Software Winners & Losers in the Age of AI (w/Alex Rubalcava & Paul Bricault) | #626

Episode Summary

Executive Summary: The episode explores how AI is reshaping venture capital, software defensibility, and workforce dynamics. Amplify LA’s Paul Bricot and Alex Rubelkava argue that AI will commoditize low-complexity software and entry-level tasks, while strengthening businesses with data moats, regulatory barriers, or deep workflow integration. They also discuss pre-seed investing, portfolio management, secondary sales, QSBS tax benefits, and opportunities in healthcare, defense, industrial automation, and space.

Main Topics: AI’s impact on software companies (Priority: 5/5): The guests debate which software businesses are vulnerable to AI disruption and which remain durable. Mission-critical systems of record, regulated workflows, and products tied to money, law, or physical operations are seen as defensible, while basic SaaS and “nice-to-have” tools face pressure. What Amplify looks for in startups (Priority: 5/5): Amplify focuses on pre-seed B2B enterprise, especially vertical AI, healthcare, fintech, defense, space, robotics, and industrial automation. They prioritize companies with proprietary data, regulatory constraints, complex workflows, or hardware/software integration that makes copying difficult. AI-driven productivity and workforce displacement (Priority: 5/5): The discussion highlights dramatic productivity gains from AI tools and simultaneous job compression, especially in SDR, CSR, legal, HR, content, analytics, and entry-level engineering roles. The guests argue future workers must be customer-facing and AI-fluent. Venture capital process and pre-seed sourcing (Priority: 4/5): They explain how pre-seed investing relies less on AI-driven sourcing and more on dense networks, founder relationships, and judgment because many opportunities are off-market or not yet visible publicly. Investment decisions can happen in days or stretch to weeks. Portfolio management, secondaries, and power-law returns (Priority: 4/5): The guests discuss taking secondary sales to de-risk winners, especially once investments reach 10x or more. They stress that venture outcomes are power-law driven, so disciplined trimming helps manage fund life cycles and emotion. Tax advantages and investor base (Priority: 3/5): QSBS expansion makes venture more attractive to taxable investors, family offices, and high-net-worth individuals. They note that early-stage venture is often more accessible than people think, especially outside top-tier oversubscribed firms. Unusual pitches and startup pivots (Priority: 3/5): The conversation ends with memorable stories about bizarre startup ideas and major pivots that eventually became successful businesses, reinforcing that early-stage company evolution is unpredictable and can create outsized returns.

Key Arguments: AI is most dangerous to software that is easy to replace, lightly regulated, and not mission critical; systems of record and workflow-embedded products are much safer. Businesses with proprietary data, physical-world constraints, or regulatory moats are better positioned to survive and even benefit from AI. AI is accelerating the creation of product and code, with some teams reportedly shipping many times more output than before. The labor market is being reshaped as AI replaces or compresses entry-level work, pushing workers to become more adaptable, customer-facing, and AI-literate. Pre-seed VC is still relationship-driven and difficult to automate because the best opportunities often happen before they are visible in public datasets. Great venture portfolios require active management of winners through selective secondaries, rather than holding everything to the end. QSBS and other tax considerations materially improve the after-tax attractiveness of venture investing for taxable investors. AI adoption is starting to correlate more clearly with company revenue growth, suggesting a widening performance gap between fast and slow adopters.

Data Points: Portfolio revenue growth and headcount: Many companies have grown revenue multiple times while keeping headcount flat or lower than at their last fundraise - Used to illustrate AI-enabled efficiency and decoupling of growth from expense Engineering productivity: 5x code output - Portfolio companies reportedly shipping five times as much code per engineer/team using AI tools Pull requests: 4x increase - Balto data shared by Alex Rubelkava after adopting cloud code Project output: 5x in 100 days - Balto engineering productivity improvement after AI adoption Trace staffing reduction: 22 SDRs and 18 CSRs down to 1 SDR and 1 CSR - Company used co-agents to replace lower-level sales/customer support roles Trace revenue growth: 50% - Same company grew revenue while reducing staffing through AI automation Placer customer turnaround: 90 minutes vs. 3 weeks - AI front end enabled a CEO to build a presentation and save a multimillion-dollar transaction quickly AI and entry-level jobs: ~30% drop - Referenced New York Times story about entry-level jobs for college grads Ramp customer study: Top quartile of AI spend doubled revenue; bottom quartile was flat - Illustrates correlation between AI adoption and growth Pre-seed round size: $1M–$2M historically; now often $2M–$3M - Shows inflation in early-stage capital requirements Extreme pre-seed outlier: $400M at a $2B valuation - Example of very large pre-seed round for a pre-revenue company Funded company ARR: Over $2M ARR - AI scribe for nurses already generating meaningful recurring revenue QSBS holding period: Partial credit under 5 years - Mentioned as part of recent QSBS expansion Fund size behavior: Pre-seed funds over $100M - Noted as a new market phenomenon, though not necessarily sensible Decision timeline: A few days to 8–9 weeks - Range of time from first founder meeting to investment decision

Pivotal Quotes: "If you're a product that a business runs on, I think that it's going to be a really poor use of money and scarce resources like developer time for a company to try to save a few hundred thousand bucks to rip out a piece of software that their company runs on." — Alex Rubelkava: Explaining why mission-critical software is more durable than low-value SaaS "AI is almost always the answer to that why now question." — Alex Rubelkava: Describing why startups can be built differently today than in prior years "As soon as the company hits 10x return, we're not taking all our capital off the table. But we're at that point going to consider taking capital off the table when there is an opportunity." — Paul Bricot: Discussing Amplify’s discipline around secondary sales and risk management

Implications: AI is creating a sharper divide between defensible, workflow-embedded businesses and easily replicated software, while also compressing labor demand at the entry level. For investors, networks, data moats, and disciplined portfolio management matter more than ever.