Software isn't dying. Its competitive edge is. In the past 24 hours, Cursor's CEO and Naval said the same thing in two different ways: writing code well is no longer a moat.

The race now is about who controls which agents, and who embeds them into the right workflow first.

What happened: Cursor CEO Michael Truell announced that over 35% of the company's internal pull requests are now created by AI agents, with no direct developer involvement. Agent usage on Cursor has grown 15x in the past year. On the same day, Naval posted six words on X that reached 2 million views: "Pure software is rapidly becoming un-investable."

Why it matters: Cursor isn't just describing its product. It's redefining the developer's role. Instead of writing code, the next-generation developer does three things: break down problems, spin up multiple agents in parallel, and review output. This is a shift from craftsman to factory manager. Naval names the financial consequence: if code can generate itself, the moat of a pure-software company collapses. What remains is proprietary data, ecosystem lock-in, and customer trust.

If the "second era" of agentic coding lasts less than a year as Truell predicts, disruption is accelerating exponentially, not linearly. Companies building software on 2024 logic may find themselves obsolete before their roadmap ships.

This isn't about AI replacing developers. It's about the definition of "building software" being rewritten entirely. The winner isn't the best coder. It's the person who asks agents the right questions, and reviews output fast enough to matter.

Markets don’t wait — and year-end waits even less.

In the final stretch, money rotates, funds window-dress, tax-loss selling meets bottom-fishing, and “Santa Rally” chatter turns into real tape. Most people notice after the move.

Elite Trade Club is your morning shortcut: a curated selection of the setups that still matter this year — the headlines that move stocks, catalysts on deck, and where smart money is positioning before New Year’s. One read. Five minutes. Actionable clarity.

If you want to start 2026 from a stronger spot, finish 2025 prepared. Join 200K+ traders who open our premarket briefing, place their plan, and let the open come to them.

Get Year-End Briefing

_{By joining, you’ll receive Elite Trade Club emails and select partner insights. See Privacy Policy.}

Trump ordered all federal agencies to stop using Anthropic products after the company refused to allow its AI to power mass domestic surveillance or fully autonomous weapons. Defense Secretary Pete Hegseth went further: banning all military contractors from doing business with Anthropic. CEO Dario Amodei held his ground.

→ Anthropic chose principles over contracts. There's no precedent for this in AI, and the entire industry is watching.

Hours after Anthropic was designated a national security supply-chain risk, Sam Altman announced OpenAI had reached an agreement to deploy its AI on the Pentagon's classified network. Altman claims the contract keeps the same two red lines as Anthropic: no mass surveillance, no autonomous weapons. The key difference: OpenAI gets to build its own safety stack and retains full control over it.

→ Same principles, different negotiation. OpenAI found a deal Anthropic wouldn't take. The real question: are these two contracts actually different, or just worded differently?

OpenAI announced 900 million WAU and 50 million paying subscribers, alongside a $110 billion funding round from Amazon, Nvidia, and SoftBank. January and February 2026 are on track to be the largest months for new subscribers in the company's history.

→ At this scale, ChatGPT is no longer an AI product. It's becoming infrastructure.

Chancellor Friedrich Merz visited China and was greeted by a live performance from Unitree humanoid robots: dancing, boxing, martial arts. Unitree's CEO called the visit an honor and expressed hopes for global collaboration. For Germany, a traditional manufacturing giant now playing catch-up in humanoids, this signals a strategic pivot toward China's digital economy.

→ Robots are now being used as diplomatic currency.

Today, I stumbled upon a fantastic guide on how to set up Claude Cowork. It's quite detailed, and I'm sharing with you how this guy explains it all in 15 minutes.

If you're on Claude Pro or above, you already have Cowork. But most users skip the most important step: context files. These are what turn Cowork from a generic agent into something that actually works like you.

Do this right after opening the app:

Then test with this prompt: "Read all files in context/. Tell me what you know about me and how I prefer to work."

If the output is accurate, your setup is working. Every time Cowork misses the mark, update the relevant context file. This is the compound interest of AI workflow.

For: founders, managers, analysts, PMs.

Naval used six words. Cursor used three pages. Same conclusion: code is a commodity. Judgment is the new moat. So, which parts of your work are you using AI for that only you know how to review?

👉 Forward it to someone trying to keep up with AI.

Read online: leftcurve.beehiiv.com