📬 This is the companion episode guide to 73% of New Enterprise AI Spend: Not OpenAI
Subscribe to get the full briefing →
Episode Guide: 73% of New Enterprise AI Spend: Not OpenAI
Companion to the Wednesday, April 1, 2026 edition of VC Brief: Startup & Early Stage Intelligence
This edition covers 11 episodes spanning Anthropic, AI Spending, Enterprise AI, OpenAI, Competitive Dynamics. Below you'll find detailed breakdowns of every episode referenced in today's briefing — including key guests, standout quotes, and links to listen.
Episode Guide
This Week in Startups — "This Bittensor Subnet Could Cut Drug Discovery Costs in HALF | E2267"
Runtime: 73 min | Host: Jason Calacanis | Guest: Mikaela Baso, Pedro Pena, Tom Bliers, Alex, Lon Harris, Max Sebti
Audience Framing: Founders and VCs in biotech or AI infrastructure should listen to understand how decentralized crypto networks are disrupting drug discovery and the creator economy.
This episode delves into how Bittensor, a decentralized network, leverages crypto incentives to drive AI contributions across various specialized subnets. Guests from Metanova Labs discuss their Subnet 68, which uses AI for virtual drug screening to drastically cut discovery costs and timelines. The conversation also explores Bitcast Network's Subnet 93, a novel approach to monetizing YouTube content creators by rewarding watch time over views, and touches on the concept of emergent behavior in AI agents within simulated environments.
"Bittensor is a decentralized network that uses crypto incentives to reward individuals who contribute useful AI models, compute or results to task specific subnets." — Jason Calacanis
Connects to: Decentralized drug discovery using Bittensor, Creator Economy Health, fine-tuning Bittensor models for specific tasks, VLMs vs. LLMs, AI bubble collapse timeline, AI models using only necessary resources, Miners' adversarial relationship as a feature, not a bug, efficient on-device inference for VLMs, Drug discovery cost and time reduction
Pivot — "Meta and YouTube Lose in Court, Insider Iran Trades, and Sora Shuts Down"
Runtime: 71 min | Host: Kara Swisher | Guest: Scott Galloway
Audience Framing: CEOs and board members should tune in to understand the latest legal liabilities for tech platforms and the significant market shifts in enterprise AI, directly impacting valuation and competitive strategy.
Kara and Scott dissect recent legal rulings holding Meta and YouTube liable for addictive features, setting a precedent that could reshape social media platform design and personal liability. They also discuss the implications of alleged White House insider trading on market dynamics and re-evaluate OpenAI's strategy following the shutdown of its Sora app, highlighting Anthropic's ascendance in the enterprise AI space.
"It's just not good policy not to have people who doubt each other and debate it." — Kara Swisher
Connects to: Social Media Addiction Lawsuits, OpenAI Sora app shutdown, Greatest levels of insider trading originating from Pennsylvania Avenue, OpenAI guaranteeing 17.5% return on private equity investment to top off round, Social media companies' insurance companies denying claims due to intentional harm, Democrats taking the Senate (Kalsha prediction markets), Lower priced models can expand market share for luxury brands (e.g., Porsche SUV, Mercedes A-class), Amazon's 'Transformer' AI-driven phone will likely be scrapped, New Disney CEO should create a Disney loyalty program with no lines at parks, SpaceX IPO to be priced low to allow for significant post-IPO gains
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — "20VC: Why You Need a $1BN Fund To Do Series A Today | OpenAI vs Anthropic: Who Wins Enterprise | SpaceX at $2TRN and Data Centers in Space | The $20BN Groq Deal Broken Down | Jeff Bezos' $100BN New Fund"
Runtime: 78 min | Host: Harry Stebbings | Guest: Jason Lemkin, Rory O'Driscoll
Audience Framing: GPs and LPs must listen to grasp the seismic shifts in AI enterprise spending, VC fund sizing, and the changing landscape of founder ambition, directly impacting allocation decisions and investment strategy.
This episode exposes the shifting power dynamics in enterprise AI, revealing how Anthropic is rapidly outpacing OpenAI in new spending. The hosts and guests dissect Jeff Bezos's rumored $100 billion AI-transformation fund and Elon Musk's ambitious SpaceX projects, questioning their $2 trillion valuation. The discussion also touches on the re-evaluation of Figma's product-market fit in the AI era and the broken traditional math for VC fund sizing in today's inflated Series A environment.
