No Meat Proxy

AI is a tool. You shouldn't be.

Don't paste AI responses to people

✕ Don't

Jim

Hey, how do the feature toggles work? Can we specify per user?

Kevin

The feature flag system here is a config based custom implementation. Definition: a FeatureToggles class — a @ConfigurationProperties data class with boolean fields (e.g. enableNewCheckoutFlow, useOptimizedSearchIndex), registered via @EnableConfigurationProperties in a FeatureConfig class. Usage pattern: inject FeatureToggles and branch on the property, e.g. if (!featureToggles.enableNewCheckoutFlow) throw a ForbiddenException. Some flags gate an early return/exception, others switch between implementations entirely. Configuration is plain YAML under app.feature-toggles.*, with per-profile overrides, and since it's standard @ConfigurationProperties, it's also overridable via env vars. This means that toggling a feature per user is not possible given today's implementation.

✓ Do

Jim

Hey, how do the feature toggles work? Can we specify per user?

Kevin

I'm afraid not. It's a hardcoded on and off switch.

By simply regurgitating what an AI gave, you made everything much harder on the one who asked the question. AI-generated answers tend to be long-winded, over-complicated and often include details that weren't requested. They asked you because they trusted your judgment. By simply forwarding AI output, you're just acting as a middleman. You're just a meat proxy.

If all you do is relay messages between people and an AI, what value do you bring?

By replying with AI answers, you're:

• making it harder for everyone else.

• highlighting your own replaceability.

• not actually being helpful.

• looking unprofessional.

• making people trust you less.

More things to avoid:

• Don't "contribute" to discussion by pasting wall-of-text AI results.

• Don't ask questions like "Hey, so I have this problem: ".

Here is what you should try instead:

• Do read and verify AI output.

• Do formulate your own answers and questions.

• Do contribute to discussions when you have something meaningful to contribute.

Answer people like a person. Treat people like people. Don't be a meat proxy!

Don't submit AI-generated work you don't understand or haven't verified

✕ Don't

function chargeCustomer(order) {

• try {

• processPayment(order);

• } catch (err) {

• logger.warn(err);

• }

}

✓ Do

function chargeCustomer(order) {

• try {

• processPayment(order);

• } catch (err) {

• logger.warn(err);

• }

}

Even if AI did most of the work, you're still responsible for it, and it has your name on it. If your name is on it, it's your work. It is your responsibility to understand, verify and own the work that has been done.

By blindly submitting AI work you're:

• putting the burden on other people to understand and verify your work.

• showing disrespect by forcing others to do what you can't be bothered to do yourself.

• showing others that your work needs to be double-checked.

• looking unprofessional.

• showcasing how replaceable you are.

More things to avoid:

• Don't reply to feedback to your work with AI responses.

• Don't blame the AI when getting feedback.

• Don't excuse the lack of knowledge because "AI did it".

• Don't let standards slip just because AI is involved.

Here is what you should try instead:

• Read, understand and verify the work.

• Reply to feedback as the owner of the work.

• Take responsibility for the work as if it was your own (it has your name on it, so it is yours).

• Show others that you care about what you're doing.

Take ownership over work that has your name on it. Be someone people can rely on. Don't be a meat proxy.