I think what's happening is Chines models genuinely try to benchmax on it and hence they suffer when a new version comes out.

A few things that stand out here - you, as a human, cannot post or update your content. You need to tell your agent to do it. It's designed like this on purpose because that's where the industry is going.

Although it's named SlopIt a bit tongue-in-cheek, obviously not trying to encourage actual slop here - I think if done the right way, LLM-assisted writing could be as good as human writing (and will only get better over time).

Some of the example blogs I've setup for my projects: - https://blog.rafikigo.com/why-we-started-rafikigo/ (Blog for a Kenyan tour company) - https://blog.simbastack.com/ (Blog for my consulting firm)

I'd love for you guys to try it out and provide feedback.

--

Getting started is super simple - just copy the prompt (shown in the Hero) and drop it into your agent and you have a new blog ready.

I do recommend spending some time and effort in using Skills and projects like Humanizer (https://github.com/blader/humanizer) to make your content more readable for humans.

Oh and the core is open source (MIT) - https://github.com/Simbastack-hq/SlopIt

Comments URL: https://news.ycombinator.com/item?id=50261080

Points: 2

# Comments: 0

I'm not a content creator (yet), I'm a developer and I build tools / agentic systems.

I was very impressed with Opus 5.5 overall, with all the animation, motion graphics and everything. But I did not realize connecting it to the DaVinci Resolve MCP can create complete edited YouTube videos just by me dropping some raw files into a folder!

It's not completely zero effort - Claude gives me options of different takes that I should pick (it figures out which options to show). But it's crazy how good it's getting to.

I thought it was just me talking to Claude and everything just gets done by the MCP. But the MCP only lets Claude run scripts inside Resolve, everything else (transcribing, picking the cuts, the graphics, the music) is stuff Claude built up over my last three videos. That’s what’s in the repo. My first video took a bit longer, the last one (a baboon reel) took 57 minutes.

Just a heads up, I’m sure many of you will find the videos a bit shitty, and I totally understand that - I'm still new at it. But without these tools I would probably procrastinate and never post videos on YouTube. So these mediocre videos I can make with the help of Claude are much better than me trying to perfect it by hand and never posting.

Thought this might be helpful for others who are considering posting things as well. Let me know your experience if anyone else has tried similar AI editing tools as well

Write-up with details: https://blog.simbastack.com/claude-edited-my-youtube-videos-...

The video: https://youtu.be/3mlRiF-LeMc

Comments URL: https://news.ycombinator.com/item?id=50003959

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=49711486

Points: 9

# Comments: 10

Comments URL: https://news.ycombinator.com/item?id=49629451

Points: 4

# Comments: 1

MIT License - https://github.com/Simbastack-hq/sentinel

Blog announcement - https://blog.simbastack.com/announcing-sentinel/

PS - Resubmitting this ShowHN since when I posted this yesterday, I completely forgot to make the repo visibility public.

Comments URL: https://news.ycombinator.com/item?id=48952249

Points: 1

# Comments: 0

Have made it public now

Comments URL: https://news.ycombinator.com/item?id=48935984

Points: 16

# Comments: 6

Comments URL: https://news.ycombinator.com/item?id=48832341

Points: 1

# Comments: 1

https://news.ycombinator.com/item?id=48222733 https://blog.simbastack.com/indexed-a-year-of-video-locally/

I wasn't familiar with your project though, interesting stuff.

I'm trying to add more photography related features to Framedex but yeah there's so much we can do locally, exciting times.

Are you planning to open source it? Or maintain it in the future?

But yeah, how it picks the frame is the weak-point here. Scene detection would definitely help - this is #1 on the Roadmap.

Could you share how your scene-detection picks the frames?

---

For the vector search, I went for the trade-off of not having it but keeping it simple with plain Markdown files for more portability. The knowledge travels with the files when an SSD moves, no index to keep in sync, and plain text that outlives the tool. But the other path you mentioned is interesting as well to explore.

But I've mentioned elsewhere - if it wasn't for all the AI-assistance, I would've put-off documenting everything that I did and not even get to the writing part.

But yeah, I'll be working on the workflow to make the next write-up better, more humanized.

You can make AI-generated content without it being slop. Slop, to me at least, is content that's wrong, padded, or generic.

I see the cadence / short-sentence issues but if there's something else beyond those, I'd actually want to know what made it feel bad.

I would've put off documenting what I did over the weekend but instead, I did document everything, spent quite some time (several iterations) and effort to make sure it does not hallucinate and writes in my own tone and voice. I'm sure it could be better but the content is not made-up.

At a time where most of us software engineers have changed our workflows to let AI write 80+% of our code using agents, I feel writing is heading the same way. It then becomes a matter of taste, whether it's done well or not.

If you're looking clues and signs for whether a content has used AI, you're going to be disappointed over the next 12 months.

If it feels jarring right now, I'll work harder on the workflow so it feels more natural next time (someone shared this project with me - https://github.com/blader/humanizer).

But this clearly allows me to make content which I wouldn't have done earlier.

Although slightly laggy, I was impressed by the fact that I was still able to work on other things and have a bunch of tabs open on my Brave browser.

You know what I REALLY want? Just point this beast at the folders and it tell me which 150 shots are good to process from these 1,500 images. That's the dream!

Although the technology is getting there, it's still a very difficult problem to solve. Taste and art is subjective. Also me as a photographer will always be concerned - "what if my best shot was in one of these rejected shots".

But yeah, I think I'll try to do some more of these experiments soon.