If you honestly can't see why this prompt from the get go was a joke, them you may have to cede that LLM have a better grasp as the subtleties of language than you expect.
This is what I don't get. I feel as though the people complaining about this are not the primary source of the information anyway. I wonder if Google already has a way for websites to opt out of the AI mode results (of course, since their sites are not adding any actual new information, there will not be a loss to them). From the search results, it seems like Google has constructed a "Knowledge Graph" LLM which it uses to answer questions in the search results and provide links to sources. How is that different from how every other LLM works?
There also seems to be a second issue about Google using YouTuber videos without their consent to train AI, which may be the more pertinent issue the EU is investigating.
I use it with Angular and Svelte and it works pretty well. I used to use Lit, which at least the older models did pretty bad at, but it is less known so expected.
I found that LLMs sometimes get confused by Lit because they don’t understand the limitations of the shadow DOM. So they’ll do something like throw an event and try to catch it from a parent and treat it normally, not realising that the shadow DOM screws that all up, or they assume global / reset CSS will apply globally when you actually need to reapply it to every single component.
What I find interesting is all the platforms like Lovable etc. seem to be choosing Supabase, and LLMs are pretty terrible with that – constantly getting RLS wrong etc.
The Fourier Transform isn't even Fourier's deepest insight. Unless we're now ranking scientific discoveries based on whether or not they get a post every weekend on HN.
The FFT is nifty but that's FINO. The Google boys also had a few O(N^2) to O(N log N) moments. Those seemed to move the needle a bit as well.
But even if we restrict to "things that made Nano Banana Pro possible" Shannon and Turing leapfrog Fourier.
It's really getting in the way of all the daily AI opinion pieces I come here to read.
More seriously, there are tens of thousands of people who come to HN. If Fourier stuff gets upvoted, it's because people find it informative. I happen to know the theory, but I wouldn't gatekeep.
It to the point that bloggers know they can say things like "OpenAI did 9/11!", provide no sources, and it still will make the front page and have a large increase in traffic to there site.
2) Most of those, while they had the two statements, the statements were not in succession.
There are maybe 4 unique examples in the search over the past 15 years, which is why it is very telling when there is an explosion of the pattern seen today, and that is most likely due to LLMs.
I was responding in particular to the "you write like a late night kitchen gizmo ad?" ... which would be a speech pattern people hear. In the audio case, it doesn't matter what punctuation symbol separates the "it isn't/it's" pattern because the comma or em dash would be invisible.
>There are maybe 4 unique examples in the search over the past 15 years,
No, (1) the Algolia search engine HN uses is not exhaustive and always returns incomplete results, and (2) I couldn't construct a regex to capture all occurrences. It didn't capture the dozens of times I used it before 2022.
More pre-2022 examples that match the "it isn't/it's" pattern that the blog author is complaining about :
The same gp mentioned that it's also common in "ad copy". That's also true with the famous Navy's "It's not just a job. It's an adventure.". E.g. 1981 tv commercial: https://www.youtube.com/watch?v=Tc9g2tagYms
That's a slogan people heard rather than read with an em dash. LLM engines picked up on a common phrasing used for decades.
I understand that there are multiple people in this conversation, but you are attempting to pick and choose points to discuss at the expense of your own internal consistency. If you were responding to "which would be a speech pattern people hear," why did you only quote written examples from the HN search and not provide video or audio clips?
>why did you only quote written examples from the HN search and not provide video or audio clips?
At the risk of stating the obvious, highlighting the HN _texts_ demonstrates in a very literal way the "write like" fragment in gp's question, "You write like a late night kitchen gizmo ad?. The other fragment was the "late night kitchen gizmo ad" which is the audio comparison. The gp was making that comparison between the writing style and the speech style when asking the question. (https://news.ycombinator.com/item?id=46165248)
Providing audio links would not show the "writes like". The gp (and you) already know what the "It isn't/It's" audio pattern sounds like. It's the written text the gp was wondering about.
The point is people really did write text like that (no em dashes required) before ChatGPT existed.
EDIT reply to: >He just said that it is traditionally associated with late-night ads, and that the explosion in use of the phrase (especially with the em-dash)
Actually, the gp (0_____0) I was responding to didn't mention the em dash in either of the 2 comments. Gp used a comma instead of em dash. Gp only mentioned the comparison to ad copy. The em dash wasn't relevant in the subthread we're in. That's something extra you brought up that's not related to gp's specific question.
EDIT reply to: >Quick HN tip: It is usually better to reply to a post instead of editing the original post.
I agree but the "reply" option was not available. This is a "cool down" mechanism HN uses to discourage flame wars. I don't know if it's 30 minutes or what the value is before the reply link shows up. It was just easier to reply in my post rather than wait an indeterminate time.
>This statement is incorrect, as the original post mentioned, "'it's not just x — it's y' format is the hallmark
Yes but that's not the ggp (ceroxylon) I was responding to. Instead, I was responding t gp (0_____0)'s question and the 2 times the writing was compared to ad copy with no mention of em dashes. Sorry for not making that clear.
>Showing fewer than a dozen uses of the phrase
Again, there are thousands of examples but the Algolia search engine will not show all of them.
Quick HN tip: It is usually better to reply to a post instead of editing the original post.
>Actually, the GP (0_____0) I was responding to didn't mention the em dash in either of the two comments. GP used a comma instead of an em dash. GP only mentioned the comparison to ad copy. The em dash wasn't relevant in the subthread we're in. That's something you brought up.
This statement is incorrect, as the original post mentioned, "'it's not just x — it's y' format is the hallmark of mediocre articles being written/edited by AI." (note the quotes in the first post), and the next post said, "It's simply how literate people write."
All of this is beside the point, however, because your statement, "The point is people really did write text like that (no em dashes required) before ChatGPT existed," was never contended in this thread, and I do not think anyone has ever thought that ChatGPT created that phrase, so it just doesn't add to the discussion. Showing fewer than a dozen uses of the phrase (with or without the em dash) in a 15-year period just further proves that it was not a common written turn of phrase before ChatGPT.
>The point is people really did write text like that (no em dashes required) before ChatGPT existed.
OK, I think I can see your point, but at best it is irrelevant. At no point did the original poster imply that ChatGPT created the phrase, or that it wasn’t in spoken or written language before then. He just said that it is traditionally associated with late-night ads, and that the explosion in use of the phrase (especially with the em-dash) is most likely attributed to increased LLM use.
That might drift in the future. I've actually found myself leaving small errors in sometimes since it suggests that I actually wrote it. I don't use literal em-dashes -- but I often use the manual version and have been doing so much longer than mainstream LLMs have been around. I also use a lot of bulleted lists -- both of which imply LLM usage. I take my writing seriously, even when it's just an internet comment. The idea that people might think I wrote with an LLM would be insulting.
But further and to the point, spelling / grammar errors might be a boutique sign of authenticity, much like fake "hand-made" goods with intentional errors or aging added in the factory.
Unless you are using a proprietary, dedicated grammar checker, auto grammar check is far from perfect and will miss some subject-verb agreement errors, incorrect use of idioms, or choppy flow. Particularly in professional environments where you are being evaluated, this can tank an otherwise solid piece of written work. Even online in HN comments, people will poke fun at grammar, and (while I don't have objective evidence for this) I have noticed that posts with poor grammar or misspellings tend to have less engagement overall. In a perfect world, this wouldn't matter, but it's a huge driving factor for why people use LLMs to touch up their writing.
That we commonly used em-dashes as a mark to set off parenthetical information. Yes, you can also use parentheses and they're somewhat interchangeable.
reply