"Ramp data suggested that Anthropic now captures 73% of all spending among companies buying AI tools. But the marginal buyer in the last six weeks, eight, 10 weeks, has massively shifted. Moving away from OpenAI doesn't mean it's the end of the world." — Harry Stebbings
Connects to: AI Spending, Enterprise AI, Competitive Dynamics, $1BN Fund for Series A investments, Unicorn Dead Zone / Lack of Exit Opportunities, broken VC math for fund sizing, OpenAI's strategic inconsistency and perceived 'air of desperation' contrasts sharply with its previous 'invincible' image and is proving damaging to its enterprise adoption rates., Despite high token costs for some AI applications, many enterprise-level LLM implementations are not sensitive to these costs, making a switch to cheaper models unappealing once a workflow is established due to high 'soft costs' of re-integration and QA., The market's rational panic over Figma's revenue durability isn't just about Google's Stitch, but a broader concern about decaying product-market fit in the AI era, especially given Figma's perceived mediocrity in AI-driven design tools., Jeff Bezos's $100 billion AI-transformation fund reflects a trend among billionaires to pursue larger, later-stage, and less operationally intensive ventures, leveraging their capital for financial engineering rather than building from scratch. This strategy is driven by a desire for quicker value creation and a shift away from the 'hard way' of starting companies at a younger age., The math for VC fund sizing, which was once simple (e.g., $60M per partner for seed), is now broken due to inflated Series A round sizes and the need for larger check sizes., Shift in Enterprise AI Spending to Anthropic, 50% ARPU growth due to AI, SpaceX valuation at $2 Trillion, Jeff Bezos $100 Billion Manufacturing Transformation Fund, Elon Musk's fab for chips (Tesla & SpaceX joint venture), Seed investments eroded to 3-4% ownership at exit, Lowest ratio of potential acquirers divided by unicorns/decacorns, SpaceX valued at $2 trillion potentially due to data centers in space, SpaceX to build data centers in space
My First Million — "Oz Pearlman: How To "Read" Minds, Influence Anyone, and Never Fear Rejection"
Runtime: 54 min | Host: Sam Parr | Guest: Oz Pearlman
Audience Framing: Entrepreneurial CEOs seeking unconventional strategies for influence, sales, and overcoming personal fear should listen to this mentalist's tactical breakdown of human psychology.
Oz Pearlman, a renowned mentalist, reveals that his "mind-reading" skills are not supernatural, but rather finely-tuned observational and influence techniques applicable to business and everyday life. He shares personal anecdotes, including overcoming rejection through psychological distancing and a transformational past encounter with the law. Sam Parr and Oz delve into practical applications of behavioral economics and mental models, from reverse-engineering desired outcomes in real estate negotiations to applying an "alter ego" to manage performance pressure.
"If I could distill all of the skills I've learned in the last 30 years doing what I do, which is really interacting with people... I believe these are the core skills that would make me successful at any business. It's a reinvention of that book [How to Win Friends and Influence People]." — Oz Pearlman
Connects to: Overcoming rejection by cognitive dissociation (Oz Pearlman's method), The 'A, B, Z' approach (focus on current step), The 'Wow' emotion as a universal product, Reverse Engineering, Overcoming Fear of Rejection, Reverse-engineering the human mind, B2B vs B2C business model preference, Reinventing material for performances, You only know if something's right if you knew what was wrong, why does anyone care?, When everyone's selling, you should be buying, Oz Pearlman predicting Josh Allen's choices (throwing ball, ATM pin, jersey numbers), guessing ATM pin code via jersey numbers
Lenny's Podcast: Product | Career | Growth — "From skeptic to true believer: How OpenClaw changed my life | Claire Vo"
Runtime: 107 min | Host: Lenny Rachitsky | Guest: Claire Vo
Audience Framing: CTOs, product leaders, and founders exploring advanced AI agent deployments should listen to learn practical strategies for managing and scaling specialized AI agents for personal and professional productivity.
Claire Vo, founder of ChatPRD, shares her transformative journey from an OpenClaw skeptic to a fervent advocate, now managing nine specialized AI agents across multiple devices for everything from family scheduling to sales. She details her unique approach to configuring "soul" and "heartbeat" for agents, emphasizing the power of multiple specialized agents over a single general-purpose one. This episode offers practical advice on setting up, managing, and securely scaling AI agents, including insights on data privacy, browser limitations, and real-world economic value generation.
