Cash flow is the net movement of cash into and out of the business over a defined period, categorized into operating, investing, and financing activities. Operating cash flow comes from running the business (customer payments minus operating expenses), investing cash flow tracks long-term asset purchases or sales, and financing cash flow captures debt and equity raised or repaid. Cash flow (not revenue or accounting profit) is the metric that actually determines whether a startup survives, because companies fail when they run out of cash regardless of what their P&L shows. It is the most operationally critical metric at most startups and the one founders most often misunderstand.
The three categories of cash flow:
Operating cash f...
Generative AI is the category of AI systems that create new content (text, images, code, audio, video, 3D) rather than classifying or analyzing existing data. The November 2022 release of ChatGPT marked the cultural and commercial inflection point that transformed generative AI from research curiosity to mainstream technology used by hundreds of millions of people within months. It's the category of AI that produces output rather than just labels or predictions.
The pre-ChatGPT history (compressed):
2014: Generative Adversarial Networks (GANs) introduced. First major generative image breakthrough.
2017: Google's "Attention is All You Need" paper introduces the Transformer architecture (the foundation for modern LLMs).
2018: Op...
An AI moat is the defensible advantage an AI startup builds to prevent commoditization by competitors. Five real moats exist in the AI era: data flywheel, workflow integration, distribution, brand and trust, and network effects. Raw access to foundation models is NOT a moat because everyone has the same APIs, making moat-building one of the most strategically important questions for any AI founder. It's the answer to "why can't anyone else build this?"
The five real AI moats:
1. Data flywheel ([Data Flywheel]):
An AI startup is a company whose product depends on artificial intelligence or machine learning as a core differentiator. The category breaks into three distinct archetypes: foundation model labs (OpenAI, Anthropic, Google DeepMind, Meta AI training the largest models), AI infrastructure (Hugging Face, LangChain, Pinecone, Weights & Biases providing tooling), and AI application companies (Cursor, Perplexity, Harvey, Glean building products on top of foundation models). Each archetype has fundamentally different economics, capital requirements, and defensibility characteristics. Understanding which category your AI startup falls into is the first step in evaluating its moat.
The three categories:
Foundation model labs:
A balance sheet is the financial statement showing a company's assets, liabilities, and stockholders' equity at a specific point in time. Unlike the P&L and cash flow statements that cover a period, the balance sheet is a snapshot, and the fundamental equation Assets = Liabilities + Equity always holds (hence "balance"). It is one of the three core financial statements (P&L, balance sheet, cash flow) that together provide a complete view of financial position. Balance sheets are more important at later-stage and public companies than at early-stage startups, where most items are minimal and cash is the only meaningful asset.
The standard balance sheet structure:
Assets (what the company owns):
Current Assets (convertible to ca...
A business plan is the written document describing a company's business model, target market, competitive position, operating strategy, team, and financial projections. It's used to align stakeholders and guide execution. Modern startup business plans rarely take the form of the traditional 30 to 40 page document; they more often appear as a pitch deck, a one-page Lean Canvas, or a short narrative memo.
The traditional business plan, with its executive summary, market analysis, organizational structure, marketing plan, operations plan, and 3 to 5 year financial projections, originated in mid-twentieth-century corporate planning and remains the format banks and SBA loan officers expect. For startups, the format has shifted. Mos...
The foundational vocabulary every founder needs before everything else. This cluster covers what a startup actually is, the categories that distinguish them (bootstrap vs venture-backed, lifestyle vs scale-up), the support ecosystem (accelerators, incubators, agencies), the early credits and grants founders chase, and the structural concepts (founder-market fit, why startups fail) that shape every decision that follows. 21 entries.
If you're new to startup vocabulary, start here. If you're a few years in, this cluster is the conceptual baseline against which everything else is read.
How startups end (and what determines who gets what). This cluster covers the major exit paths (IPO, acquisition, SPAC, direct listing), deal structures and terms (LOI, definitive agreement, earnout, holdback, reps and warranties), the rights that affect exit outcomes (drag-along, tag-along, ROFR, lockup), and the mechanics specific to exits (liquidation waterfall, exit multiples, QSBS). 26 entries.
Exits are the moment when years of equity decisions become real money. Founders should know this vocabulary years before they need it.
A foundation model is a large-scale AI model trained on broad, diverse data and designed to be adapted to many downstream tasks. Adaptation happens via fine-tuning, prompting, or API access. The term was coined by Stanford's Center for Research on Foundation Models in 2021 and now describes GPT-4, Claude, Gemini, Llama, Mistral, and similar models that form the base layer of the modern AI stack. The foundation model is to AI applications what AWS is to web applications: shared infrastructure that powers everything built on top.
What distinguishes foundation models:
Scale: hundreds of billions to trillions of parameters. Trained on hundreds of billions to trillions of tokens of data.
General-purpose training: trained on broa...
The vocabulary of the AI era of startups. This cluster covers the foundational concepts of modern AI (foundation models, LLMs, generative AI), the architecture and operations that power AI applications (Transformer, training data, fine-tuning, prompt engineering, RAG, context window), the economics that determine viability (inference cost, GPU cost, token economics), the strategic moats AI companies build (data flywheel, AI moat, wrapper vs thick wrapper), the safety considerations (alignment, safety), and current-era terms (multimodal, agents, vibe coding). 22 entries.
This cluster is the freshest in the lexicon. If you're building anything AI-adjacent in 2025, every entry here is operational vocabulary.