Google Delays its GPT-4 Competitor Gemini Until Next Year – Report
Google has reportedly postponed the launch of its highly anticipated Gemini multimodal AI model. According to sources familiar with the matter, Google has decided to delay the planned launch events for Gemini until January of next year. The announcement came directly from Google CEO Sundar Pichai.
The delay was prompted by concerns over Gemini’s ability to effectively respond to queries in languages other than English. Although Gemini has shown promising results and has even reached a level comparable to GPT-4 in some aspects, it still needs further improvement in terms of multilingual capabilities. While smaller versions of Gemini are currently being tested, the development of the larger Gemini model is still ongoing.
Previously, it was leaked that the cloud version of Gemini would be delayed. Now, the Bard chatbot and other AI-powered products won’t receive the anticipated Gemini upgrade until next year.
Looking ahead, Sundar Pichai recently shared Google’s plans to roll out a series of next-generation models based on Gemini in the coming year. The success of Gemini’s launch is pivotal for Google, as it aims to demonstrate its ability to compete with, and possibly surpass, OpenAI. Additionally, Gemini sets the stage for a new era of the internet, in which information flow shifts from traditional search and the World Wide Web to chatbots.
Gemini’s triumph would also serve as a message to the industry that OpenAI’s GPT-4 is not the ultimate benchmark. It would highlight the potential for significant advancements using the underlying Transformer technology and the scaling principle of incorporating more data and training. With Google’s access to vast amounts of data and computing power, it has the advantage, but partnerships such as Microsoft’s collaboration with OpenAI have posed challenges.
Since March 2023, no company, whether it be a tech giant or innovative startup, and regardless of whether it operates in closed or open source, has managed to release a language model that can match GPT-4. The market currently offers language models that are on par with GPT-3.5, showcasing the ease of achieving that level of sophistication.
The reason behind GPT-4’s advancements can be attributed to its larger, more complex, and expensive architecture. GPT-4 consists of a network of interconnected AI models known as a Mixture of Experts rather than a single large model. It is highly likely that Google’s Gemini is based on a similar concept.
However, such a complex architecture comes with the drawback of high costs for inference, which is why OpenAI is striving to reduce prices with models like GPT-4 Turbo, even if it means compromising on quality.
As Google faces this delay in the launch of Gemini, it is clear that the competition between Google and OpenAI is fierce. The race to develop the most advanced AI models continues, and the success of Gemini’s eventual release will undoubtedly have significant implications for both companies and the industry as a whole.
In conclusion, Google’s decision to postpone the launch of Gemini until next year showcases the company’s commitment to ensuring a high-quality and versatile AI model. This delay acknowledges the need for further improvements, particularly in terms of multilingual capabilities. As the battle between Google and OpenAI unfolds, the release of Gemini could shape the future of AI, information retrieval, and the way humans interact with computers.
However, for now, we eagerly await the arrival of Gemini in January, hoping it will deliver on its promise to revolutionize the AI landscape and solidify Google’s position in the industry.
