CHAPTERS
- 0:00 – 2:22
Goldman’s funding lesson: scaling deposits and escaping wholesale funding risk
Solomon opens with a candid look at a key strategic vulnerability for large financial institutions: unstable wholesale funding. He explains how Goldman built a meaningful deposit base from essentially zero, improving liquidity resilience and reducing reliance on short-term markets.
- •Wholesale funding can be a dangerous “largest in the world” position during stress
- •Deposits are a more stable funding source than institutional/wholesale markets
- •Goldman went from 0 deposits to ~ $500B total, including a large digital platform
- •Funding and liquidity are framed as long-term strategic risks, not just treasury issues
- 2:22 – 5:30
From partnership to public company: keeping the culture while adding top-down strategy
Solomon describes Goldman’s transition from private partnership to public company and why permanent capital was necessary to stay globally relevant. He highlights the tension between preserving a partnership mindset and operating with the strategic direction required of a scaled public enterprise.
- •Private partnerships create strong mutual agency and entrepreneurial behavior
- •Going public in 1999 enabled permanent capital and global capital markets expansion
- •Goldman aims to retain a partnership culture (partner aspiration, correlated comp)
- •A public company needs clear strategic direction to make “1+1+1+1 > 4”
- •Core values emphasized: client service, partnership, integrity, excellence
- 5:30 – 9:28
CEO priorities: scale and funding as existential strategic moats
Looking forward, Solomon argues that scale is increasingly decisive in mature financial services, especially during turbulence. He also reiterates funding stability as a second major strategic pillar, explaining why Goldman must continue building a stronger deposit franchise to compete with larger universal banks.
- •US capital markets create globally dominant financial institutions
- •Goldman and Morgan Stanley are the smallest among the top US financial giants
- •Scale provides leverage and latitude during volatility and crises
- •Balance sheet scale comparisons: Goldman ~$1.9T vs JPM ~$4.5T
- •Deposits reduce dependence on less stable wholesale funding
- 9:28 – 10:27
Founding a16z in the downturn: raising capital when nobody has money
Horowitz reflects on launching a16z in 2009 amid skepticism and scarce capital. He frames contrarian fundraising as a durable investing advantage and sets up how the firm’s strategy evolved alongside major platform shifts like mobile and cloud.
- •2009 was an unusually good vintage to start investing (mobile + cloud inflection)
- •Contrarian fundraising: the best time to raise is when others can’t
- •Investors often chase highs and retreat at lows—structural behavioral error
- •Early luck plus disciplined positioning helped a16z establish staying power
- 10:27 – 11:58
Becoming “top tier” by building a better VC product for founders
Horowitz lays out the core venture logic: without top-tier status, firms struggle to access the best entrepreneurs and deals. a16z’s first phase focused on differentiating through founder enablement—treating venture as a product built for entrepreneurs, not just LP returns.
- •Top-tier status is necessary to reliably win the best deals
- •Historically, top-tier was reputational and hard to catch up to post-2009
- •a16z designed services to strengthen founder control and CEO success
- •Shift away from the earlier norm of replacing founders
- •Building brand/power/access as a repeatable founder-support system
- 11:58 – 13:58
Scaling venture for “Software is Eating the World”: from 15 winners to 150
The second phase of a16z was premised on software expanding the number of breakout companies dramatically. Horowitz explains the organizational challenge: scaling firm capacity without degrading investment quality, and why this pushed a16z beyond traditional Silicon Valley venture models.
- •Traditional VC assumed ~15 $100M-revenue tech breakouts per year
- •If software eats the world, the opportunity set expands massively
- •A small “basketball team” partnership model doesn’t cover a 150-company world
- •Scaling requires organizational design while keeping decision-making tight
- •This strategy helped drive a16z’s growth to a leading share of US VC fundraising
- 13:58 – 15:38
From firm success to national competitiveness: policy, dynamism, and China
Horowitz connects a16z’s market leadership to a broader mandate: growing the overall market and strengthening US technological leadership. He frames policy work as necessary to ensure the country remains competitive over the next century, particularly against China.
- •Andy Grove lesson: industry leaders must grow the market itself
- •a16z sees responsibility beyond firm outcomes—US technological “winning”
- •Policy engagement expands to crypto, AI, international posture, and “American dynamism”
- •Competitive framing: long-term implications of losing tech races to China
- 15:38 – 19:30
“The sweetest macro spot” in decades: stimulus cocktail, AI productivity, and fragile geopolitics
Solomon argues the US is in an unusually supportive macro regime for financial and investable assets, driven by overlapping fiscal stimulus, monetary easing, capex super-cycle, and deregulation. He balances that optimism with warnings about multipolar geopolitics and information-driven volatility.
