Harry Stebbings

Harry Stebbings × Naveen Rao

Rao sees US spare power becoming scarce within years, tying AI growth to construction.Unconventional co-designs models and hardware for efficiency as falling token prices pressure inference margins, despite a $10 trillion market forecast within 10 years.He sees robotics 100 times larger than agentic coding, with prototypes in two to three years and a market in five to seven.

Harry Stebbings × Harrison Chase

Chase says vertical-agent routing belongs in a state-aware harness, where domain-specific context and verifiers differentiate products.He estimates most OpenRouter usage is gateway access; smart routing is less common than assumed, and better frontier models can erode its value.His inference-market reversal—passing on a gateway was “100% the wrong decision”—signals room for Fireworks, AWS, and Crusoe.

Harry Stebbings × Mike Cannon-Brookes

Atlassian is betting that agents will expand interface value, especially for collaborative work requiring taste, context and human interaction.The Teamwork Graph turns each customer’s people, decisions, documents, code and workflows into queryable context for agents.Atlassian currently uses no Chinese models in production while routing Nemotron, Mistral and other models by quality, cost and speed; specialized models must sustain “exponential gains” as general models improve.

Harry Stebbings × Alex Mashrabov

Higgsfield says it reached $1B in annualized revenue 18 months after $1M, using trailing-four-weeks revenue ×13 and counting only live revenue.The economics are unusually expansionary: month-12 business NRR exceeds 300%, while routing own/open-weight models produces over 80% margins versus 20–30% for closed-source models; monitor whether at least 30% monthly growth survives the projected substantial deceleration.

Harry Stebbings

Meta's Muse, called the first real ChatGPT competitor, pairs autonomous agents with a free consumer LLM and added roughly $100B to Meta's market cap.Jev targets 20% of LLM API spend at one-fifth the price, potentially turning $20B into $4B of revenue.OpenAI's projected $278B net burn and roughly $700B of CapEx keep capital intensity central to the AI trade.

Harry Stebbings × Daniel Dines

Daniel Dines argues enterprise AI will diffuse process by process, because models retain context without being transformed by experience, while “millions of Einsteins” are not yet hireable people.The investable asset is the portable “map of work,” with AI designing deterministic, governed software around probabilistic models; UiPath’s 14% growth and production gaps in vibe-coded procurement tools leave timing and near-free human replacements as key risks.

Harry Stebbings × Thomas Sohmers

Positron co-founder Thomas Sohmers frames inference as a memory-wall problem: GPU FLOPS rose 120x from 2014–2024, versus 17x for memory bandwidth.Positron targets cutting a 500-megawatt Nvidia deployment to 100 megawatts, while cached tokens cost roughly one-thousandth of recomputation and support reported 80-point Anthropic API gross margins.Planned capacity may shift locations rather than disappear, but debt, regulation, and concentration remain risks to monitor.

Harry Stebbings × Matteo Franceschetti

AI is letting Eight Sleep scale output: engineers direct hundreds of AI engineers, while email marketing generates close to $100M with a team of zero and paid media produces $200–300M with two people.Watch whether this productivity supports the three-year target of 250 people making a billion, amid talent retention, Anthropic’s cost curve, regulation, and model self-learning risks.

Harry Stebbings

Jensen Huang’s AGI claim meets Rory O’Driscoll’s skepticism, while Fable 5.1 delivers Jason Lemkin’s first step-function coding experience since last year’s “three dot” models.Instinct’s quickly cloned WhatsApp distribution, $2.5B pre-revenue valuation, Wonderful’s $5B repricing, and Anthropic’s reported roughly $6B deal reversal expose a market where rule-breaking and capital access remain central.

Harry Stebbings × Jean-Denis "JD" Grèze

Town converts more than 15% of trial users to paid and generates over $700 per user annually despite losing 30% at its mandatory email-and-calendar gate, while teams spend at least $75K per engineer on AI tools.Its agent-to-agent network effect could make team switching difficult, but zero pricing power at the frontier leaves economics dependent on whether inference costs fall and how much workload remains frontier.