TL;DR:
- Zoho, a SaaS leader, is focusing on developing smaller AI models with 7 billion to 20 billion parameters.
- Founder-CEO Sridhar Vembu emphasized that smaller models are more effective for domain-specific challenges.
- Zoho aims to establish its own GPU infrastructure for long-term cost savings.
- The company recently introduced a suite of 13 generative AI extensions powered by OpenAI’s ChatGPT.
- Zoho is enhancing AI capabilities by integrating third-party intelligence into its AI-powered assistant, Zia.
- Zoho is undertaking a project to create its own Large Language Model (LLM), akin to OpenAI’s GPT and Google’s PaLM 2 models, overseen by Vembu himself.
Main AI News:
In the dynamic realm of Software-as-a-Service (SaaS), Zoho, a prominent player led by its founder-CEO Sridhar Vembu, has recently unveiled a strategic shift in its AI endeavors. Contrary to the prevailing trend of massive AI models that boast billions of parameters, Zoho is turning its focus towards crafting smaller AI models ranging from 7 billion to 20 billion parameters. This bold move comes as a result of the company’s recognition that smaller AI models exhibit greater efficacy in tackling specific domain-centric challenges.
Vembu elaborated on this paradigm shift during his appearance at CNBC TV18-Moneycontrol’s Global AI Conclave, stating, “We are working on models that are based on 7 billion to 20 billion parameters…we are not doing the 500 parameter models as of now. We also want to have our very own graphics processing units (GPUs) infrastructure as that is cheaper in the long term.” This deliberate departure from the ‘bigger is better’ mentality signifies Zoho’s commitment to delivering precision and cost-effectiveness in the AI landscape.
This strategic pivot follows Zoho’s recent unveiling of a suite of 13 generative AI extensions, complemented by seamless integrations into its applications, all powered by OpenAI’s ChatGPT. Furthermore, Zoho is actively bolstering its AI capabilities by seamlessly integrating third-party intelligence into Zia, the company’s AI-powered assistant. This holistic approach underscores Zoho’s dedication to delivering AI-driven solutions that cater to diverse business needs.
In a parallel development earlier this year, in June, Zoho announced an ambitious project to create its own Large Language Model (LLM), akin to OpenAI’s GPT and Google’s PaLM 2 models. Notably, this pioneering initiative is under the direct oversight of Sridhar Vembu himself, with the dedicated research and development (R&D) team diligently advancing the project’s development. Zoho’s foray into the realm of large language models underscores its commitment to innovation and pushing the boundaries of AI technology.
Conclusion:
Zoho’s strategic shift towards smaller AI models indicates a commitment to precision and cost-efficiency in addressing specific domain challenges. This move positions Zoho as a player that adapts to evolving AI trends while focusing on delivering tailored AI solutions to the market. It underscores the importance of flexibility and innovation in the competitive AI landscape.
