痛点为 AI 基于上游原始证据的初步提炼；未包含额外中国市场检索。
用户在使用LLM辅助写作时，最初觉得AI生成的文本比自己写的好很多，但几个月后发现整个互联网上充斥着相同的句子结构和表达模式，导致AI辅助写作的内容失去独特性。这种“AI味”让用户难以判断自己产出的质量，因为当自己不擅长某个领域时，很容易被AI的表面流畅性迷惑，从而依赖AI输出，最终造成个人风格被稀释、内容同质化严重。评论中用户指出LLM写作的“千篇一律”是坏的，建议只让LLM做结构批评等客观改进，而不直接使用其生成的词汇，说明现有流程中用户难以在提升写作质量和保持个人风格之间取得平衡，容易陷入重复劳动和风格丧失的困境。
Late last year I started writing a math blog and decided to use LLMs to polish/enhance my writing. The LLM generated writing obviously felt significantly bet...
> The LLM generated writing obviously felt significantly better than my own writing. A general pattern for LLMs is that they look really good at things you are bad at. What that means is that if you find yourself thinking of its output as significantly better than yours in a particular domain, there's a high chance that you are not equipped to judge that quality effectively.
The LLM writing sameness is bad. Use LLMs to help your writing! But don't include a word they generate, even just a vocabulary adjustment, in your own output. Have them critique structure and flow, spot overused words and passive constructions and dumb picks for topic sentences. It's great for that, and those are all objective improvements in your writing that won't mess up your style. The LLM sameness in web design is good. Most sites shouldn't try to be idiosyncratic. The best design for a site with real utility is legibility , and LLMs are better at that than the median developer. Always laying out the same buttons? Always using the same type scales? Good! If it looks good to you, you weren't going to do better on your own, and you were very likely to do worse.
- “(The) honest caveat:” (or “genuine caveat:”, both with the colon) - “(The) honest answer:” (again, with colon) - “The thing to internalize:” - “The smoking gun:” (really, sentences that start with “The :” are a strong tell, but those four are the most prolific) - “load bearing” (when not talking about architecture) - “blast radius” (when not talking about actual explosives, but rather the effect of an event/action) - “smoke test” (esp. when “sanity check” is more apropos) - Lists of three clauses/adjectives where the third is really just a combination of the first two - Referring to the “shape” of things figuratively - Social media posts that end with “Curious if anyone…” - Stories or anecdotes using. “Oh. Oh.” (where the second “oh” is italicized) Edit: Yes, some of those last ones are terms that we often use as devs...but I would argue about the actual frequency of their use. Plus, these tells live on in prose generated by the latest models.
Scrolling down a LinkedIn feed is hilarious at the moment. My favourite one today from today: “The tax isn't the problem. The mindset is.”
Thank you, these are all things I've noticed too.
{
"author": "speckx",
"hn_item_id": 48313810,
"external_url": "https://shvbsle.in/various-llm-smells/"
}{
"by": "speckx",
"id": 48313810,
"url": "https://shvbsle.in/various-llm-smells/",
"kids": [
48315309,
48316058,
48315567,
48315598,
48316097,
48315159,
48315654,
48315970,
48315281,
48315282,
48315948,
48315480,
48315242,
48315299,
48315487,
48315250,
48315554,
48315997,
48315770
],
"time": 1779994938,
"type": "story",
"score": 88,
"title": "Various LLM Smells",
"descendants": 50
}{
"id": "6870c0aa-beb1-4045-8065-a613905fe39d",
"daily_ranking_item_id": "6c39e276-281e-4237-9899-02f0e44dbff0",
"source": "hacker_news",
"external_id": "48313810",
"fetched_at": "2026-05-28T22:01:23.716Z",
"story_raw": {
"by": "speckx",
"id": 48313810,
"url": "https://shvbsle.in/various-llm-smells/",
"kids": [
48315309,
48316058,
48315567,
48315598,
48316097,
48315159,
48315654,
48315970,
48315281,
48315282,
48315948,
48315480,
48315242,
48315299,
48315487,
48315250,
48315554,
48315997,
48315770
],
"time": 1779994938,
"type": "story",
"score": 88,
"title": "Various LLM Smells",
"descendants": 50
},
"stats_raw": {
"time": 1779994938,
"score": 88,
"descendants": 50
},
"aux_raw": {
"external_url": "https://shvbsle.in/various-llm-smells/",
"hn_comment_url": "https://news.ycombinator.com/item?id=48313810",
"normalized_text": null,
"external_article": {
"title": "Various LLM smells",
"excerpt": "Late last year I started writing a math blog and decided to use LLMs to polish/enhance my writing. The LLM generated writing obviously felt significantly better than my own writing. It had better vocabulary, interesting sentence structures etc etc. I swear it did not seem like AI-slop to me at the time. Then about 3 months later, I see the exact sentence structures appearing ACROSS THE ENTIRE F***** INTERNET. And what is fascinating to me is that ai-smell seems like an artifact that emerges across various AI assisted tasks that you can now easily recognize. A few examples that I've collected so far to show the \"ai-smells\" across two domains:\n\nSome picks from my math blog (now deleted) and the drafts that accompanied it",
"final_url": "https://shvbsle.in/various-llm-smells/",
"fetched_at": "2026-05-28T22:01:17.855Z",
"description": "Late last year I started writing a math blog and decided to use LLMs to polish/enhance my writing. The LLM generated writing obviously felt significantly bet..."
