| author | Alan Dipert
<alan@tailrecursion.com> 2025-12-29 16:43:21 UTC |
| committer | Alan Dipert
<alan@tailrecursion.com> 2025-12-29 16:43:21 UTC |
| parent | c889258beb4ed26fdfb87f825f3596d1b6c72766 |
| md/Coherence.md | +6 | -6 |
| md/Home.md | +1 | -1 |
| md/Writings.md | +1 | -1 |
diff --git a/md/Coherence.md b/md/Coherence.md index 887f14b..7074a51 100644 --- a/md/Coherence.md +++ b/md/Coherence.md @@ -1,6 +1,6 @@ # Coherence -Most engineers use Codex like autocomplete and stop at convenience. Leverage comes from **coherence**: keeping intent, artifacts, and execution aligned so the AI works longer than you do. +Most engineers use an AI coding agent like autocomplete and stop at convenience. Leverage comes from **coherence**: keeping intent, artifacts, and execution aligned so the AI works longer than you do. **Coherence means three alignments:** @@ -10,7 +10,7 @@ Most engineers use Codex like autocomplete and stop at convenience. Leverage com When a seam drifts, you land in a slow read–correct–retry loop. -Treat Codex like an employee, not a pair programmer. Employees need direction, expectations, memory, and feedback from the real world. I give those up front, then let it work asynchronously. A quick check: Codex should be working at least **2× longer** than I’m chatting with it. If I keep intervening, I’m doing its job. +Treat the AI agent like an employee, not a pair programmer. Employees need direction, expectations, memory, and feedback from the real world. I give those up front, then let it work asynchronously. A quick check: the agent should be working at least **2× longer** than I’m chatting with it. If I keep intervening, I’m doing its job. Fixing generated code is expensive; designing first is cheap. I’ll even ask one AI to draft the brief another AI will implement. That approach is cost control, not overengineering. @@ -20,10 +20,10 @@ The biggest gains come from **non-code artifacts** that lock intent in place bef - `AGENTS.md` and other operating guides - Failing tests that force alignment -Codex needs the same feedback loop I have. If I’m building a site, it should see the browser. If I’m building a CLI, it should run in a close match of the target environment. If I’m stuck manually validating everything, I’ve kept myself in the loop and throttled iteration. +The agent needs the same feedback loop I have. If I’m building a site, it should see the browser. If I’m building a CLI, it should run in a close match of the target environment. If I’m stuck manually validating everything, I’ve kept myself in the loop and throttled iteration. -Let Codex exercise the work and fill its own backlog: browse the site like a new reader, run the CLI in a staged environment, and capture TODOs for confusing flows or visual glitches without waiting on me to spot them. +Let the agent exercise the work and fill its own backlog: browse the site like a new reader, run the CLI in a staged environment, and capture TODOs for confusing flows or visual glitches without waiting on me to spot them. -Commits record outcomes, not dead ends. I start and end work by logging an issue in `git-bug` (or similar) so failed approaches and constraints live beside the code. Without that memory, Codex will “fix” problems by creating new ones forever. +Commits record outcomes, not dead ends. I start and end work by logging an issue in `git-bug` (or similar) so failed approaches and constraints live beside the code. Without that memory, the agent will “fix” problems by creating new ones forever. -The failure smell is simple: Codex keeps producing more work for me instead of less. When that happens, stop generating code, rebuild the artifacts, and restore alignment. Coherence keeps intent, artifacts, and execution in lockstep—humans set direction and taste; Codex supplies execution and endurance. +The failure smell is simple: the agent keeps producing more work for me instead of less. When that happens, stop generating code, rebuild the artifacts, and restore alignment. Coherence keeps intent, artifacts, and execution in lockstep—humans set direction and taste; the agent supplies execution and endurance. diff --git a/md/Home.md b/md/Home.md index 299bbcf..3768529 100644 --- a/md/Home.md +++ b/md/Home.md @@ -25,7 +25,7 @@ Updates | Date | Note | |:-----------|:-------------------------------------------------------------------------------------------------------------------------------------------| | 2025-12-27 | Added [AIAndRisk](./AIAndRisk.md), reflecting on why AI upside favors owners who can offload risk while salaried engineers must retain understanding. | -| 2025-12-27 | Added [Coherence](./Coherence.md) on treating Codex like an employee by keeping intent, artifacts, and execution aligned. | +| 2025-12-27 | Added [Coherence](./Coherence.md) on treating an AI coding agent like an employee by keeping intent, artifacts, and execution aligned. | | 2025-11-19 | Added [WhoIsGod](./WhoIsGod.md) as the hub linking to the names and covenants reference tables. | | 2025-11-19 | Added [CovenantsOfGod](./CovenantsOfGod.md), summarizing the major biblical covenants and their fulfillment in Christ. | | 2025-11-19 | Added [CovenantNamesOfGod](./CovenantNamesOfGod.md) to gather the covenant-compound YHWH titles in one reference. | diff --git a/md/Writings.md b/md/Writings.md index f7612fe..5579c57 100644 --- a/md/Writings.md +++ b/md/Writings.md @@ -3,7 +3,7 @@ I keep long-form pieces here when I want to explore a topic beyond a paragraph or status post. * [AIAndRisk](./AIAndRisk.md) weighs how AI shifts upside toward owners who can offload risk, while salaried engineers must keep understanding in their own heads. -* [Coherence](./Coherence.md) explains how I use Codex like an employee by aligning intent, artifacts, and execution. +* [Coherence](./Coherence.md) explains how I use an AI coding agent like an employee by aligning intent, artifacts, and execution. * [WhoIsGod](./WhoIsGod.md) is the entry point for the [Names](./NamesOfGod.md), [Covenant Names](./CovenantNamesOfGod.md), and [Covenant](./CovenantsOfGod.md) tables so readers can find every reference from a single hub. * [OurEmptyPromises](./OurEmptyPromises.md) reflects on the limits of human promises and why certainty only comes from God. * [RandallRDipertEulogy](./RandallRDipertEulogy.md) shares the eulogy I gave for my dad and the lessons I carry forward.