Community guides · Technology
How to comment in r/artificialintelligence without getting downvoted
Reply in r/artificialintelligence and your code-review instincts are showing. The community spots vague answers, marketing-speak, and 'just use X' shortcuts instantly.
What this sub rewards
How replies earn upvotes here
- Acknowledge limitations, trade-offs, or uncertainty when relevant. Name what the approach doesn't solve. Universal claims get downvoted.
- Senior-level technical depth is expected; avoid generic advice. Back claims with measurement — 'P95 latency dropped from 800ms to 230ms after switching to Redis'.
- Measured and factual; no hype, hyperbole, or breathless speculation. Stick to current tense and verified state. 'May enable' / 'expected to' triggers skepticism.
The one rule that gets posts removed
Distinguish practical AI from speculative futurism.
Your pre-reply checklist
Before you hit Reply — four quick checks
Does my comment add something specific, not just an opinion?
A useful reply names a number, a method, a personal experience, or a concrete example. If you could replace your draft with "I agree" and lose nothing, rewrite.
Have I read the top 3 comments already on this thread?
If the highest-upvoted reply already says what you were about to say, your comment will sink. Add what's missing — counterexamples, edge cases, the next step — not what's there.
Would my reply still work if my username was removed?
If it only makes sense because it promotes you, your product, or your platform, this sub will spot it. Frame the value as "this solved it for me," not "check out my…"
Did I skim the sidebar rules one more time?
Every sub has at least one rule that surprises outsiders — peer-review only, no images, no specific tags, no off-topic. Five seconds in the sidebar can save you a removal.
Want to draft this reply with the rules already baked in?
Try Redimates — free, 10 replies / week