Published on November 17, 2023, 10:30 am

Google Delays Launch of Gemini, Its Multimodal AI Model, to Compete with OpenAI’s GPT-4

Google’s plan to counter OpenAI’s advancements in the field of artificial intelligence (AI) has hit a roadblock as the launch of its multimodal model called Gemini has been delayed. Initially expected to be made available to cloud customers in November, insiders now report that it won’t be ready until the first quarter of 2024.

While smaller sub-models of Gemini have already been tested by external parties, the main challenge lies in developing a large model that can compete with OpenAI’s GPT-4. There are doubts as to whether Google will be able to match GPT-4’s quality or even surpass it.

The delay in the cloud version launch could also indicate Google’s intention to introduce this new technology first in its consumer products. However, Google has not made any official comments on this matter. During a public event, CEO Sundar Pichai merely mentioned that Google aims to release Gemini 1.0 as soon as possible and ensure its competitive and state-of-the-art nature. Pichai further emphasized that this next-generation AI model will serve as a foundation for a series of subsequent models set to be released throughout 2024.

This setback presents a challenge for Google in its competition against Microsoft and OpenAI. While Microsoft successfully markets OpenAI technologies to enterprise customers through its cloud services, Google is struggling to reach consumers with its competing solution known as Bard. Additionally, due to the limited number of users, Google faces difficulties in creating an extensive feedback database required for continuously improving their language models offered through the PaLM suite on their cloud platform. As a result, OpenAI’s GPT-4 outperforms Google’s offering in terms of quality.

To develop Gemini, Google is leveraging the AI capabilities from both its Deepmind subsidiary and Google Brain division. However, integrating the talents and models from these different teams has proven complex. Sergey Brin, co-founder of Google, is actively involved in guiding the developers for four to five days a week but does not officially make decisions.

In terms of computing power, Google’s earlier advantage is now overshadowed by OpenAI’s partnership with Microsoft, which is expected to offer even greater computational capacity along with its own chips next year. Both Google and OpenAI also face a shortage of AI experts, with reports suggesting that OpenAI has been spending millions to attract top talent away from Google.

The stakes are high for Google as delaying the launch of Gemini could put them at a disadvantage compared to Microsoft in the cloud business. Furthermore, advancements in AI features on popular platforms like YouTube and improvements in Google Assistant may have to be put on hold as OpenAI’s ChatGPT continues to gain ground. On top of that, Gemini holds potential benefits for Google’s advertising business. Advertisers can use Gemini’s longer memory for user interactions to create new variations of their most successful campaigns from recent months.

The delay may be seen as an obstacle for Google, but they remain determined to ensure that Gemini is competitive and state-of-the-art when it finally becomes available. The race between tech giants in advancing generative AI models intensifies as each company strives for dominance in this groundbreaking field.


Comments are closed.