially Google Bard (in fact, the trial for Bard started today). If at the top Gemini Ultra Ultra, the most powerful large language model Google has ever created, is an option designed mostly for data centers and mostly enterprise applications. located in my location.
YOU MAY BE INTERESTED IN
According to the statement, Gemini Pro can be used via Google Generative AI Studio or Vertex AI in Google Cloud. It will be available starting from December 13th. The full release of Gemini Ultra will be in the new year, while Gemini Nano will first enter Google’s Pixel series of new generation smartphones. Currently only available in English However, the big language model will later support many other languages, including Turkish. As it will not come as a surprise in the future, it will also be integrated into Google Search.. Generative artificial intelligence based “Call” The internet giant, which has been testing and improving its infrastructure with its existing systems for a long time, said Gemini was OpenAI’s most advanced large language model in 30 out of 32 benchmarks. It says it surpasses GPT-4, which actually seems pretty significant.
Google, which is far behind ChatGPT with Bard and seems to close the gap with Gemini and even get ahead in some tasks, states that Gemini Ultra can easily handle surprisingly complex tasks. The new big language model, which does not separate analysis / comprehension categories such as sound, text and visual, but rather blends them together, Of course it will get better over time And understand input more accurately And will start to give more detailed answers.
AlphaCode 2 It is built on the basis of creating a code called and creates an impression that will please the software developers. Gemini, Trained on Google’s own Tensor units and built on the company’s previous models such as PaLM can work both faster and less costly. The real star of the show is the widest version, Gemini Ultra, at MMLU, which combines 57 subjects such as mathematics, physics, history, law, medicine and ethics to test knowledge and problem-solving abilities, with a score of 90.0 percent according to Google. (Massively multitasking language understanding) It is highlighted as the first model to outperform human experts.
Its release today is directly “1.0The system made with the code “ It’s really important for the future of Google looks and it already reveals great potential. The company, which will use Gemini to carry its many different systems into the future, can become the dominant power in the productive artificial intelligence side thanks to its great financial power and trained personnel. It may be the only name that can compete with OpenAI in the field.
Of course, other companies are also working on artificial intelligence. For example, software giant Microsoft is already developing together with OpenAI, while the latecomer in the process As far as we hear, Apple is working behind closed doors to release iOS 18.. Names such as Meta and Amazon are also making huge investments in the field, but at the top of the competition, at this stage, only Google And OpenAI It is located.