Allen Pike

Building ⚡️ I also host and write monthly about what I’m learning. Follow me elsewhere:

[Allen Pike @apike]

My articles automatically post here, but if you use Bluesky, LinkedIn, Email, or Mastodon then better to follow along that way: https://t.co/W1PJ6Ms4bb

[Allen Pike @apike]

@simonw Yes! This is a great way to implement the “it should be clear what is AI-only writing, and what is human-vetted text” rule ⭐️

[Allen Pike @apike]

Like posting on social media, obviously

[Allen Pike @apike]

The core challenge with coding agents: resisting side quests.

Like, this panel encourages users to get help, share feedback, etc. I'll get more feedback once I iterate the weak "message from Allen" here. Rotate them at least? A/B test? But there are higher leverage todos today!

[Allen Pike @apike]

Guidelines and Guardrails Belong in Git: On making not just your agents work better, but your whole team's. https://t.co/5wg4MByjxm

[Allen Pike @apike]

Made some updates to 2026’s Vancouver Election Guide based on feedback and the latest polling. This is the last week to vote! https://t.co/CYxLnlqvFg

[Allen Pike @apike]

Nailed it. The only way out is talking to (and reading from) other teams about how they actually evaluate what is and isn't working https://t.co/4dPiAHG7XZ

[Allen Pike @apike]

@davidhariri Bonus: a well-defined state machine ensures all the arrows between the screenshots are labeled with exactly the possible transition events between those screens

[Allen Pike @apike]

@jorybraun To your question about no-shows at free meetups, here’s what I learned to do:

1. Use past data to estimate the no-show rate
2. Decide if you’d rather have too much food or not enough, then make peace with it
3. Try to make the event so compelling people prioritize it

[Allen Pike @apike]

Vancouver Election Guide, 2026 edition: 56 Council candidates, 9 parties, and 1 pragmatic attempt to help people make their vote count. https://t.co/CYxLnlqvFg

[Allen Pike @apike]

Telling agents to fix mistakes they made isn’t satisfying. Having them help ensure future agents won’t make the same mistakes is much moreso

[Allen Pike @apike]

@scottastevenson Brevity is most useful when the goal is simply to convey.

More often, the goal is to compel, which benefits from repetition.

[Allen Pike @apike]

Scaling coding agents across teams requires strong guidelines (what does good code look like) and guardrails (rejecting bad code).

Fast system-1 models like Jev are a big opportunity for faster, cheaper, and more plentiful guardrails.

[Ohans Emmanuel @OhansEmmanuel]

best coding agent use case for Jev:

enforcing rules a linter can't

coding agents tend to break rules and some rules can't be codified

Jev can now score every turn against such rules and tell the agent what to fix asap

open-source: https://t.co/yfw9KCm2Cu

[Allen Pike @apike]

If you want trains that can run every 75 seconds, all you need to do is automate!

Otherwise you get merge conflicts, which are annoying for code, but worse for trains

[Allen Pike @apike]

I understand that screaming is cheap and kids love it. There's a reason why Mark Rober started out engineering things to go into space, yet his latest Netflix show is "zany challenges with screaming."

But I do not need to like it

[Allen Pike @apike]

YouTube's parental controls have, for now, deterred our children from getting up at 6am to watch brainrot on the TV.

Feature request: A checkbox to block all content that samples "Industrial Samples & Impacts by Bluezone" aka the Vine Boom. Also screaming

[Allen Pike @apike]

Working title for my talk: “We need to move faster! … Wait, not like that”

[Jory. @jorybraun]

Vancouver: build faster, stay aligned.

Allen Pike (@apike), co-founder of Powerplant, joins BIOS SPHERE to share how teams adapt as coding agents speed up work.

Sponsored by @DevinAI and @cognition.
Nov 2 · Science World, Vancouver
https://t.co/7kMaPQPRXY

[Allen Pike @apike]

Cool thing about letting agents see the prompts behind past work: they can help fix recurring mistakes (e.g. over-verbose UI copy) by checking how we're reprimanding the bots.

All three of us had been pushing back on this mistake, but hadn't yet fixed our shared AGENTS file.

[Allen Pike @apike]

@PaulSolt @davidhariri @AppleSupport $5 says that will fix it! $5 more says one day the problem will recur haha

[Allen Pike @apike]

With macOS 27, we’ve refined the experience from top to bottom, while still making sure to keep classic experiences that users have come to expect from their Mac

[Allen Pike @apike]

@davidhariri @PaulSolt @AppleSupport “But it’s simpler if all search results use the same index. And if there are bugs with index correction and rebuilding, we should fix those, not add a clarifying message for a state the user should never see” – every DRI of this feature over 20 years

[Allen Pike @apike]

Folks saying “It worked for me” may not be aware that preferences occasionally returning no search results has been a recurring event since Mac OS X Tiger, or by modern naming conventions “macOS ‘05”.

It is not a bug – it is a part of our cultural heritage