"I'm now running like eight different agents on OpenClaw. You really have to pull the thread on these tools and you have to spend enough time with them to see not where they are today, but where they are in a week and where they are in a month." — Claire Vo
Connects to: OpenClaw use cases changing Claire's life (family scheduling, sales, podcast prep), OpenClaw's 'Soul' and 'Identity' concept, Multiple specialized AI agents vs. one general-purpose agent, Brain transplant (OpenClaw agent knowledge transfer), OpenClaw installation and setup process, Avoiding OpenClaw mistakes (don't install on main computer), OpenClaw's 'Heartbeat' concept (cron jobs), OpenClaw browser unreliability, memory management in OpenClaw, Multiple specialized agents vs. one general-purpose agent, browser limitations, memory issues, and practical workarounds with AI agents, security risks and handling them in AI agents, onboarding experience for AI agents, OpenClaw agent scheduling (cron jobs/heartbeat), Yappers API, management skills to AI agents, Treat AI agents with respect and politeness, Start OpenClaw on a clean machine (e.g., old MacBook or Mac Mini), Create a separate admin account and email for OpenClaw agents, Use high-quality models (e.g., Opus 4.6, Sonnet 4.6, GPT 5.4) for OpenClaw agents, Use Telegram for initial OpenClaw setup due to its beginner-friendly nature, Use AI to employ an assistant to make life better, Use Google Workspace for AI agent collaboration, Watch Age of Attraction on Netflix (for 'garbage television'), OpenClaw moving past peak hype cycle to practical use cases, OpenClaw will change personal and professional life, AI agents will help individuals in personal logistics, Unexpected personal and professional impact of AI agents, AI will change personal and professional life unpredicted three years ago, Skepticism towards initial OpenClaw hype, OpenClaw as sandboxed
This Week in Startups — "The 5-Step Framework for AI Agents That Improve While You Sleep | E2269"
Runtime: 87 min | Host: Jason Calacanis | Guest: Shubham Saboo, Mike Nosov, Hakam Aujla, Lon Harris
Audience Framing: Product managers and engineering leads building agentic AI systems should listen for a structured approach on designing, deploying, and managing self-improving AI agent teams efficiently.
Google's Senior AI Product Manager, Shubham Saboo, shares a detailed 5-step framework for building efficient, self-improving AI agent teams using OpenClaw, emphasizing careful onboarding, scheduled autonomy, shared memory, and agent self-reviews. The episode also features demos of a "Minecraft"-inspired virtual workspace for agents called MoltWorld, and AgentMail for AI communication, alongside a discussion on Jason Calacanis’ viral tweet about founders avoiding traditional media.
"Start with one agent and make your way up by talking to your agent. That's the best thing that OpenClaw has done. So you really don't need to be a technical expert or you don't need to be a coding ninja to run these agents, deploy them, or make multiple agents out of it." — Shubham Saboo
Connects to: 5-step framework for designing an efficient AI agent team, Onboard one agent like a new hire, Put agents on fixed schedules (cron schedules), Add shared memory for agents, Agents run self-reviews and rewrite own instructions, 5-Step Framework for AI Agent Design, Shubham Saboo's 5 OpenClaw tips, Japanese immigration policies and debate, Trust in Media at All-Time Low, Horseshoe Theory in Media, Adversarial vs. Advocacy Journalism, Reliability in open source AI projects, Founders should avoid journalists, Twitter X bringing peace to regions of the world, mainstream publications are facing tough times
The a16z Show — "Marc Andreessen on Evaluating Founders and AI's Consumer Surplus"
Runtime: 68 min | Host: Harry Stebbings | Guest: Marc Andreessen
Audience Framing: Venture capitalists and established founders should listen to gain insights into Marc Andreessen’s contrarian views on founder evaluation, the perils of introspection in VC, and the economic implications of AI for industry centralization and labor.
Marc Andreessen offers a deep dive into his investment philosophy, highlighting the "scalded stove phenomenon" and the greater risk of "mistakes of omission" for VCs. He outlines his rigorous framework for identifying exceptional founders based on IQ, courage, and primal drive, advocating for backing individuals over business plans. Andreessen also challenges the prevailing narrative around AI's impact, arguing against labor displacement, emphasizing AI's consumer surplus, and noting the surprising re-centralization of the tech industry in Silicon Valley due to AI.
