Skip to content
The Twenty Minute VCThe Twenty Minute VC

Nabeel Hyatt, GP @ Spark Capital: To Win in AI, Investors Need to Change Their Approach | E1255

Nabeel Hyatt is a General Partner @ Spark Capital, one of the leading firms of the last decade with portfolio companies including Twitter, Anthropic, Coinbase, Affirm, Discord, Deel and more. ---------------------------------------------- Timestamps: (00:00) Intro (00:50) Adapting Investment Mindset for the Age of AI (04:17) Adaptation of Venture Firms for Rapid Creativity & Innovation (10:00) Heuristics for Evaluating the Explosion of Startups (12:53) Venture's Shift from High Margin to Commoditized Industry (18:07) Effective Coaching & Mentorship Within a Partnership (19:29) Doubting Yourself as an Investor During the COVID Era (22:19) VC Service vs. Founders Who Don’t Need Help (25:23) What Happens When Nabeel Loses Faith in a Founder? (29:40) Sustainability of Value in a Rapidly Changing World (32:46) Why Focus on Product When It’s the Most Transient Element? (36:19) How Many Companies Does Nabeel Meet a Week? (39:48) Regrets Over Turning Down Deals Due to Price (42:37) Secondary Market Activity & The Need for Liquidity (44:23) Three Categories of AI Startups & How to Approach Them (50:53) Does DeepSeek Change OpenAI and Anthropic's Strategy? (54:03) The Future of the Agent Economy and Replacing Labor (57:39) Biggest Loss (01:00:20) Key Challenges Facing Europe Today (01:02:09) How to Approach the Path to Raising for AI Startups (01:05:36) Quick-Fire Round ---------------------------------------------- In Todays Show with Nabeel We Discuss: 1. The Rules of Investing: - What have been Nabeel’s biggest lessons on price sensitivity? When did he not pay up and with the benefit of hindsight, wish he had of paid up? - How important is ownership to Nabeel and Spark? How does Nabeel think about reserve investing and doubling down? - Why does Nabeel not engage in secondary markets? How does Nabeel think about when is the right time to sell? - Why does Nabeel think the majority of market sizing is total BS? 2. The Venture Landscape: Run by Principles and Broken: - Why does Nabeel believe this generation of AI investing will require a different mindset to the one that made VCs successful over the last decade? - Why does Nabeel believe that venture is currently run by principals and associates? Why is that such a problem? - Why does Nabeel believe that the majority of venture firms today are dead but do not know it yet? - What does Nabeel believe happens to the mega multi-stage firms who have raised billions and billions? 3. How to Win the VC Game in a World of AI: - Infrastructure, models, apps: where does Nabeel believe the most value will accrue in the next decade of AI investing? - What does Nabeel mean when he says there are three categories of AI apps today? Where does Nabeel believe the most valuable will be built? - Does Nabeel believe Deepseek hurt or helped the future for Anthropic? How could Anthropic be a $100BN company from this point? - What does no one see about the next 10 years of AI that everyone should see? ----------------------------------------------- Subscribe on Spotify: https://open.spotify.com/show/3j2KMcZTtgTNBKwtZBMHvl?si=85bc9196860e4466 Subscribe on Apple Podcasts: https://podcasts.apple.com/us/podcast/the-twenty-minute-vc-20vc-venture-capital-startup/id958230465 Follow Harry Stebbings on Twitter: https://twitter.com/HarryStebbings Follow Nabeel Hyatt on Twitter: https://twitter.com/nabeel Follow 20VC on Instagram: https://www.instagram.com/20vchq Follow 20VC on TikTok: https://www.tiktok.com/@20vc_tok Visit our Website: https://www.20vc.com Subscribe to our Newsletter: https://www.thetwentyminutevc.com/contact ----------------------------------------------- #20vc #harrystebbings #nabeelhyatt #sparkcapital #venturecapital #investing #ai #openai #startups