},
"selected_comments": [
{
"id": 48315309,
"raw": {
"by": "Planktonne",
"id": 48315309,
"kids": [
48315795,
48315365,
48315492,
48315351
],
"text": "> The LLM generated writing obviously felt significantly better than my own writing.<p>A general pattern for LLMs is that they look really good at things you are bad at. What that means is that if you find yourself thinking of its output as <i>significantly</i> better than yours in a particular domain, there's a high chance that you are not equipped to judge that quality effectively.",
"time": 1780001634,
"type": "comment",
"parent": 48313810
},
"body": "> The LLM generated writing obviously felt significantly better than my own writing. A general pattern for LLMs is that they look really good at things you are bad at. What that means is that if you find yourself thinking of its output as significantly better than yours in a particular domain, there's a high chance that you are not equipped to judge that quality effectively.",
"is_op": false,
"author": "Planktonne",
"raw_body": "> The LLM generated writing obviously felt significantly better than my own writing.<p>A general pattern for LLMs is that they look really good at things you are bad at. What that means is that if you find yourself thinking of its output as <i>significantly</i> better than yours in a particular domain, there's a high chance that you are not equipped to judge that quality effectively.",
"created_at": 1780001634,
"reply_count": 4
},
{
"id": 48316058,
"raw": {
"by": "tptacek",
"id": 48316058,
"text": "The LLM writing sameness is bad. Use LLMs to help your writing! But don't include a word they generate, even just a vocabulary adjustment, in your own output. Have them critique structure and flow, spot overused words and passive constructions and dumb picks for topic sentences. It's great for that, and those are all objective improvements in your writing that won't mess up your style.<p>The LLM sameness in web design is good. Most sites shouldn't try to be idiosyncratic. The best design for a site with real utility is <i>legibility</i>, and LLMs are better at that than the median developer. Always laying out the same buttons? Always using the same type scales? Good! If it looks good to you, you weren't going to do better on your own, and you were very likely to do worse.",
"time": 1780005297,
"type": "comment",
"parent": 48313810
},
"body": "The LLM writing sameness is bad. Use LLMs to help your writing! But don't include a word they generate, even just a vocabulary adjustment, in your own output. Have them critique structure and flow, spot overused words and passive constructions and dumb picks for topic sentences. It's great for that, and those are all objective improvements in your writing that won't mess up your style. The LLM sameness in web design is good. Most sites shouldn't try to be idiosyncratic. The best design for a site with real utility is legibility , and LLMs are better at that than the median developer. Always laying out the same buttons? Always using the same type scales? Good! If it looks good to you, you weren't going to do better on your own, and you were very likely to do worse.",
"is_op": false,
"author": "tptacek",
"raw_body": "The LLM writing sameness is bad. Use LLMs to help your writing! But don't include a word they generate, even just a vocabulary adjustment, in your own output. Have them critique structure and flow, spot overused words and passive constructions and dumb picks for topic sentences. It's great for that, and those are all objective improvements in your writing that won't mess up your style.<p>The LLM sameness in web design is good. Most sites shouldn't try to be idiosyncratic. The best design for a site with real utility is <i>legibility</i>, and LLMs are better at that than the median developer. Always laying out the same buttons? Always using the same type scales? Good! If it looks good to you, you weren't going to do better on your own, and you were very likely to do worse.",
"created_at": 1780005297,
"reply_count": 0
},
{
"id": 48315567,
"raw": {
"by": "spdustin",