"In venture, I think you're always much more worried about the mistake of omission than you're worried about the mistake of commission." — Marc Andreessen
Connects to: The meme of 'retard maxing', Scalded Stove Phenomenon, Retard Maxing, AI reconcentrating tech industry in Silicon Valley, AI's consumer surplus, Mistake of Omission vs. Mistake of Commission, Extreme Ownership mindset, Inception point/early stage investing is critical for venture, Second Chances and Recovery in Career, Critical View of Managerialism, Backing the person over the business plan, American gestalt for risk-taking and big bets (reinforced by AI boom), Founder greatness evaluation: IQ, courage, drive, Mistake of omission vs. mistake of commission, flat organizations, Don't ever do diamonds in the rough, only do diamonds, Extreme Ownership (Jocko Willink), Startups are like baking a cake (sugar analogy for early decisions), Don Valentine's 'More companies die from indigestion than from starvation', High valuations set difficult hurdles for future funding rounds (down rounds), Conflict between investment mentalities in growth stage funding, Tech industry more centralized in Silicon Valley in the next decade due to AI, most large companies are overstaffed by 25-75%, Tech industry more centralized in Silicon Valley in the next decade, AI is not causing current layoffs, IPO as a branding event, not primarily financial or liquidity, Labor displacement theory around AI is totally wrong, Overfunding is dangerous to company operations, Always invest in 'diamonds' not 'diamonds in the rough' (avoid investor ego for hidden gems)., Overfunding is dangerous to company operations (Don Valentine's indigestion analogy), You don't need to 'like' founders you invest in; maintain professionalism., Do not bring your whole self to work., Naive view of vertical integration, Wealth inequality is not greater than it's ever been (historical context of monarchies/slavery), Nobody ever does a down round in someone else's company, Founders should not take unsolicited advice on overfunding despite its truth.
The a16z Show — "The SpaceX and Tesla Playbook for Hard Tech Startups"
Runtime: 51 min | Host: Erin Price-Wright | Guest: Chandler Luzsicza, Turner Caldwell
Audience Framing: Hard tech founders and executives leading complex R&D projects will benefit from these operational lessons on speed, flat organizational structures, vertical integration, and aggressive goal setting.
Chandler Luzsicza (Galadyne) and Turner Caldwell (Mariana Minerals) share invaluable lessons from their time at SpaceX and Tesla, emphasizing the "factory mindset" applied to complex hardware and manufacturing. They discuss the critical importance of speed, maintaining flat organizational structures for efficient information flow, and strategically managing critical paths to avoid burnout. The conversation also provides insights into selective vertical integration and the rigorous hiring practices that define these industry-leading companies.
"The gap between mythology and method is where the useful knowledge lives." — Erin Price-Wright
Connects to: Vertical Integration Strategy for Hard Tech Startups, mission alignment to avoid burnout, decision velocity, Takt Time Analysis, End-to-end project experience for engineers, Vertical integration for company existence (binary outcome), Engineers should experience end-to-end project cycles multiple times before founding a company, Strategic vertical integration for bottleneck removal, Over-indexing on technical basis before company building for founders, burnout caused by churn, not long hours, Naive view of vertical integration
Y Combinator Startup Podcast — "How François Chollet Is Building A New Path To AGI"
Runtime: 57 min | Host: Y Combinator | Guest: François Chollet
Audience Framing: AI researchers, entrepreneurs, and deep tech investors should listen for a contrarian perspective on AGI, deep learning's limitations, and the emerging paradigm of symbolic descent from the creator of the ARC Prize.
François Chollet, founder of the ARC Prize and NDIA lab, introduces symbolic descent as a groundbreaking machine learning paradigm poised to eventually replace deep learning. He argues that current LLM-based approaches, while useful for verifiable tasks, are inherently inefficient for achieving true AGI. Chollet redefines AGI as human-level skill acquisition efficiency and details the ARC-AGI benchmarks (especially V3) which measure intelligence in novel, interactive environments, projecting AGI's arrival by the early 2030s with surprisingly minimal code.
