Notes on go-to-market, AI, and building things that last.https://jdav.is/https://jdav.is/white-papers/the-ai-rent-stack-july-2026.pdfhttps://jdav.is/white-papers/the-ai-rent-stack-july-2026.pdfThe economics underneath the drive for FDE make it a double-edged sword for enterprises: an opportunity and a liability. AI vendors have conceded that the model is no longer the moat and moved the contest to your last mile—here's what the deployment land grab is really about, and how to tell whether the offer on your desk leaves you more capable or more captive.Fri, 17 Jul 2026 00:00:00 GMTWhite paperhttps://jdav.is/articles/the-second-invoice-what-ai-generated-code-costs-after-the-meter-stops/https://jdav.is/articles/the-second-invoice-what-ai-generated-code-costs-after-the-meter-stops/The token bill is the cheap part. AI-generated code moves most of its real cost downstream—into review time, rework, failed builds, and the compute burned re-running pipelines. Here is how to find that cost before it finds your budget.Thu, 16 Jul 2026 00:00:00 GMTOperationshttps://jdav.is/articles/the-multi-model-mirage-governance-cost-and-accountability-when-every-developer-picks-their-own-ai/https://jdav.is/articles/the-multi-model-mirage-governance-cost-and-accountability-when-every-developer-picks-their-own-ai/Enterprise AI coding environments now support a dozen model choices per developer. That flexibility is real. So is the governance vacuum it creates—and the cost exposure that accrues quietly under uncapped usage-based billing.Fri, 10 Jul 2026 00:00:00 GMTGovernancehttps://jdav.is/articles/benchmark-theater-the-leaderboard-is-a-press-release/https://jdav.is/articles/benchmark-theater-the-leaderboard-is-a-press-release/Frontier AI models arrive roughly every two days, each announced with curated benchmark scores. For enterprise teams making procurement decisions, the noise has become the problem. A framework for what actually predicts model performance in production.Thu, 09 Jul 2026 00:00:00 GMTStrategyhttps://jdav.is/articles/the-atrophy-risk-what-happens-to-developers-when-ai-goes-dark/https://jdav.is/articles/the-atrophy-risk-what-happens-to-developers-when-ai-goes-dark/Controlled trials, outage data, and cognitive science converge on a warning most teams haven't heard yet: the developer who can't function without AI is a liability, not an asset.Mon, 06 Jul 2026 00:00:00 GMTStrategyhttps://jdav.is/articles/polaris-rising-what-the-copilot-model-swap-actually-changes-for-enterprise/https://jdav.is/articles/polaris-rising-what-the-copilot-model-swap-actually-changes-for-enterprise/Microsoft's Project Polaris replaces GPT-4 Turbo as GitHub Copilot's default engine in August. The shift touches architecture, IP liability, inference economics, and how much control enterprises actually have over the stack they thought they understood.Fri, 03 Jul 2026 00:00:00 GMTStrategyhttps://jdav.is/articles/the-token-tax-thesis-why-government-access-to-frontier-ai-may-be-about-more-than-security/https://jdav.is/articles/the-token-tax-thesis-why-government-access-to-frontier-ai-may-be-about-more-than-security/The Trump administration's restrictions on GPT-5.6 and Anthropic's models are being framed as cybersecurity measures. The fiscal math suggests something larger is developing in the background.Thu, 02 Jul 2026 00:00:00 GMTGovernancehttps://jdav.is/articles/scaling-ai-in-the-enterprise-what-the-evidence-says-works/https://jdav.is/articles/scaling-ai-in-the-enterprise-what-the-evidence-says-works/The data on enterprise AI adoption has matured. Organizations now have clear empirical signals—from Stanford, McKinsey, Deloitte, and IBM—on what separates durable, at-scale deployment from stalled pilots. Here is what that evidence looks like in practice.Wed, 01 Jul 2026 00:00:00 GMTAdoptionhttps://jdav.is/articles/the-developer-premium-what-eighteen-months-of-ai-replacement-attempts-actual-cost/https://jdav.is/articles/the-developer-premium-what-eighteen-months-of-ai-replacement-attempts-actual-cost/Enterprise leaders replaced developers with AI and called it transformation. The remediation bills, rehiring costs, and security incidents are calling it something else. A data-driven accounting of what AI-first workforce strategies actually cost—and a framework for doing it right.Mon, 29 Jun 2026 00:00:00 GMTAdoptionhttps://jdav.is/articles/the-prototype-trap-why-vibe-coding-platforms-have-limited-place-in-enterprise-systems/https://jdav.is/articles/the-prototype-trap-why-vibe-coding-platforms-have-limited-place-in-enterprise-systems/Generative app platforms like Replit, Base44, Lovable, and Bolt.new—and spec-driven IDEs like AWS Kiro—can compress weeks of scaffolding into hours. They should not be allowed anywhere near a production codebase. Here is why—and what to do instead.Thu, 25 Jun 2026 00:00:00 GMTSecurityhttps://jdav.is/articles/the-reassignment-dividend-why-displacing-workers-costs-more-than-redeploying-them/https://jdav.is/articles/the-reassignment-dividend-why-displacing-workers-costs-more-than-redeploying-them/Eighty percent of enterprises deploying AI have cut headcount. None of them got better returns for it. The data is in—and it argues for a fundamentally different strategy.Tue, 23 Jun 2026 00:00:00 GMTStrategyhttps://jdav.is/articles/the-ai-subsidy-economy-is-ending/https://jdav.is/articles/the-ai-subsidy-economy-is-ending/Frontier AI companies are burning tens of billions of dollars a year at negative operating margins while investors mark up paper valuations. Microsoft's shift to usage-based billing is not a pricing change. It is a reckoning with physics.Sun, 21 Jun 2026 00:00:00 GMTStrategyhttps://jdav.is/articles/the-token-is-the-unit/https://jdav.is/articles/the-token-is-the-unit/Token-based pricing has arrived across every major AI platform. Here is why it was inevitable, what it means for enterprise budgets, and how financial and technology leaders can build the governance frameworks to manage it.Mon, 15 Jun 2026 00:00:00 GMTGovernance