Nov 5, 2026, 10:14 AM ETInterview

Google Loses $500B in 12 Hours | DoorDash Buys Instacart for $17B | Kalshi Raises $1B

OpenAI's valuation decreased from $1.4 trillion to $1.25 trillion in a $20 billion funding round, with only $11 billion currently committed.
The broader AI sector experienced a $5 trillion market cap loss last week, with Alphabet dropping 12% and Intel falling 29%.
Analysts attribute valuation corrections to price discovery and liquidity constraints rather than a fundamental "AI wave cracking," noting public competitors like Google are now trading at 1.5–2 trillion.
Risk assessment for the AI trade is shifting from "euphoria" to "tempered euphoria," creating volatility in debt markets where CoreWeave debt yields hit 15%.
Broadcom's need to guarantee Anthropic's debt is cited as a specific indicator of wobbly financial conditions in the sector.
If AI demand remains strong, existing infrastructure debt (e.g., CoreWeave bonds) remains secure due to the necessity of GPUs and data centers; weakness in demand would disproportionately impact these asset-heavy players.

Anthropic's quarterly revenue gaps were approximately $5 billion (Q1), $11 billion (Q2), and $17 billion (Q3), representing a slowdown from 100% QoQ growth to roughly 50% QoQ growth.
Despite high absolute growth, the market perceives deceleration due to year-over-year growth rates projected to drop from 12x to 2x by next year.
Pricing pressure is identified as the primary factor for deceleration, with OpenAI reportedly offering equivalent products at 30% discounts, approaching the 40% discount target of open-weight models.
Anthropic faces potential challenges in commanding price premiums as competitors like Haiku 5.5 and Jev demonstrate 72% cost reductions for similar B2B workflows.
While the Fortune 500 represents only 0.1% of addressable market spend, the slowing rate of bookings suggests a local maximum in Anthropic's ability to command premiums.

DoorDash acquired Instacart for $17 billion, roughly half of Instacart's 2021 private valuation, to consolidate an estimated 80% market share in food delivery.
DoorDash reported that AI agents processed only several hundred orders against a daily volume of three million, though analysts warn this highlights a significant long-term risk of agentic disruption.
The primary threat from agents is the potential to bypass platform fees by optimizing orders across competitors or self-delivery options (e.g., finding restaurants that offer direct delivery for lower fees).
Salesforce aggressively pursued M&A, acquiring Cresta, Listen, and other AI-native companies, with SVPs often proposing acquisitions to protect jobs and modernize legacy stacks (Service Cloud).
Salesforce prioritized deals with at least $100 million in revenue to justify deploying their sales force and GTM machines, noting that sub-$100 million deals yield lower investment performance.

Kalshi raised $1 billion at a $40 billion valuation, transitioning from a prediction market to a sports gambling platform, prompting a lawsuit from Utah over state-level betting regulations.
The Canadian government invested up to $1.14 billion in Cohere to support a sovereign AI initiative, contrasting with Kalshi's regulatory challenges.
Physical Intelligence reached a $20 billion valuation despite having minimal revenue, highlighting a market preference for future growth potential over current revenue multiples compared to mature companies like Dropbox.
Robotics deployment is expected to lag behind LLM adoption due to the longer timeline for hardware integration, though foundation model companies are likely to acquire robotics firms to double their total addressable market (from 81M knowledge workers to 87M physical workers).
Security risks are escalating with the rise of AI-powered pen-testing tools (e.g., Ardex) being repurposed for offensive attacks, including recent breaches affecting IoT devices like European refrigerators.

The 90-year-old law firm Weil, Gotshal & Manges dissolved due to partner disputes, while the AI-native firm Crosby bid to acquire its assets and brand.
Crosby's model relies on standardizing transactional legal work via AI to offer lower costs, though skeptics question the sustainability of the partner equity model and the firm's potential to scale to a $10 billion exit.
The legal sector is bifurcating: high-value, rainmaker-driven advisory work remains resistant to AI, while transactional work is a viable target for automation and consolidation.

a16z faced a committee decision between three high-profile AI companies:
- Serval ($3B valuation): Preferred for high upside and potential to disrupt ServiceNow, though concerns exist regarding TAM expansion and a long "grind" to replace incumbent software.
- Cognition ($80B valuation): Viewed as the last standalone coding agent cracking enterprise; offers high revenue velocity but faces valuation risks similar to OpenAI/Anthropic.
- Base 10 ($30B valuation): An inference provider with strong product but high competition and commoditization risks.
a16z prioritizes revenue velocity for inference (Base 10) but leans toward Serval or Cognition for superior multiple expansion, with a preference for deals large enough to move their fund needles.
The committee noted that "vertical AI" plays often fail to expand TAM, instead competing for slices of existing markets held by incumbents like Salesforce or ServiceNow.