"I personally don't think that machine learning or AI in 50 years is still going to be built on this stack. I think it's inevitable that the world of AI will trend over time towards optimality." — François Chollet
Connects to: ARC AGI Benchmark, Deep Learning, Symbolic compression in science vs. curve fitting, Intelligence as skill acquisition efficiency, AGI by 2030 Timeline, NDEA, ARC V3, ARC AGI V3: Measuring “Agentic Intelligence”, Reasoning Models, LLM Stack, ARC V2, ARC V1, Program Synthesis, Removing humans from the AI improvement loop, Verifiable Reward Signals in AI, Importance of scaling approaches without human bottlenecks, Symbolic Descent, Program synthesis for AGI, Deep learning guided program search, Human-level skill acquisition efficiency, Neural networks were considered a 'failed approach' in the late 2000s, LLMs not optimal for long-term AI, AGI in less than 10,000 lines of code, AGI by 2030s, around ARC 6 or ARC 7, AGI by 2030 or early 2030s, AGI will be less than 10,000 lines of code and fit on 1980s compute, Explore alternative AI approaches not reliant on deep learning, Hire enthusiastic power users for open source projects, Study older AI research from 70s/80s, Read older AI research papers (70s-80s), Explore alternative AI approaches
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — "20VC: The Venture Model is Broken | You Need to be Greedy and Selfish to Win Early Stage Investing | Why Margins Do Not Matter for Early-Stage Startups | The Growth Rate that is Required in a World of AI with Gili Raanan, Founder @ Cyberstarts"
Runtime: 53 min | Host: Harry Stebbings | Guest: Gili Raanan
Audience Framing: Venture GPs, LPs, and growth-stage founders should listen to grasp how Cyberstarts achieves outsized returns by mastering the new math of early-stage investing, navigating inflated valuations, and strategically managing talent liquidity.
Gili Raanan of Cyberstarts makes a stark case: the traditional venture model is "broken" due to market imbalances and soaring entry prices, yet Cyberstarts has achieved an exceptional hit rate with its targeted strategy in cybersecurity. He challenges the relevance of margins for early-stage companies, prioritizing aggressive growth and deep product-market fit in an AI-driven world. Raanan also details their innovative "employee liquidity fund" as a critical tool for talent retention, defying traditional VC illiquidity.
"The market is not balanced. A lot of that cash that's flowing into the market would be wasted." — Gili Raanan
Connects to: Venture business viability with high entry prices, Margins in the age of AI, Trajectory, velocity, and growth rates as key business indicators, Extension of private markets and secondaries, Employee Liquidity Fund at Cyberstarts, Importance of founder chemistry, Teams shipping faster than ever with Cursor, Claw, Code, Codex AI, High new ARR growth rates (e.g., 4x annually for first 5 years), Companies with fast growth DNA tend to continue fast growth, Increasing seed entry prices in cybersecurity, Public market multipliers for tech companies, Low likelihood of cybersecurity unicorns (1-2 per year outside 2021), IPO as a branding event, not primarily financial or liquidity, Selling Wiz shares early was a mistake despite positive LP reaction, Greedy and selfish traits are good for early-stage investors, Founders must pick financing partners wisely, Gili Raanan is not being a 'boomer' regarding high entry prices due to market facts, Not being in every important AI or cybersecurity company, Many venture players will face 'catastrophe', Mega funds may not return 'venture like economics' due to entry prices, Not worried about too much money defocusing founders, GP/LP misalignment regarding early liquidity vs. long-term holding
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — "20VC: Marc Andreessen on The Future of Venture Capital: Will a16z Go Public | Why Labour Displacement with AI is Wrong | Why Introspection is Dangerous | Why \"Diamonds in the Rough\" is BS in VC | Why a16z Invested $300M into Adam Neumann"
Runtime: 72 min | Host: Harry Stebbings | Guest: Marc Andreessen
Audience Framing: Venture capitalists and established founders need to hear Marc Andreessen's unfiltered insights on VC strategy, founder psychology, and how AI is fundamentally reshaping the tech ecosystem and conventional wisdom.
Marc Andreessen doubles down on his contrarian views, arguing against introspection in VC as a "scalded stove phenomenon" that leads to missing future opportunities. He elaborates on his founder evaluation framework (IQ, courage, drive) and explains why blindly chasing "diamonds in the rough" is a fallacy. Andreessen also challenges the AI labor displacement narrative, asserts Silicon Valley's renewed centrality due to AI, and defends controversial investments like Adam Neumann's Flow, emphasizing the critical importance of early-stage, founder-led initiatives.
"In venture, I think you're always much more worried about the mistake of omission than you're worried about the mistake of commission." — Marc Andreessen
More from VC Brief: Startup & Early Stage Intelligence
- Episode Guide: Anduril’s $20B Army Contract vs. Nvidia’s Muted Growth
- Episode Guide: Anthropic vs. Pentagon: AI Nationalization Risk
- Episode Guide: The $200M Moral Stand: Anthropic, the Pentagon, and 1.5-Person GTM Teams
- Episode Guide: White-Collar Reckoning & $20B Wipeout
- Episode Guide: AI Agents: Shipping Code, Not Ready for Mainstream
Get the next edition delivered to your inbox
