“Zoho CEO Vembu Unveils Development of Substantial Yet Compact AI Models”

3

In a strategic shift, Zoho, the Software-as-a-Service (SaaS) behemoth, is pivoting towards the development of smaller artificial intelligence (AI) models, revealed its founder-CEO, Sridhar Vembu. This alteration stems from Zoho’s findings that smaller AI models outperform larger ones, particularly when addressing specific domain problems.

Zoho Ventures into Smaller AI Models, CEO Vembu Emphasizes Focus on Specific Domain Problems
Zoho Ventures into Smaller AI Models, CEO Vembu Emphasizes Focus on Specific Domain Problems

In a strategic shift, Zoho, the Software-as-a-Service (SaaS) behemoth, is pivoting towards the development of smaller artificial intelligence (AI) models, revealed its founder-CEO, Sridhar Vembu. This alteration stems from Zoho’s findings that smaller AI models outperform larger ones, particularly when addressing specific domain problems.

Vembu highlighted this paradigm shift during CNBC TV18-Moneycontrol’s Global AI Conclave, mentioning Zoho’s focus on crafting AI models ranging from 7 billion to 20 billion parameters. The company has opted out of pursuing 500 parameter models, opting instead for a narrower scope. Moreover, Vembu outlined Zoho’s intention to establish an in-house graphics processing units (GPUs) infrastructure, foreseeing its cost-effectiveness in the long run.

The revelation coincides with Zoho’s recent unveiling of 13 generative AI extensions, powered by OpenAI’s ChatGPT, integrated into its suite of applications. Concurrently, the Chennai-based company is intensifying its AI capabilities by amalgamating third-party intelligence into Zia, its AI-driven assistant.

Furthermore, Zoho announced an ambitious project earlier this year: the development of its own Large Language Model (LLM), akin to industry-leading models like OpenAI’s GPT and Google’s PaLM 2. Spearheaded by Vembu himself, the company’s research and development (R&D) team is actively engaged in advancing this initiative.

For context, a Large Language Model (LLM) represents a specific type of AI model engineered to comprehend natural language intricacies. These models possess the capability to process and generate textual content, empowering tasks such as language translation, summarization, and Q&A functionality.

Zoho’s strategic shift towards smaller AI models underscores a nuanced approach to problem-solving within designated domains. This departure from the pursuit of larger models reflects the company’s recognition of the efficacy and efficiency of tailored solutions for specific challenges.

As the tech landscape continues to evolve, Zoho’s foray into crafting smaller AI models and the ongoing development of its LLM project signal the company’s commitment to innovation and its endeavor to deliver more targeted, contextually relevant AI solutions to its users.

Sources By Agencies

About The Author

3 thoughts on ““Zoho CEO Vembu Unveils Development of Substantial Yet Compact AI Models”

Leave a Reply

Your email address will not be published. Required fields are marked *