- •Rare alignment: fiscal stimulus + rate-cut cycle + capex super-cycle + deregulation
- •Mega-cap spending: top 4 companies’ ~$400B capex added ~1% to GDP growth
- •Markets pull forward expectations of AI-driven productivity gains
- •Policy responsiveness: leadership that ‘marks to market’ can shorten drawdowns
- •Key risk: higher geopolitical fragility in a multipolar world + social media volatility
- 19:30 – 21:50
M&A and IPO outlook: from “no” to “maybe,” with regulatory caveats
The conversation turns to dealmaking, where Solomon predicts a potentially historic year for M&A and a stronger IPO pipeline as confidence returns. Horowitz agrees on activity but flags ongoing uncertainty around FTC posture—especially in tech—potentially pushing activity toward IP-centric transactions.
- •Confidence drives IPOs and M&A; tougher regulation suppresses confidence
- •Solomon: prior four years felt like every strategic question was “no”
- •Now: “whatever the question is, the answer is maybe” → activity unlocks
- •Prediction: could be the biggest M&A year on record + improved IPO market
- •Horowitz: FTC uncertainty may shift deals toward IP transactions vs classic M&A
- 21:50 – 23:38
AI competition dynamics: why “leads aren’t what they once were” and capital matters
Horowitz argues AI changes competitive moats: with enough proprietary data and compute, incumbents can ‘buy’ their way into capabilities faster than in classic software. This compresses lead time advantages and increases the need for rapid scaling capital, potentially accelerating IPO timing.
- •Classic software benefited from the ‘mythical man-month’ constraint
- •AI reduces the durability of early product leads when money+compute can replicate
- •Proprietary data + GPUs can solve many problems quickly—“it is magic”
- •Hypergrowth examples: companies hitting $100M+ or even $1B run-rates extremely fast
- •Public markets may become a funding tool to sustain AI competition intensity
- 23:38 – 25:40
Crypto policy agenda: stablecoins (GENIUS Act) and market structure (Clarity Act)
Horowitz explains why a16z prioritized crypto policy, framing crypto as a foundational technology for digital property rights and new economic coordination. He describes the GENIUS Act’s passage and positions the Clarity Act as the crucial next step to define token market structure and regulatory categories.
- •Crypto framed as societal infrastructure: property rights, creator economies, coordination
- •Claim of prior ‘ban by enforcement’ tactics (e.g., debanking, Wells notices)
- •GENIUS Act / stablecoin legislation: passed and enacted
- •Clarity Act (market structure): aims to define how different tokens are classified
- •Critique of “everything is a security” approach extending even to NFTs/artists
- 25:40 – 28:50
AI policy agenda: don’t regulate math, prevent 50-state compliance chaos, and protect training access
Horowitz outlines an AI policy stance focused on preserving innovation velocity while regulating harmful applications. He warns that fragmented state-by-state rules could cripple startups, and argues that restrictive training/copyright approaches could disadvantage the US versus China.
- •Core principle: models are mathematical predictors—regulate applications, not the model
- •Concern about fear-driven narratives and potential regulatory capture
- •Priority: avoid 50 different state AI regimes that block innovation
- •Training data/copyright: allow statistical learning without reproducing works
- •Geopolitical framing: China’s weaker IP constraints could yield stronger models faster
- 28:50 – 33:29
One GS 3.0: using AI to reimagine core processes, not just boost individual productivity
Solomon distinguishes between AI as a tool that makes top talent more productive and AI as a lever to redesign enterprise workflows end-to-end. He describes One GS 3.0, a program targeting six major processes for automation and redesign to free up capacity for growth investments while maintaining return discipline.
- •Track 1: put AI tools in employees’ hands to increase day-to-day productivity
- •Regulatory clearance is a major gating factor for financial institutions’ AI adoption
- •Track 2: reimagine operating processes for major efficiency and automation gains
- •Capital discipline: efficiency can fund higher tech spend without lowering returns
- •One GS 3.0 targets six initial processes; change management is the hard part
- 33:29 – 34:29
Will AI agents change investing? Limits of models and the rarity of true outperformers
They debate agentic investing, noting that models are bounded by available historical facts and struggle with unprecedented regime shifts. Solomon questions whether AI built on broadly accessible information can replicate the sustained edge of the small set of legendary investors.
- •Investing is often driven by novel, unforeseen events not in historical data
- •Models can adapt quickly after new facts emerge, but don’t anticipate unknowns
- •Most investors underperform; a small minority persistently outperform
- •Open question: can AI break the ‘average information’ trap to create real alpha?
- 34:29 – 35:34
Closing lightning round: favorite DJs
The session ends with a personal bonus question that contrasts Solomon’s current electronic pick with Horowitz’s classic hip-hop choice. It provides a brief, human closing beat after heavy macro, policy, and AI discussion.
- •Solomon’s pick: John Summit for contemporary club/house energy
- •Horowitz’s pick: DJ Jazzy Jeff as an underrated all-time great
- •A light, culture-focused wrap-up to end the conversation
