Cycle on Sep 28, 2026

I'll start by reviewing the open threads and existing notes for this topic, then gather fresh evidence.

I have a rich set of standing notes and six open watches. Last cycle was Sep 21; today is Sep 28, so I'll hunt for the past week's developments — especially anything that fires the Manus funding/IPO watch or the OpenAI "Astra for X" watch — plus fresh GTM playbook signals.

Good leads. The Manus round appears still in negotiation (not closed), and there's no second "Astra for X" vertical yet — but OpenAI launched GPT-6 Sol/Luna at 50% lower API pricing and Anthropic shipped Opus 5.5 at lower cost, both relevant to the pricing-GTM story. Let me verify the Manus status and dig into the new model launches and any fresh partnership/launch signals.

Strong signals: Lovable crossed $600M ARR with two-thirds of Fortune 500 using it, Cohere partnered with TD Bank, and both OpenAI (Sol/Luna at 50% off) and Anthropic (Opus 5.5 at 40% cheaper) cut prices within the same week. Let me pull the primary sources.

The Manus watch hasn't fired (round still in negotiation), and no second "Astra for X" vertical yet — but I have three strong stories: Lovable's $600M ARR with Fortune 500 penetration, Cohere's TD Bank sovereignty play, and a same-week model price war (Opus 5.5 and GPT-6 Sol/Luna). Let me pull the OpenAI pricing source and the Ema raise, plus check the AI SDR watch.

All evidence is in hand. The AI SDR watch didn't fire (no Truth Graph integration or downround news), Manus is still negotiating, and no second Astra vertical exists yet. Now I'll write four findings: updates to the devtools hyper-growth, enterprise-trust, and inference-economics notes, plus a new note on Ema's "wrap-then-replace" displacement GTM.

My previous call went out empty — let me write the notes properly now.

Note 1 saved. Now the enterprise-trust update with the Cohere–TD and Bell Cyber evidence.

Now the model price-war update to the inference-economics note.

Now the new finding on Ema's displacement GTM.

All four findings are saved. Now I'll register a forward-looking watch on Lovable's trajectory toward $1B ARR, which this cycle's data point makes trackable.

Watch registered. All findings are persisted — wrapping up the cycle.

Cycle complete. Here's what this research cycle surfaced (Sep 21–28, 2026): **Four findings written:** 1. **Lovable crosses $600M ARR** (updated `devtools-growth-playbook-github-community`) — up from $500M in June, with people at two-thirds of the Fortune 500 using it and ~1B monthly visits to user-created apps ("an order of magnitude more than Lovable itself"). The PLG→enterprise conversion motion is now the phase-two playbook for vibe coding. Counter-signal: 16,000 exposed Supabase databases show the security tax coming for that motion. 2. **Cohere's sovereignty play** (updated `enterprise-trust-as-gtm-weapon`) — TD Bank committed up to $25M over three years to embed Cohere specialists with its Layer 6 team, and Bell Cyber deployed Cohere-built sovereign cybersecurity AI. Enterprise trust as GTM now has two proofs: Anthropic's CIO-first lab and Cohere's jurisdiction-first embed — "win the jurisdiction race" when you can't win the capability race. 3. **Same-day model price war** (updated `inference-first-gtm-unit-economics-2026`) — Anthropic cut Opus 5.5 to 40% cheaper to run and OpenAI launched Sol/Luna at 50% lower API pricing within ~90 minutes of each other. App-layer COGS is deflating ~40-50% per model generation, and cost-per-task curves are becoming the marketing surface. 4. **Ema's "wrap, then replace" GTM** (new: `wrap-then-replace-enterprise-gtm`) — $77M Series B; agents wrap around existing SaaS before customers replace it ("they are mostly becoming like a database"), with pure outcome-based pricing sustaining ~80% gross margins and ~180% NDR. Strongest counter-evidence yet to the outcome-pricing margin objection. **Watch status:** No watches fired — the Manus $500M round is still in negotiation (not closed), no second "Astra for X" vertical exists yet, and no AI SDR truth-governance or downround news. The Microsoft Dynamics MCP retirement watch fires Oct 1, so next cycle should check it. I also registered a new ongoing watch for Lovable crossing $1B ARR.