We just open sourced an AI that places real phone calls on your behalf. We named it after the dog.

Quick roadmap:

• Meet the actual Banjo, and why the name isn’t just a cute mascot choice
• What the software does, in one real example
• Five ways the dog’s actual personality shows up in the architecture
• Two bugs that made it to a live phone call before we caught them
• What’s still held together with tape, said plainly
• How to run it yourself

Meet Banjo

Banjo is a rescue terrier mix. Nobody in the family agreed on his exact breed mix and everyone agreed on the verdict: the little terrier with the big personality. He has ears that stand up like he’s perpetually mid-question and a face that looks skeptical of your plans by default. There’s a photo of him in a pirate costume on the beach in Asbury Park, standing his ground against wind that’s actively trying to remove his hat. He did not ask for the costume. He tolerated it, the way he tolerates most indignities, with visible reservations and zero actual protest.

In 1991, a federal judge ruled that Biz Markie’s use of a Gilbert O’Sullivan sample was theft, full stop — and for the better part of a decade, hip-hop producers had to clear a sample before they could trust it wouldn’t end their career. The technique wasn’t illegal. But the presumption around it changed overnight, and the presumption did more damage than the ruling.

Cost Optimization Just Got a Second Author

A model rewrote its own GPU serving code, cut OpenAI’s inference cost, and got itself a price cut in the field — while nobody prompted it to think about money.

That’s not a caching trick. That’s not a router sending easy questions to a cheaper tier. That’s the thing traditionally done to a model — the entire FinOps-for-AI playbook — being done by one, autonomously, at the infrastructure layer. And it landed in the same week I was already forwarding myself an open-source tool that does the mirror-image job on AWS: scan the account, find the waste, hand you a ranked list with a dollar figure attached, no ticket, no Slack thread, no waiting on next quarter’s audit. Cost optimization didn’t get better tooling. It got a second author, on both sides of the meter, in the same seven days.

KPMG didn’t buy AI for 276,000 people. It bought a control plane — and threw in a quarter-million seats to keep you looking the other way.

Read the June 9 announcement again. Everyone fixated on the headcount: more than 276,000 professionals getting Copilot. That’s the press release. The load-bearing noun is one clause down — Agent 365, the system KPMG adopts to “manage how AI agents are deployed, managed, monitored and updated,” giving it “centralized governance and control of AI agents operating across systems, data and business processes.”

For thirty years, the first question before building anything was “will this be worth it?” That question is dead. I killed a small piece of it on a Saturday morning, on my deck, with a cup of coffee and one sentence typed into my phone.

I wanted a clean read on the World Cup schedule dashboard. Game times, teams, location, which streaming service had each match (with live updates as a bonus and only client-side - no backend). That information existed everywhere and nowhere — scattered across a dozen pages, none of them showing me the one thing I actually wanted in one glance. So I opened the Claude app on my phone, typed a simple short prompt describing what I wanted, and hit go. About ten minutes later I had https://www.hatch.org/world-cup/ — a clean, shareable schedule that did exactly one thing. I sent it to family and friends. It will be useful for about a month. Then the tournament will end and so will the app’s reason to exist (for maybe another 4 years).

There’s a new letter, and you’re not it.

For a century, getting rich topped out at a word you could say with a straight face: millionaire. An M. Then we let the B’s happen — billionaires — and quietly agreed it was the same word with a worse vowel. It is not the same word. And as of today there’s a third tier nobody is emotionally prepared for: the T. A trillionaire is not a spicy B. The jump from a billion to a trillion isn’t a raise, it’s a change of species — Elon didn’t climb the ladder, he left the food chain. You can’t even be jealous of him correctly. Envy doesn’t reach that far; it’s a goldfish resenting the moon.

The best deal in software just ended, and the replacement bill is non-deterministic.

For two years, flat-rate coding-agent pricing was the deal of the decade. Twenty bucks a month, point an agent at your codebase, let it churn. You were almost certainly consuming more value than you paid for. That wasn’t generosity. It was customer acquisition, financed by venture money, priced below cost on purpose. And in a two-week window this spring, three different vendors quietly clawed it back.

The $965 billion number is a distraction. Here’s what actually matters.

Last week, Anthropic raised $65 billion and saw its valuation approach a trillion dollars. Every business publication on the planet covered the number. Almost none of them covered the insight.

The insight came from Simon Willison, posting on May 27 — one day before the funding closed — and it’s worth more to your business than any valuation figure: enterprise AI has found product-market fit. Not “gaining traction.” Not “showing promise.” Fit. The kind where customers stop subscribing and start paying by the token because they’re consuming so much that the subscription ceiling became a constraint on their business.

The enterprise AI agent story of 2026 is a demo story. Somewhere between “watch this” and “we run this at 2am,” the wheels come off — and almost everyone is blaming the wrong part of the car.

Quick roadmap:

• The headline stat and why the usual diagnosis is wrong
• What actually kills agents in production (it’s not the model)
• What the teams that crossed the gap did differently
• The uncomfortable truth: you can’t buy your way out of this

The Number Nobody Wants to Explain

A March 2026 survey put enterprise agent pilots at roughly 78%. Production deployments below 15%. That is not a rounding error. That is a structural failure, happening industry-wide — across models, across vendors, across use cases.

Subtitle: On the dark psychology of token scarcity, the $200 plan that still isn’t enough, and why Anthropic’s usage bar is the most stressful thing on the internet

It’s 11:47 a.m. on a Saturday and I am staring at a message in my beloved terminal that has become the most familiar sentence in my life: