Databricks CEO: AI is here, focus on data context

In a candid discussion, Databricks CEO and co-founder Ali Ghodsi shared his perspective on the current state of artificial intelligence, asserting that artificial general intelligence (AGI) has already arrived. Ghodsi, speaking at a Bloomberg Tech event, challenged the common narrative around AI development by emphasizing that the focus should not be on making AI models 'smarter' in a vacuum, but rather on providing them with the necessary data context to be more effective and productive.

StartupHub data

Companies working on this

Profiles of the companies named in this story, with funding and a one-liner from our database.

Databricks

$190.0B

A unified data analytics and AI platform built on the lakehouse architecture.

Bloomberg L.P.

$42.6B

Global financial, software, data, and media company providing real-time information and analytics.

Microsoft

A global technology leader providing software, cloud services, AI, and devices for individuals and businesses.

Unilever

A multinational consumer goods company that owns over 400 brands.

Ghodsi's View on AGI and AI Evolution

Ghodsi stated that he frequently asks audiences how many believe AGI has already arrived, and typically around 10% of the audience raises their hands. He then asks how many believe the AI models they use daily are smarter than the people they interact with, and a much larger percentage, around 90%, raise their hands. Ghodsi views this as evidence that AI is already highly capable, but its true potential is unlocked when it is properly contextualized. He argues that instead of pursuing increasingly complex models, the industry should concentrate on feeding these models with relevant data. "We don't need AI to get smarter. It just needs to be given context," Ghodsi explained. He believes that by providing AI with the right data, its ability to perform tasks and answer complex questions can be significantly enhanced.

The full discussion can be found on Bloomberg Technology's YouTube channel.

The Importance of Data Context for AI

Ghodsi highlighted Databricks' strategic focus on this aspect of AI development. The company is actively working on solutions that allow businesses to infuse their proprietary data into AI models, thereby enabling more relevant and actionable outputs. He drew an analogy, stating that while AI models are sophisticated, they are like brilliant students who need the right textbooks and curriculum to excel. The Databricks platform aims to be that comprehensive educational resource for AI. Ghodsi revealed that Databricks has been actively engaging with its clients and has conducted numerous studies, finding that companies are increasingly looking to integrate their vast datasets with AI to drive business value. He noted that while AI can generate code and automate tasks, the real power comes from grounding these capabilities in specific business contexts and data.

Databricks' Strategy and Market Position

The CEO also touched upon Databricks' partnerships with leading AI companies such as Alphabet Inc. (NASDAQ:GOOGL) subsidiary Google DeepMind (Gemini), Microsoft (NASDAQ:MSFT)-backed OpenAI, and Anthropic. He emphasized that Databricks provides the data foundation for these models, enabling them to be trained and fine-tuned with specific organizational data. Ghodsi stated that Databricks is not just building AI models but is focused on creating a comprehensive data intelligence platform that can serve as the backbone for AI adoption across enterprises. He pointed to a slide showcasing Databricks' customer base, including major companies like AT&T, Rivian, Adidas, Mercedes-Benz, Unilever, Virgin, and Bayer, illustrating the broad applicability of their data and AI solutions.

Ghodsi concluded by stating that while the current market might be experiencing a slowdown in IPOs, particularly for software companies, Databricks is well-positioned due to the fundamental need for data management in the age of AI. He believes that as more software is generated and more AI agents are deployed, the demand for robust data infrastructure like that provided by Databricks will only grow. He also mentioned that in the next nine months to two years, the amount of software being written will dwarf all previous human history, underscoring the critical role of data in this AI-driven software creation process.