Databricks CEO: AGI Is Already Here, Why Companies Are Behind, and the Superintelligence Debate
Nov 17, 2026 · 1h 2m
More Technology podcasts · More Business podcasts
Summary
In this episode, All-In hosts interview Ali Gazi, CEO of Databricks, who discusses the company’s origins in AI research and its strategy for enterprise data governance. Gazi argues that while AGI has effectively arrived, most organizations still use AI rudimentarily due to significant barriers involving data context, security, and cost efficiency. He demonstrates how Databricks’ Genie product uses an organizational ontology to automate decision-making and reduce human blockers, effectively creating a "management moneyball" for corporate leadership.
Topics discussed
Intro: Databricks origin and AI adoption
Sponsor: Airwallex AI-native payments
Guest intro: Ali Ghodsi and Databricks growth
Berkeley roots and the AMP lab
Early bets on enterprise AI and data
The big data movement and cloud computing
Betting on cloud, AI, and open source
Netflix Prize and the XPRIZE era
Human feedback and Bay Area culture
Defining superintelligence vs. AGI
The compute limits of superintelligence
Enterprise AI: beyond glorified search
The need for human oversight in agents
Data access and liability in enterprises
Cybersecurity risks of open AI access
Managing exponential token costs
CFO concerns and operational efficiency
Anthropic's rapid product expansion
Organizational autonomy and innovation
Governance and secure data dashboards
Agentic tech and org chart latency
Cross-functional AI for speed
Company world models and digital natives
Demo: Running Databricks with AI agents
Genie: Ontology and company graph
AI inbox and background work
Analyzing leadership and work-life balance
Deep dive: Culture metrics and actions
Sales vs. Product culture insights
Org structure and reporting lines
Engineering cost analysis
Data-first culture and theory testing
Commit activity and automated factories
Customer meeting analysis
The CEO as detective
Reducing human blockers
Accountability and data-driven resolution
Major decisions bot and decision speed
Identifying all-stars and decision makers
Accountability for missed targets
Culture principles and leader alignment
Global culture insights and transparency
Staying private: Strategy and alpha
New markets and growth paths
Employee sentiment and secondary market
Equity liquidity and holding skills
Competition: Snowflake and Anthropic
Developer resources and profitability
Stock volatility and employee pay
Distraction of daily stock checks
Superintelligence: Context and RSI
Coding agents and existential crisis
Bostrom's evolving view on risk
Scaling laws and compute requirements
Conditions for true superintelligence
Domain-specific RL and pre-training
Cybersecurity: ScriptKitty example
Physical vs. digital security
AI defense vs. human hackers
Asymmetric attack risks
Latent bugs and code scanning
Shadow patching and vulnerability cleanup
The EA community and their beliefs
Dario Amodei and false hope
Cyber risks vs. RSI evidence
Social impact and news manipulation
Healthcare breakthroughs and Regeneron
Curing cancer: timelines and trials
Drug discovery and Crisis Text Line
Economic impact and sci-fi influence
GDP growth and the 10-year horizon
Internet analogy and Robert Morris
Closing remarks
More from All-In with Chamath, Jason, Sacks & Friedberg
Anthropic Goes Public, Nvidia in Crisis, Markets Tank, Open Source Crackdown?
Nov 20, 2026
Jensen Ousted as Nvidia CEO: "Spy" Scandal, Export Restrictions, Market Impact
Nov 14, 2026
Freedom vs. Safety: Axon CEO Rick Smith on the Future of Policing
Nov 10, 2026
Midterm Blood Bath, $4T AI Sell Off, Data Center Freeze, What's Next for Trump
Nov 7, 2026
Anthropic S-1 Shocker, Greatest Business Ever? Google's $500B Wipeout, Midterm Predictions
Oct 30, 2026
Listen ad-free on Castria