Summary of Alex Sacerdote - How to Invest Through Technology Cycles - [Invest Like the Best, EP.477]

Overview of Invest Like the Best with Alex Sacerdote

Patrick O’Shaughnessy speaks with Alex Sacerdote, founder of Whale Rock Capital Management, about how he invests through technology cycles and why he believes AI is the biggest technology S-curve he has ever seen. The conversation centers on Anthropic, the broader AI stack from chips to applications, and how Whale Rock uses its framework of S-curve positioning, competitive advantage, and underappreciated earnings power to find winners early. Sacerdote also explains how his firm has expanded into private markets and built a research-driven “learning machine” around tech investing.

The Core Investment Framework

Sacerdote’s process is built on three questions:

1. Where is the company on the S-curve?
   He wants to buy when adoption is still early but inflecting.
2. Does it have a durable competitive advantage?
   He looks for moats such as network effects, scale, brand, critical IP, or industry standard status.
3. Is earnings power underappreciated?
   The market often underestimates how quickly profits can compound once a technology is truly adopted.

His central belief: in technology, earnings can grow nonlinearly once a product hits the right stage of adoption and the right moat is in place.

Why Anthropic Is His Highest-Conviction Position

Sacerdote lays out why Anthropic has become Whale Rock’s highest-conviction private investment:

• The AI market is consolidating into a few leading foundation model players.
• Early fears that the model layer would be a commodity have faded as differentiation has become clear.
• Anthropic has emerged as a strong enterprise-focused leader, while OpenAI has been stronger in consumer.
• Google remains a major competitor, but the market increasingly looks like a three-horse race.

The Coding Breakout

A major part of his thesis is the rapid growth of AI coding tools:

• Claude Code and related tools have shifted from helper software to something close to agentic coding.
• Sacerdote argues coding alone could represent a half-trillion-dollar market because it addresses a massive global labor pool.
• He sees coding as the biggest near-term unlock for AI monetization and labor replacement.

Why He Thinks Anthropic Has Moats

He points to several advantages:

• Critical intellectual property
• Strong enterprise brand
• Scale and escape velocity
• Recursive improvement, where better code generation improves the model, which then improves code generation again
• A growing ecosystem of tools around the model, including SDKs and orchestration layers

The AI Stack: Chips, Models, Applications

Sacerdote sees AI as a new compute paradigm that creates a new stack:

1. Infrastructure and Chips

He believes the bottom of the stack remains one of the best ways to play AI because:

• Compute demand is exploding
• There is not enough supply in the world
• Hardware is being pushed to its physical limits
• Many parts of the data center are being decommoditized

2. Foundation Models

He believes this layer is now clearer than it was two or three years ago:

• Many startups have disappeared
• The market is converging toward a few durable leaders
• Quality, compute access, and feedback loops matter more than many expected

3. Applications

He is much more cautious here:

• Most AI software products are still early
• Many incumbents have not yet built compelling AI revenue
• It is still unclear which applications will develop durable moats

Why He Is So Bullish on Chips and Infrastructure

Sacerdote spends a lot of time on the downstream beneficiaries of AI compute growth.

What Changed in the Data Center

For decades, data centers were largely commoditized. AI has changed that:

• Workloads are growing rapidly, often 10x per year
• Servers are more complex, more expensive, and more specialized
• Every layer of the stack is seeing demand pull-through

Examples He Cites

He highlights companies and categories such as:

• NVIDIA
• TSMC
• SK Hynix
• ASML
• Celestica
• Corning
• Delta Electronics
• Advanced Energy
• PCB and memory suppliers
• High-bandwidth memory and advanced networking components

His broader point is that AI is creating not just demand growth, but a hardware renaissance with rising ASPs, tighter supply, and better margins.

Why He Sold Most of His Software Exposure

Five years ago, Whale Rock had much more software exposure. That changed as AI matured.

