Bessemer Venture Partners benchmarks 60% gross margin as the investor threshold for AI products. Most teams don't know if they're above or below it until someone asks in a board meeting. Built a free tool to fix that. Plug in your token flow and API cost. It outputs your cost-to-serve, your current gross margin, and how much you need to raise your price to clear 60%. Free trial. Starter. Pro. Enterprise. None of those numbers mean anything until you know your cost-to-serve per credit. Token Flow Calculator is live. kylekelly.co/tools/token-cal…

Hot take: For the overwhelming majority of agents I've built for small businesses. You do not need a virtual machine to run a 24/7 agent anymore. The labs absorbed the runtime. A webhook fires, a serverless function wakes, the agent drafts, it goes back to sleep at zero cost. The always-on Linux box was a 2025 mental model. Let it go.

This is kind of a game changer.

New in Claude Code: /ultraplan Claude builds an implementation plan for you on the web. You can read it and edit it, then run the plan on the web or back in your terminal. Available now in preview for all users with CC on the web enabled.

BREAKING 🚨: ANTHROPIC IS WORKING ON ITS OWN ALWAYS-ON AGENT SOLUTION CALLED CONWAY! CONWAY WILL HAVE A SEPARATE UI INSTANCE, WILL BE ABLE TO OPERATE BROWSER, CONNECTORS, CLAUDE CODE (EPITAXY?) AND COULD BE INVOKED VIA WEBHOOKS. IT WILL ALSO SUPPORT EXTENSIONS, AN UPCOMING CNW ZIP STANDARD FROM ANTHROPIC TO BUILD CUSTOM TOOLS, UI TABS, AND CONTEXT HANDLERS! LET CLAUDE COOK 🔥

Computer use is now in Claude Code. Claude can open your apps, click through your UI, and test what it built, right from the CLI. Now in research preview on Pro and Max plans.

You can now enable Claude to use your computer to complete tasks. It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk. Research preview in Claude Cowork and Claude Code, macOS only.

The companies winning with AI right now aren't the ones with the best models. They're the ones with the best operators. I've spent the last year watching organizations adopt AI. The pattern is clear: - Companies that hire an "AI team" and isolate them → slow progress - Companies that embed AI into existing workflows with skilled operators → compounding results The difference isn't technology. It's operational design. An AI tool without an operator who understands the business context is just expensive autocomplete. An operator who understands both the business AND the tool? That's leverage. The skill gap isn't "can you use ChatGPT." It's "can you redesign a workflow around what AI actually does well." That's the operator gap. And it's widening every month.

The Agency

This is next on my setup list.

Nice work @jamesclift

Been deep in the OpenClaw rabbit hole. Woke up this morning and realized I'M the bottleneck. The agent wants to work. I'm the one holding it back. Then I found this: a directory of MCPs that let your agent actually do things. Email, domains, deployments, payments. All of it. mcpservers.org/

Every company in the world today, needs to have an openclaw strategy, an agentic system strategy. - Jensen Huang

Anthropic just launched Anthropic Academy Totally free — 13 official courses, complete with certificates, and zero subscription required. Some highlights: → Claude 101 (perfect starting point) → Claude Code in Action → Building with the Claude API (seriously in-depth, 8 hours of content) → Intro to MCP Advanced MCP → Agent Skills → Claude on AWS Bedrock & Google Vertex AI anthropic.skilljar.com/

Been listening to 33 Laws of War by Robert Greene. Law 2 hit different: never fight based on the last war. There's something eerily similar happening in tech right now. Companies applying build, buy, partner frameworks built for a world where capability advantages lasted years. Frontier AI capabilities are now being replicated in months. The strategy has changed.

OpenClaw 2026.3.12 🦞 🎛️ dashboard v2 — slick new control UI ⚡ /fast mode for models 🔌 ollama/sglang/vllm → plugins (core goes on a diet) 🛡️ device tokens now ephemeral ⏰ cron windows reliability fixes touch grass? nah, touch main 🌿github.com/openclaw/openclaw…

The New Math of AI, by @GoKyleKelly open.substack.com/pub/kyleke…

The AI displacement narrative is missing the real story. The underuse of AI (in red) is far more interesting to me. New research from Anthropic shows that while AI could theoretically cover 60-94% of tasks in sectors like legal and finance, observed usage is just a fraction of that. This isn't a technology gap. It's a commercialization gap. We saw the same thing with the cloud. The tooling, pricing, and deployment architecture took a decade to catch up. AI is on the same curve, just steeper. The companies that close this gap first will build a significant competitive advantage through workflow depth that is hard to replicate. Models commoditize. Deployment does not.

Every AI company is becoming an energy company by accident. Training runs now cost more in electricity than engineering salaries. Inference at scale requires direct power purchase agreements. The CFO conversation shifted from "how much compute?" to "can we guarantee baseload capacity?" When your OpEx depends on megawatt-hours, you're in a different business than you thought.

you don't need to overthink this when you look at this goldman sachs chart long enough it becomes pretty obvious how people will build the next wave of $10m-$100m ARR vertical ai companies ill break it down so we all know every business function produces something tangible 1. a recruiting pipeline produces candidate summaries 2. a finance team produces monthly reporting packages 3. a real estate team produces market analyses and listing packages those outputs come from repeatable processes that pull information from a handful of systems and sources. builders who win in this environment start by understanding how those outputs get created today they collect real examples, reconstruct the process step by step, then design software that gathers the inputs and assembles the finished output automatically as adoption grows, the system expands into adjacent responsibilities until the product becomes the infrastructure that function runs on most people still think in terms of software categories. CRM. ATS. ERP. project management. that framing misses what is happening the next great vertical ai companies will be built around finished work. they will own the artifact the customer actually cares about, then expand outward until they own the function so the opportunity isnt really “build an ai tool for real estate” which is what i see a lot of on twitter the opportunity is much more specific: 1. build the ai employee that creates the broker opinion of value 2. build the ai employee that prepares the insurance renewal package 3. build the ai employee that drafts the first version of the investment memo 4. build the ai employee that assembles the lender reporting package every month that is how small software companies become very large ones in this market start with one painful output, automate it well, then expand until you own the workflow basically you go from automation to ai employees if you don't remember anything from this long post, remember that it's obvious that this is where its all going you dont need to overthink it you're in the robot business now