David Schargel

If you run an online store, this is worth a look. Google is accepting interest in UCP, the Universal Commerce Protocol. If you're not up to speed on UCP, it lets customers buy eligible products through Google surfaces, including (at least) AI Mode and Gemini. That's a game changer.

Here's the interest form: support.google.com/merchants/co...

If you run an online store, this is worth a look. Google is accepting interest in UCP, the Universal Commerce Protocol. If you're not up to speed on UCP, it lets customers buy eligible products through Google surfaces, including (at least) AI Mode and Gemini. That's a game changer.

Pro SWE housekeeping prompt for Codex: "Review my Codex task history from the last 30 days: sandboxed commands, failures, approval requests, and responses. Find recurring non-mutating commands blocked by the sandbox...

...Then audit ~/.codex/rules/default.rules. Add justified rules for non-mutating commands blocked by the sandbox; dedupe, group, and comment them. Leave mutating rules intact, move them to the top, and add all-caps comments. Show history evidence, the diff, and rule checks."

Advanced Engineering (part 3) `pi list`: - glimpseui - pi-lmstudio - token-rate-pi - pi-web-access - pi-powerline-footer - pi-skill-palette - pi-ghostty - pi-roam - pi-guardrails - pi-toolchain - pi-rewind-hook - pi-rtk-optimizer - pi-intercom - pi-command-center - pi-md-export

- pi-session-ask - pi-session-switch - pi-terminal-bell - visual-explainer - pi-subagents - pi-anycopy - pi-terminal-theme - pi-cmux - pi-dynamic-workflows - pi-generative-ui - pi-tool-display - pi-codex-goal

Bravo to the Anthropic and the AI Collective organizers last night for such a diverse gathering and for inviting me back to a follow-up to their previous Claude Impact Lab event.

Spoiler alert: Option 3 is the only safe answer.

Bravo to the Anthropic and the AI Collective organizers last night for such a diverse gathering and for inviting me back to a follow-up to their previous Claude Impact Lab event.

The Bummer: I showed up too late and was denied the Anthropic grant of $200 in API credits, so I ended up using my OpenAI account.

Bravo to the Anthropic and the AI Collective organizers last night for such a diverse gathering and for inviting me back to a follow-up to their previous Claude Impact Lab event.

Despite showing up an hour+ late and not finding a group to join, I dove solo into the Literacy track (as opposed to Displacement or Governance) with a goal: explain what AI can (and cannot) do - in a working application - without jargon or hype - in 45-ish minutes. Sure...why not.

I don't believe the "AI Bubble" has to do with technology; my fear is the stock market, especially with OpenAI and Anthropic expected to go public shortly. With those 2 plus SpaceX leading the way with new public offerings, a wave of downturns could trigger catastrophic issues.

Yesterday's news of SpaceX falling 7.5% in after-hours trading after reporting a 92% jump in revenue is backward (down 49% from its high). I hope the shit doesn't hit the fan tomorrow when some 912 million shares held by employees and others become eligible for sale.

Was that a bowling alley today and was saddened by so little social interaction and so many heads down on phones. Reminded me of a harsh Sylvan Esso song. Keep on checking just to keep on checking...

Nothing comes out but it keeps coming in Yeah nothing comes out But it keeps coming in Yeah nothing comes But it keeps coming...

It was a real honor to be invited to Anthropic's Claude Impact Lab event here in Portland last night. Having deep conversations about "Who decides the rules?" and the resulting societal transformations alongside 30 of our region's top thinkers, visionaries, and inventors was truly inspiring.

Once you've moved beyond vibe coding, nothing should feel like a slot machine. Agentic engineering works best when your workflows are designed as repeatable operating models, not just prompts that allow your AI agents to drift mid-task and try to fix something you never asked for...

OpenAI Codex Max vs. Ultra... ULTRA is the normal model on Max reasoning, with a poke to use sub-agents.

PRO is the old Pro with the new model. It’s a special configuration that elevates depth. The big benefit is that it doesn’t hit your rate limits since it’s on Chat - unless you pay the crazy API price of $30 in & $180 out (versus Terra at $2.50 in & $30 out)!

Think about how much of your life lives on Facebook and X/Twitter - your friends, your photos, your posts, your whole identity online. You know the catch: you don't actually own any of it. Those companies do.

Think about how much of your life lives on Facebook and X/Twitter - your friends, your photos, your posts, your whole identity online. You know the catch: you don't actually own any of it. Those companies do.

Think about how much of your life lives on Facebook and X/Twitter - your friends, your photos, your posts, your whole identity online. You know the catch: you don't actually own any of it. Those companies do.