"id": 48315567,
"kids": [
48315985,
48315869
],
"text": "- “(The) honest caveat:” (or “genuine caveat:”, both with the colon)<p>- “(The) honest answer:” (again, with colon)<p>- “The thing to internalize:”<p>- “The smoking gun:”<p>(really, sentences that start with “The <tag suggesting the next clause is the key point>:” are a strong tell, but those four are the most prolific)<p>- “load bearing” (when not talking about architecture)<p>- “blast radius” (when not talking about actual explosives, but rather the effect of an event/action)<p>- “smoke test” (esp. when “sanity check” is more apropos)<p>- Lists of three clauses/adjectives where the third is really just a combination of the first two<p>- Referring to the “shape” of things figuratively<p>- Social media posts that end with “Curious if anyone…”<p>- Stories or anecdotes using. “Oh. Oh.” (where the second “oh” is italicized)<p>Edit: Yes, some of those last ones are terms that we often use as devs...but I would argue about the actual frequency of their use. Plus, these tells live on in prose generated by the latest models.",
"time": 1780002815,
"type": "comment",
"parent": 48313810
},
"body": "- “(The) honest caveat:” (or “genuine caveat:”, both with the colon) - “(The) honest answer:” (again, with colon) - “The thing to internalize:” - “The smoking gun:” (really, sentences that start with “The <tag suggesting the next clause is the key point>:” are a strong tell, but those four are the most prolific) - “load bearing” (when not talking about architecture) - “blast radius” (when not talking about actual explosives, but rather the effect of an event/action) - “smoke test” (esp. when “sanity check” is more apropos) - Lists of three clauses/adjectives where the third is really just a combination of the first two - Referring to the “shape” of things figuratively - Social media posts that end with “Curious if anyone…” - Stories or anecdotes using. “Oh. Oh.” (where the second “oh” is italicized) Edit: Yes, some of those last ones are terms that we often use as devs...but I would argue about the actual frequency of their use. Plus, these tells live on in prose generated by the latest models.",
"is_op": false,
"author": "spdustin",
"raw_body": "- “(The) honest caveat:” (or “genuine caveat:”, both with the colon)<p>- “(The) honest answer:” (again, with colon)<p>- “The thing to internalize:”<p>- “The smoking gun:”<p>(really, sentences that start with “The <tag suggesting the next clause is the key point>:” are a strong tell, but those four are the most prolific)<p>- “load bearing” (when not talking about architecture)<p>- “blast radius” (when not talking about actual explosives, but rather the effect of an event/action)<p>- “smoke test” (esp. when “sanity check” is more apropos)<p>- Lists of three clauses/adjectives where the third is really just a combination of the first two<p>- Referring to the “shape” of things figuratively<p>- Social media posts that end with “Curious if anyone…”<p>- Stories or anecdotes using. “Oh. Oh.” (where the second “oh” is italicized)<p>Edit: Yes, some of those last ones are terms that we often use as devs...but I would argue about the actual frequency of their use. Plus, these tells live on in prose generated by the latest models.",
"created_at": 1780002815,
"reply_count": 2
},
{
"id": 48315598,
"raw": {
"by": "rimeice",
"id": 48315598,
"kids": [
48316068
],
"text": "Scrolling down a LinkedIn feed is hilarious at the moment.<p>My favourite one today from today:<p>“The tax isn't the problem. The mindset is.”",
"time": 1780002922,
"type": "comment",
"parent": 48313810
},
"body": "Scrolling down a LinkedIn feed is hilarious at the moment. My favourite one today from today: “The tax isn't the problem. The mindset is.”",
"is_op": false,
"author": "rimeice",
"raw_body": "Scrolling down a LinkedIn feed is hilarious at the moment.<p>My favourite one today from today:<p>“The tax isn't the problem. The mindset is.”",
"created_at": 1780002922,
"reply_count": 1
},
{
"id": 48316097,
"raw": {
"by": "newer_vienna",
"id": 48316097,
