Roundtable: How AI Is Changing Enterprise – Insights from Box CEO Aaron Levie and Panel

The discussion, hosted by Garry and featuring Jared Friedman, Harj Taggar, Diana, and Box CEO Aaron Levie, explores the AI revolution and its implications for enterprises.

AI Revolution and ChatGPT: Levie notes that while ChatGPT is popular, it has both value and limitations; enterprises need complete software solutions rather than relying solely on models.

Enterprise Demand for AI: From a B2B perspective, companies care about results, not the model itself. Successful AI products integrate with existing systems and automate workflows.

Future of Model Companies: Levie observes that pure model companies are rare; most AI firms sell software services (e.g., Anthropic) and rely on enterprise contracts, security, compliance, and governance.

Startup Opportunities: As intelligence becomes a commodity, startups should operate like software companies, building APIs that connect intelligence to specific verticals or functions, leveraging the decreasing cost of AI tokens.

Open‑Source Inference Models: Open models such as DeepSeek bring new use‑cases; adoption rates are still low (around 10% for chat assistants, 1% for advanced agents), but will grow as models improve.

Enterprise Attitudes Toward Underlying Models: Some firms focus only on outcomes, while others with technical expertise care about model details; over time, model differences will diminish and the focus will shift to delivered value.

AI Business Pricing and Models: Various models exist, from outcome‑based fees to usage‑based pricing; annual contracts can lock in revenue, but flexible usage‑based models are gaining traction.

Internal vs. External AI Solutions: Core, value‑defining AI applications (e.g., recommendation engines) should be built in‑house, while generic functions (HR, ERP) can be sourced from external vendors.

Security and Open‑Source Models: Enterprises are increasingly comfortable deploying external or open‑source models, leveraging cloud experience to manage security and privacy.

Economic and Societal Impact: Levie likens AI’s potential to the cloud computing wave, predicting expanded software capabilities, increased revenue, more hiring, and broader benefits for education, healthcare, and overall quality of life.