AI vocabulary · it is important to note
"It is important to note": why the data doesn't back this one
This is one of the phrases people mean when they say a piece of writing sounds like AI, and it earned that reputation somewhere, probably in essay-mill drafts and long-form report generation rather than the advice-and-conversation register underslop logged. What the evidence shows: in the twelve-model, dozens-of-conversation corpus underslop is built on, the phrase never appeared once, despite an explicit search for it. That does not mean it never happens anywhere. It means underslop does not score a rule its own logs cannot back.
"It is important to note" shows up on almost every AI-tell list, but underslop's own logs do not back it as a reliable signal. Across the twelve-model corpus underslop is built on, a direct search for it, alongside Furthermore, Moreover, and In conclusion, turned up zero instances. Those same models suppress hedging generally, favoring flat declaratives over qualifiers, so a blanket importance-flag phrase is rarer in practice than its reputation suggests. underslop's shipped linter does not score it. The fix is still worth making when you do spot it: delete the frame and check whether the claim underneath survives on its own.
What reaching for it signals
The mechanism, when this phrase does appear, is a model inserting an evaluative label about a claim's significance instead of letting the claim demonstrate it. "It is important to note that X" adds nothing to X; it is metacommentary asserting X deserves attention, placed in front of the sentence rather than earned by what the sentence says. It behaves like a hedge in shape, a soft framing device, but like a booster in effect, since it raises the reader's expectation before the content arrives. Given that these models otherwise suppress genuine epistemic hedging, might, perhaps, I think, in favor of flat assertions, an importance flag would be an odd substitute: it announces uncertainty is being managed without doing any of the managing.
Whether underslop counts it
underslop's shipped linter does not score "it is important to note," and it isn't even on the reported, calibration-only list. That's a deliberate omission: the full twelve-model corpus was checked for this exact phrase, along with Furthermore and In conclusion, and returned zero hits. underslop separately calibrates importantly and notably as report-only adverbs, a thinner version of the same habit, but the full phrase itself isn't something the evidence supports scoring. If a corpus in a different register, formal reports rather than advice chat, turns up real frequency, underslop will add it then, not before.
What to write instead
- It is important to note that the deadline moved.The deadline moved.
- state the fact plainly; if it needs emphasis, put it first in the paragraph rather than announcing it
- It's worth noting that the sample size was small.The sample size was 40, which limits how far this generalizes.
- replace the vague flag with the number that actually makes it worth noting
- Importantly, the results held across all three trials.The results held across all three trials.
- let the count carry the weight instead of the adverb; add it back only if a reader could plausibly expect otherwise
- One important consideration is cost.Cost is the constraint: the budget caps this at twelve thousand dollars.
- fine as a transition when the actual number follows immediately, not when it stands alone as filler
Examples
It's worth noting that response times improved after the update. It is important to note that this isn't just a minor tweak, it's a full rework, and some edge cases remain.
Response times improved after the update. This isn't a minor tweak—it's a full rework, and some edge cases remain.
It is important to note that this recommendation depends on your specific circumstances. I hope this helps!
This recommendation depends on your specific circumstances.
Questions
What are common AI phrases in writing?
Lists vary, but the recurring names are validation openers that praise the question before answering it, offer-to-continue sign-offs, and structural moves like the negative-parallelism reframe. underslop scores all three because its own logs and independent reviewers both confirm them at real frequency. This phrase is different: it is widely cited, but the search for it came back empty.
Is it is important to note always an AI tell?
The evidence doesn't support calling it an AI tell at all. It is a genuine hedge phrase used by careful human writers too, and the twelve-model corpus built specifically to catch AI patterns did not produce it once.
Why doesn't underslop's linter flag this phrase?
Because scoring it would mean claiming evidence underslop does not have. The scored list is built from tells actually observed in underslop's own model logs or that at least two independent style reviews corroborate. This phrase cleared neither bar.
Should I still remove it from my own writing?
If a sentence works without the frame, cut it regardless of whether it's an AI signal. The check is simple: read the sentence without it. If the claim still stands, the frame was padding.
What are common AI patterns in writing generally?
The patterns with real evidence behind them are behavioral more than lexical: opening with unearned validation, closing with an offer to keep helping, and converting plain answers into bullet lists with bold labels. Single stock phrases like this one turn out to be weaker signals than the shape of the whole response.
underslop scores a small set of vocabulary tells and computes a larger greylist it deliberately does not count. Paste a draft to see which of the two your writing is tripping. The free tier caps at 500 words a run.