Nabeel HyattguestHarry Stebbingshost
Feb 2, 20251h 15mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Nabeel Hyatt: Why Venture Must Abandon Playbooks To Win In AI

  1. Nabeel Hyatt argues that the venture playbook honed in the B2B SaaS boom—heavy on spreadsheets, pattern‑matching, and industrialized processes—is misaligned with today’s AI-driven, highly uncertain environment.
  2. He believes early-stage venture must return to being an artisanal, founder-centric craft that embraces “mysteries” rather than “puzzles,” with small, deeply product-native teams making subjective, long-term bets.
  3. Hyatt critiques current VC incentive structures (principals chasing markups, packaging companies for the next round) and emphasizes genuine service to founders, taste, and curiosity over coverage, branding, and heuristics like ARR thresholds.
  4. In AI specifically, he outlines a framework of adaptation, evolution, and revolution, backs both foundational models (e.g., Anthropic) and applications, and stresses that enduring value will come from continuous reinvention, data exhaust, and expert workflows, not thin GPT wrappers.

IDEAS WORTH REMEMBERING

5 ideas

AI demands a shift from metrics-driven ‘puzzles’ to embracing ‘mysteries.’

The B2B SaaS era rewarded industrialized venture—checklists, ARR milestones, and pattern-matching. In AI, models and markets change too fast for that; investors must accept fog-of-war, rely on judgment, and get comfortable making calls without clear spreadsheets.

Firm structure and incentives now matter more than playbooks.

Hyatt argues that when partnerships balloon and firms become associate/principal-led, incentives tilt to markups and promotions instead of long-term outcomes. Winning in AI requires small partnerships where every GP writes checks, uses products, debates truthfully, and isn’t optimizing for internal politics.

Stop over-relying on revenue and growth heuristics as quality signals.

In AI, products can hit $10M ARR in months yet be dead two years later. Speed to revenue is no longer a reliable proxy for durability; investors must probe for defensibility, depth of insight, and a founder’s ability to continuously reinvent in response to rapid model and market shifts.

Use product analysis to understand founders, not to rate features.

Hyatt doesn’t back companies solely because the product looks good; he uses the product to ask how and why it was built. The goal is to distinguish hucksters from real executors by interrogating the decisions embedded in the product and how the team reasons about taste, tradeoffs, and users.

Avoid over-capitalizing companies, even when money is cheap and plentiful.

He believes too much capital can structurally damage a startup—distorting hiring, slowing learning, and raising expectations to impossible levels. Growth investors playing a deployment game may destroy potential $5B outcomes by pre-emptively shoving in oversized rounds before a company is ready.

WORDS WORTH SAVING

5 quotes

We are in the industrialization of startups playbook land where everybody's trying to churn out some piece of ridiculous arbitrage every week in order to get through the end of their incubator and raise their seed round.

Nabeel Hyatt

The industry today is run basically by principals, associates, and junior GPs… A principal is not actually waiting for an exit, they just want a promotion, man.

Nabeel Hyatt

Why are we building a bunch of playbooks if the whole thing is about exceptions?

Nabeel Hyatt

There is absolutely a belief, for me at least, that too much capital can mess up a company.

Nabeel Hyatt

Venture is an incredibly simple business. Very hard, but very simple.

Bruce Dunlevie, as quoted by Nabeel Hyatt

How AI changes venture investing from metrics-driven puzzles to uncertain mysteriesStructural problems in modern VC: incentives, principals chasing markups, and ‘packaging’ behaviorDesigning venture firms for the AI era: small partnerships, deep product usage, and serviceHyatt’s framework for AI startups: adaptation, evolution, and revolutionEvaluating founders: taste vs execution, conflict-avoidance, and product as a window into peopleCapital, pricing, and the dangers of overfunding in a hyper-competitive marketWhere value accrues in AI: models, data exhaust, vertical agents, and founder continual reinvention

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome