The Brain Sandwich

The whole actual, not necessarily the "openly admitted", utility of AI is to allow people to ship while not understanding the work, and not caring that much about it.

So such posts will make as much change as "trade, but be ethical" would make on a late 80s Wall Street trading floor:

"Before you tell the agent to start building anything, you go there and read and understand it first."

Yeah, not gonna happen. 1 in 100 might do that.

The basic concept here, that you take personal responsibility for the input and output being transmitted to/from the LLM, is an obvious but critical one, so I appreciate this terminology as a way to convey it.

I'll try using Brain Sandwich as guidance for less technical folk.

I fully agree with the approach, but sadly that doesn't make the overwhelming amount of code reviews in the AI age much better. At best, it might slow down your colleagues' deluge of review requests a little. But also, you still have to spend time reviewing your own LLM code, of which there is more of and you're naturally less familiar with than if you'd hand-written it, so it takes longer to review.

Like the concept.

Pedantic addendum: this is an AI sandwich, not a brain sandwich.

> Before you tell the agent to start building anything, you go there and read and understand it first.

The sentence is maybe ambiguous, but it sounds like it's saying to not use the AI at all until you understand what's happening in the area that needs to change.

I don't agree with this as you can spend quite a bit of time digging around to first find the correct place which needs to be updated and then more time reading the code (likely in multiple files or places) to build up a mental model of what's going on.

I typically have more than one AI conversation on the go at any given time, so I'll always start by asking it to make a plan to solve a problem and while it's busy exploring to make a plan I review some other AI conversation's work in the meantime, when the AI's plan is ready I review it.

If at this point there is anything that is unclear to me about its approach to implement the plan, or I don't already have familiarity with the part of the system it's proposing to update, I start questioning the AI on how the system works. I will scrutinize the AI's plan until I'm absolutely sure that it's going to implement something that is sound from a system design perspective and as a senior software developer with architecture level experience I am capable of making such judgments.

I'm not sure what it might be like for comparitively inexperienced developers, but I would expect that AI's capacity for infinite patience allows them to ask as many questions as they need to get a clear mental model of what's happening, so if I were mentoring such a person in this age of AI I would drill into them on how to interact with AIs to ensure they have this clarity before the AI starts implementing anything.

The difference between vibe coding and highly accelerated agentic development is having the expertise and judgment to ensure you're building properly sound software.

Hard to truly understand the work unless if you're actually doing it.

in the post, the brain sandwich gets explained as "brain, ai, brain" but sandwich experts note that a sandwich is usually named after its contents and not the bread

I agree completely, I recently went thru this. Customer API session/journey security is a beast to secure, so many layers, so many moving parts, so many approaches.

> before sending your stuff to other people for review, you need to make the work reviewable.

Yes, but the amount of work needed here is overstated. With good orchestration/prompting in the task[1] (and token budget), the work on this end is significantly reduced.

… and then the brain drain issue returns.

[1] https://claude.dev/blog/solving-code-review-at-anthropic/

This is far better than just hitting return, but there's zero chance it's as good as doing things the hard way. Tradeoffs, of course.

This is the same approach I use. I don't know if it actually produces good results yet as I've only been doing this for about three months. At the very least it feels like I have some agency in my work which I think is one of the things people complain they lose to AI.

Do people not use AI this way in prod?

sounds like 'design' is back

"Other people use slop the wrong way."

https://en.wikipedia.org/wiki/Third-person_effect

I have a very simple framework I use for working with John von Neumann ... the idea is very simple, and can be summarized by: whenever you work on something with John von Neumann, you should use:

1. Your brain, and then

2. John von Neumann, and then

3. Your brain

Explaining a bit more… Before you tell John von Neumann to start working on anything, you go there and read and understand it first. And when John von Neumann brings you the solution, you will know if it’s right, or if it’s too much, or if it’s completely solving the wrong problem, etc.

[dead]

This approach appears valuable for personal growth as an engineer, yet it demands discipline that may not yield returns in corporate settings. Specifically:

> The rule of thumb is: if you can’t explain what’s changing and why, it’s not ready for someone else to take a look

I have observed firsthand that the metric or "understanding" can be irrelevant to management. When delivery dominates, understanding can be deferred to future agents who will gladly consume your codebase and convince you they grasp it, and further, that you do too. If I invest substantial effort in my PRs only for them to be /review'd, does it still matter? Perhaps reviewers should practice brain sandwich as well.

The worst offenders seem to sit at either extreme: fully AI firms that treat understanding as a human weakness and prioritize outcomes alone, or large corporate teams building internal CRUD apps whose middle management merely pretends to value good engineering to satisfy finicky senior engineers.

Not a criticism on the authors work at all, I like the idea (and I don't think they were trying to solve anything to do with work culture), but ultimately if my company doesn't find my practice valuable, I can't easily justify taking it on. Especially if I have to sell to them that I'm going to be spending _more_ time delivering. Fun times!