"text": "Thank you, these are all things I've noticed too.",
"time": 1780005554,
"type": "comment",
"parent": 48313810
},
"body": "Thank you, these are all things I've noticed too.",
"is_op": false,
"author": "newer_vienna",
"raw_body": "Thank you, these are all things I've noticed too.",
"created_at": 1780005554,
"reply_count": 0
}
],
"presentation_fields": {
"title": "Various LLM Smells",
"tagline": "shvbsle.in",
"website_url": "https://shvbsle.in/various-llm-smells/",
"canonical_url": "https://news.ycombinator.com/item?id=48313810"
},
"external_url_hostname": "shvbsle.in",
"selected_comments_raw": [
{
"by": "Planktonne",
"id": 48315309,
"kids": [
48315795,
48315365,
48315492,
48315351
],
"text": "> The LLM generated writing obviously felt significantly better than my own writing.<p>A general pattern for LLMs is that they look really good at things you are bad at. What that means is that if you find yourself thinking of its output as <i>significantly</i> better than yours in a particular domain, there's a high chance that you are not equipped to judge that quality effectively.",
"time": 1780001634,
"type": "comment",
"parent": 48313810
},
{
"by": "tptacek",
"id": 48316058,
"text": "The LLM writing sameness is bad. Use LLMs to help your writing! But don't include a word they generate, even just a vocabulary adjustment, in your own output. Have them critique structure and flow, spot overused words and passive constructions and dumb picks for topic sentences. It's great for that, and those are all objective improvements in your writing that won't mess up your style.<p>The LLM sameness in web design is good. Most sites shouldn't try to be idiosyncratic. The best design for a site with real utility is <i>legibility</i>, and LLMs are better at that than the median developer. Always laying out the same buttons? Always using the same type scales? Good! If it looks good to you, you weren't going to do better on your own, and you were very likely to do worse.",
"time": 1780005297,
"type": "comment",
"parent": 48313810
},
{
"by": "spdustin",
"id": 48315567,
"kids": [
48315985,
48315869
],
"text": "- “(The) honest caveat:” (or “genuine caveat:”, both with the colon)<p>- “(The) honest answer:” (again, with colon)<p>- “The thing to internalize:”<p>- “The smoking gun:”<p>(really, sentences that start with “The <tag suggesting the next clause is the key point>:” are a strong tell, but those four are the most prolific)<p>- “load bearing” (when not talking about architecture)<p>- “blast radius” (when not talking about actual explosives, but rather the effect of an event/action)<p>- “smoke test” (esp. when “sanity check” is more apropos)<p>- Lists of three clauses/adjectives where the third is really just a combination of the first two<p>- Referring to the “shape” of things figuratively<p>- Social media posts that end with “Curious if anyone…”<p>- Stories or anecdotes using. “Oh. Oh.” (where the second “oh” is italicized)<p>Edit: Yes, some of those last ones are terms that we often use as devs...but I would argue about the actual frequency of their use. Plus, these tells live on in prose generated by the latest models.",
"time": 1780002815,
"type": "comment",
"parent": 48313810
},
{
"by": "rimeice",
"id": 48315598,
"kids": [
48316068
],
"text": "Scrolling down a LinkedIn feed is hilarious at the moment.<p>My favourite one today from today:<p>“The tax isn't the problem. The mindset is.”",
"time": 1780002922,
"type": "comment",
"parent": 48313810
},
{
"by": "newer_vienna",
"id": 48316097,
"text": "Thank you, these are all things I've noticed too.",
"time": 1780005554,
"type": "comment",
"parent": 48313810
}
]
},
"selection_meta": {
"discussion_depth": "top_comments_v1",
"external_article": {
"status": "ok",
"final_url": "https://shvbsle.in/various-llm-smells/",
"status_code": 200,
"content_type": "text/html; charset=utf-8",
"failure_reason": null
},
"snapshot_version": "hn_story_v3",
"selected_comments_count": 5,
"external_article_resolved": true,
"text_normalization_applied": false
},
"created_at": "2026-05-28T22:01:23.852Z",
"updated_at": "2026-05-28T22:01:23.852Z"
}