Great topic. 

There was a case some time ago when someone had commented on a post of mine that got me suspicious, I went on to check her other recent comments and noticed they were both in big chunks similar to this paragraph will become I assume in the span of a minute or two. I kept observing the account and while it was being careful, it kept making a few mistakes here and there as the usual way to consume content on Hive would be to read a post, maybe not fully but still enough to leave a long comment, then move on to another post and consume that one and leave another lengthy comment, but doing so takes time for most humans, even if you're the fastest reader and fastest typer it's going to take longer than a minute to leave comments of this length on two separate posts while also having consumed them. 

I asked the author if they were using AI to consume and generate comments but of course they said no but the reply was like 8h later after having left a few more comments here and there on other posts. Either way, after more digging we noticed they had also plagiarised posts in the past, etc. 

My point is, similar to some users who may be jumping from alt account to alt account to try avoid curation project voting cooldowns while keeping both accounts active enough to pretend to be regular users - one fuck up and your reputation is usually toast. So I guess the same goes for those trying to cheat their way with pretending to be consuming and engaging with posts through close to no effort other than copy-pasting posts to consume and generated comments to reply with through an AI platform.  

Authors in general do better the longer they can uphold a healthy reputation so hopefully people aren't dumb enough to stoop down to these levels to fake activity in an effort to maybe get rewarded more because they may hit a wall sooner or later and wish they could go back to just leaving regular comments now and then and earning alongside everyone else not attempting any cheap tricks. 

Yeah it’s pretty funny to think that people are so lazy that they throw a post in AI to read it and then ask AI to come up with a comment for it lol wtf bros. 

I use AI in some degree but mainly just to use the PeakD post summary feature. Helps me get to many posts in a session for comments! I didn’t use it on this one though :D

Lately it seems like AI really overdoes it on the em dashes. Which really stinks, because I have always liked using em dashes in my writing but now I deliberately try to avoid them just because I don't want to be mistaken for AI. 

For awhile the questions at the end designed to increase engagement was a very big telltale sign of AI. But it is getting a lot better. For awhile it was also really going overboard on Gen Z slang, but that is also being dialed back lately, making it more tricky to tell. 

I really can't tell in that second comment you post. Slang spelling and casual abbreviations, incorrect grammar, dropping words... AI now does all of that. I guess it would depend on who wrote it and if I trust them or not. 

[Amusingly enough, I have a post about my usage of em dashes scheduled to drop tomorrow morning. But despite that post, I now find myself using them less and less because of AI]

I must admit I sometimes check posts with an AI detector before I vote on them if the language is just a little too flowery. I think you can usually tell by the vocabulary and the too perfect grammar...and I'm always gentle!

I never quite know what to think. I love reading about AI advances. Beyond the simple generation of text, which is just the upper part of the iceberg, the rest is too interesting not to be interested in. If only for research.

Being French, with very bad English, I use DeepL's API to translate my text.

When I write a comment, it's often on my phone and my keyboard offers me a translation of my text. Honestly, I don't know what my English looks like from the point of view of a native speaker. All I know is that I can make myself understood much better than with my own English. It is a major step forward to be able, easily, to make oneself understood by more people. 

To generate a translation, DeepL uses seven years of data to power its AI. So it's actually AI-generated content. But I think it's still human enough to be enjoyable to read. 

On the other hand, I know that there is a risk of AI spam being used so badly. 

I think it's about education rather than punishment. For example, by creating tutorials on "how to use ChatGPT/Claude/Mistral ethically on Hive."  To promote this, use personal photos, screenshots, videos, and real-life experiences.

Again, I'm unsure if my AI use is ethical. I believe it is, but if not, I'd rather be told frankly than punished.