Sacerdote says the firm initially thought AI would strongly benefit software incumbents, but they changed their view as:

• AI products from incumbents were weaker than expected
• Monetization was limited
• AI budgets shifted toward model usage and infrastructure
• Some software companies are under pressure on pricing, hiring, and seat growth

His Concern About Enterprise Software

He argues that many incumbents face three problems:

• AI features are often too weak to move the needle
• Customers may prefer spending on AI tokens instead of legacy software
• New AI-native competitors could emerge and attack entrenched workflows

The One Area He Still Watches Closely

He is not dismissing software entirely. He thinks the next phase may involve:

• AI agents operating inside existing tools
• Network-based software like Slack, CRM, and HR systems becoming more entrenched
• Incumbents potentially surviving if AI becomes a layer on top rather than a replacement

How He Thinks About S-Curves in Practice

Sacerdote gives a detailed view of how he uses S-curves to invest:

• Technologies often exist for years before adoption goes vertical
• The key is removing adoption barriers like cost, usability, infrastructure, or range
• He looks for the moment when the market shifts from “interesting” to “must-have”

Examples He Uses

• iPhone / smartphone adoption
• Tesla and electric vehicles
• AWS and cloud computing
• AI and chat-based interface adoption

He stresses that investors should care not just about the start of the S-curve, but also:

• How tall the curve is
• How long it lasts
• Whether it stalls at 30% to 40% penetration
• Whether a company keeps its lead as the category matures

What Makes a Technology Leader Win

Sacerdote believes category leaders often separate because of:

• Network effects
• Scale advantages
• Becoming the industry standard
• Strong brand
• Proprietary IP
• First-mover advantages
• Ecosystem lock-in

He emphasizes that the best companies often combine several of these at once.

How Whale Rock Does Research

Whale Rock’s process is highly hands-on and relationship-driven.

Research Method

• 2,500–3,000 face-to-face management meetings per year
• Deep scuttlebutt-style research inspired by Philip Fisher
• Conversations with customers, competitors, suppliers, and operators
• Heavy use of AI for notes, summaries, and speed, but not as a substitute for judgment

His View of AI in Research

AI helps analysts:

• get up to speed faster
• organize information
• draft notes and summaries

But it does not replace:

• original insight
• pattern recognition
• judgment about what matters
• the ability to forecast the future

He says the analyst’s role is still to provide the “wisdom on top,” not just reporting.

Whale Rock’s Product Evolution

The firm has expanded beyond its original long-short hedge fund model:

• Long-short fund as the original core product
• Long-only fund launched in 2020
• Private investing formalized over time
• Hybrid fund for higher private exposure
• Mega Cap Tech fund focused on the largest global tech companies

Why Mega Cap Tech

Sacerdote argues that many investors structurally underweight mega-cap tech because:

• they assume large caps have less alpha
• they are overweight private and small/mid-cap exposure
• they miss the fact that the largest tech leaders can still compound dramatically

Biggest Risks He Sees

Despite the bullish tone, Sacerdote is explicit about risks:

• Regulatory backlash against AI
• Public negativity toward AI adoption
• Slower model improvement, which could let open-source models catch up
• Competitive failures among major AI players
• The possibility that one or more leaders lose their position, reducing compute demand

He notes that even if models slow down, adoption can still continue — but the investment implications may change.

Notable Themes and Takeaways

• AI is not just another software cycle; it is a new compute paradigm
• The most compelling opportunities may currently be in chips, infrastructure, and foundation models
• Many software incumbents may face pressure before they benefit
• The market often misunderstands how long exponential growth can last
• Whale Rock’s edge comes from a deep, repeatable research process and long experience across multiple tech cycles

Closing Reflection

The episode ends on a personal note, with Sacerdote describing his father as the kindest thing anyone ever did for him. He credits his father — a smart, humble, and respected investor — for mentoring him and helping build Whale Rock in its early years. It reinforces a broader theme of the conversation: great investing is not just about models and markets, but about long-term learning, judgment, and relationships.