Anthropic + Blackstone Launch $1.5B "Ode" — Betting AI Implementation Beats Model-Building
Anthropic and Blackstone have launched Ode, a $1.5 billion joint venture that embeds elite engineers directly inside enterprises to accelerate AI adoption. The thesis: the next trillion-dollar AI business isn't in building better models — it's in helping companies actually use them. Ode deploys "forward-deployed engineers" who work hands-on with enterprise teams to implement AI workflows. This follows a pattern we've seen emerging: the model layer is commoditizing, and the value is shifting to execution.
Read on TechCrunch →PwC's AI-Generated "Thought Leadership" Riddled with Fake Citations
PwC Middle East published AI-generated reports on topics like electric vehicles and government services that were found by GPTZero to contain fake footnotes, misattributed claims, and citations that don't exist. One footnote cited a teenage blogger with 280 followers as the source for a JPMorgan "success story" from 2017 — five years before ChatGPT existed. This follows similar retractions from EY and KPMG. "Do as I say, not as I do" is an interesting business model for firms advising clients on responsible AI adoption.
Read on Financial Times →GrowthLoop Ships Engagement Suite — AI Marketing Directly from Data Platforms
GrowthLoop launched Engagement Suite to execute multi-channel marketing actions directly from enterprise data warehouses. The application uses AI to analyze behavioral data, build target audiences, generate message content, and trigger campaigns — all without moving data to separate marketing tools. This continues the "composable CDP" trend: instead of copying data to yet another platform, activate it where it already lives.
Read on MarTech →CMSWire: Can Your Martech Stack Support AI Agents — Or Is It Just in the Way?
Microsoft's Sandip Patel warns that AI agents "must read and act across the full customer journey" and break when data and permissions are scattered across dozens of point tools. The conversation is shifting from "how many platforms do we own?" to "how well do those platforms work together?" Enterprises are consolidating around connected, governance-ready platforms built for AI readiness — not just cutting vendors, but fixing the workflows beneath them.
Read on CMSWire →MarTech: AI Pushes Composability Beyond Software — 15,505 Products and the "Hypertail"
The martech landscape now contains 15,505 commercial products. But AI is pushing composability in a new direction: the "hypertail" of custom-built solutions. Organizations are increasingly building their own low-code automations and AI agents on open platforms rather than buying more SaaS. Gartner predicts 40% of enterprise applications will incorporate task-specific AI agents by year's end, up from less than 5% at the start of 2026.
Read on MarTech →💡 My Take
Read this one: The Anthropic/Blackstone Ode story signals something important. When the company behind Claude — one of the leading AI labs — decides that the real money is in implementation consulting, not model development, it tells you where the market is headed. Models are becoming infrastructure. The differentiation now lives in who can actually deploy them inside enterprise workflows. The PwC embarrassment is the counterpoint: if you just bolt AI onto existing processes without understanding what you're doing, you end up citing teenage bloggers as evidence for JPMorgan initiatives.