Congratulations @riverflows! You have completed the following achievement on the Hive blockchain And have been rewarded with New badge(s)

<table><tr><td><img src="https://images.hive.blog/60x70/https://hivebuzz.me/@riverflows/upvoted.png?202505020920"></td><td>You received more than 740000 upvotes.<br>Your next target is to reach 750000 upvotes.</td></tr>
</table>

<sub>_You can view your badges on [your board](https://hivebuzz.me/@riverflows) and compare yourself to others in the [Ranking](https://hivebuzz.me/ranking)_</sub>
<sub>_If you no longer want to receive notifications, reply to this comment with the word_ `STOP`</sub>



**Check out our last posts:**
<table><tr><td><a href="/hive-122221/@hivebuzz/pud-202505-feedback"><img src="https://images.hive.blog/64x128/https://i.imgur.com/zHjYI1k.jpg"></a></td><td><a href="/hive-122221/@hivebuzz/pud-202505-feedback">Feedback from the May Hive Power Up Day</a></td></tr><tr><td><a href="/hive-122221/@hivebuzz/pum-202504-result"><img src="https://images.hive.blog/64x128/https://i.imgur.com/mzwqdSL.png"></a></td><td><a href="/hive-122221/@hivebuzz/pum-202504-result">Hive Power Up Month Challenge - April 2025 Winners List</a></td></tr><tr><td><a href="/hive-122221/@hivebuzz/pum-202505"><img src="https://images.hive.blog/64x128/https://i.imgur.com/M9RD8KS.png"></a></td><td><a href="/hive-122221/@hivebuzz/pum-202505">Be ready for the May edition of the Hive Power Up Month!</a></td></tr></table>

Want people to know it's you, the actual human, making a comment? Be *specific.* Never be vaguely congenial and complimentary, like, "I appreciate the perspective you offer in this post. It's refreshing to see your creativity here!"

Oh, I can totally tell the difference between the two comments. First one is definitely AI, because why does it feel like it did a simple recap on everything you had said? And also, should I say it's too verbose? I'm careful about this because I get very wordy in my writing as well, and use big words that may look like I was sitting beside a dictionary to write it, or probably quoting AI. 

But honestly though, I think AI is getting better at not sounding generic. I just hope that I can still spot them out. Most times I start praising an author for being so dexterous with words, and then find later that AI was employed in everything. Disheartening, I'll tell you. 

Ohh such a disgrace in your self if you are using AI to do a comment. do we know longer have the brains to think

The high dependency on Ai for almost everything is really alarming, we have brains and feelings, can we put them to use please.

Im super annoyed by it honestly
 (Have been digging through a ton of application letters this week and a *lot* of them have this AI  feeling to it. I get that you use it for inspiration, but come onn....if you are not putting in human touch.. whats the point)

Usually when I see a sort of AI comment i check what what other comments they leave on there and even make some remake to it.

You can just see it (i guess this will be different in a year of time)

I've never used AI to write *anything,* So, no. Where is the fun in that? Not for me. 

I didn't know people uses AI to write comments, like 🙄? Why would you use AI to write a comment, what happened to your brain. As much as I like what AI is doing but I think there's a limit to everything 

> What do you think?

Well, I think, it takes more time comment using AI, and someone would only do that, if they are doing a lot of comments, to farm rewards. And as you said, normal human comments would not be always this big, so if someone is always leaving a big comment, then there is something to suspect.

Oh man, this has been such a big annoyance to me reading comments. Sometimes when I don't know the person, or doubt a little (often because I don't know their writing style), I answer anyway without an upvote or just a short message.. But some lay it on there SO thick, that they even copy these AI layouts that you get when you ask ChatGTP for some topic ideas to write about. Including the emojis lol.

I've had my suspicions about some long term members as well reading their comments... I prefer that someone just doesn't say anything and not engage over leaving me a comment that's AI written without any true engagement in the end.

AI comments often use fancy words in between and defo leaves out the spelling errors. It just feels off, if you're around long enough here, you can spot them quite easily imo.

Anyway, let's hope it's not going to be the norm ;)

Have a great weekend over there!

Well, there are different approaches to how we perceive things. I guess this post came alive from our conversation, and I like how you were plain and open about it. It is always good to feel connected with the person you are talking to. Nobody wants to talk to a bot. We humans have feelings, and it will surely run through what we write or do. The fact that we are human makes us prone to error. In writing, we often overlook some grammatical errors because we are often in a hurry to comment or post. However, even when we are not in haste, there are still typos errors in our write-up. There are AI that can just do that for us. Like myself, I used voice recording and often used grammatical to correct to make it readable.