2026-05 - W1

AddyOsmani.com - Cognitive Surrender

Another link for the cognitive surrender, but now from a renowned personality.

Cognitive offloading is a superpower. Cognitive surrender is the failure mode of using it without noticing the line between them. The job, increasingly, is to stay calibrated about which side of that line you’re standing on at any given moment.

If your code is shipping and your understanding of the system is shrinking, you’re paying with cognitive debt. If your code is shipping and your understanding of the system is growing, you’re doing the actual job, just faster than before.

The tools are the same in both cases. The posture is what differs. That’s the part still entirely yours.

Harness engineering beyond skills: Using sensors to keep your coding agent in check - YouTube

Video of the author of the really good article: Harness engineering for coding agent users.

Must read/watch to have a better mental model of the word “harness” in the AI world.

Vibe coding and agentic engineering are getting closer than I’d like

The problem is that as the coding agents get more reliable, I’m not reviewing every line of code that they write anymore, even for my production level stuff. Claude Code does not have a professional reputation! It can’t take accountability for what it’s done. But it’s been proving itself anyway—time and time again it’s churning out straightforward things and doing them right in the style that I like.

The Normalization of Deviance in AI - Embrace The Red

Such dangerous, almost invisible consequence… I can believe that, because we are human and we can’t see too far in the future, and companies are working on quarterly timeframe, so short term benefit are louder.

Such a drift does not happen through a single reckless decision. It happens through a series of “temporary” shortcuts that quietly become the new baseline. Because systems continue to work, teams stop questioning the shortcuts, and the deviation becomes invisible and the new norm.

Especially under competitive pressure for automation, cost savings, a drive to be first, and the overall hype, this dangerous drift is evident. The incentives for speed and winning outweigh the incentives for foundational security. Over time, organizations forget why the guardrails existed in the first place.

How Do I Lead When I’m Not the Boss? | Andi Roberts

Really difficult topic. Involve lots of human interaction (obviously), and quite hard to evaluate the impact of each small actions of one on their community, besides “vibes”.

The post is quite meta, and difficult to put in practice, in a sense that it’s not possible to adopt all the advices in one go, and some requires specific context, timing, etc… to be able to “shine”. Especially on those critical moment, I don’t think I will remember those advice, and stress will most likely kick in instead.

Regardless, quite a good read. Might be worth to make some sort of plan to improve those “soft skills” (careful with procrastination with excessive planning).

Idempotency is easy until the second request is different

Really good article. Might be useful someday if I want to design harden idempotency.