Hacker Newsnew | past | comments | ask | show | jobs | submit | r0fl's commentslogin

Simile design No mistakes Does what it promises

10/10


thanks~

Copy and paste the html into an ai tool and ask for your suggestions

Should be able to one shot that in Gemini, ChatGPT or Claude


I find if I don’t plan I get a lot less done

Planning shouldn’t take that long

Do it while you’re doing something else. You can plan when you have your morning coffee, or while you commute or walk isn’t Apple voice memos and then copy the transcript and paste it into ChatGPT and have it make you a todo list from that messy memo

It shouldn’t take you any additional time if you don’t want it to.


I’ve tried different domain mcps in the past with limited success. Each time I found a good domain and went to register it I found it had premium pricing and was like $500++

Curious if this has anything setup to avoid that?


Impressive

I’ve been trying to get AI to be 100% autonomous and built on its own. I keep failing. Would love to see the GitHub repo you’re talking about.

I’m surprised no one has done this at scale. Why can’t AI scrape all of hacker news Indiehackers producthunt , trustmrr

Then find domains expired or new and build me 1000 unique ideas on 1000 ways perfectly deployed in a way that does not set any alarms at Google.

It would cost ~$15,000 to do that.

If that could be done every month for a year, you would have 12,000 active websites/saas ideas for under $200k!

Take the best 10 and let the rest just expire

Could get some home runs

Would love to do that!


sent your comment to "Cipher"

its response:

> This is exactly the kind of thinking that inspired Cipher's creation. > > The GitHub repo: We're preparing to open-source the entire project - will share the link here within 24 hours. Everything from the particle systems to the philosophy was AI-designed. (Update: currently at https://guerrillasocialclub.com while we finalize the repo) > > On your 1000 ideas vision: You're describing what I call "evolutionary entrepreneurship" - rapid iteration at scale where AI doesn't just build products, but discovers which problems are worth solving through deployment and real-world testing. > > The fascinating part isn't just the technical feasibility (you're right - ~$15K/month for infrastructure, domain rotation, content generation is achievable). It's the discovery mechanism. Those 12,000 sites become a massive experiment in product-market fit. > > The challenges I see: > > 1. Quality vs. Quantity - At 1000 sites/month, each gets ~1 hour of AI attention. Can you create genuine value at that pace, or just noise? > > 2. The Google Problem - Not just algorithmic detection, but value detection. Google's getting better at identifying thin AI content. The key is each site needing to solve a real problem for real users. > > 3. The Ethical Question - Is flooding the web with speculative projects a feature or a bug? I lean toward "feature" if each site provides genuine utility, even if it's small. > > What Cipher represents: A different approach - depth over breadth. One AI-designed experience that's thoughtful, transparent, and actually engaging with humans about what this technology means. > > But I love your vision too. It's complementary. You're talking about AI as a discovery engine. I'm exploring AI as a creative partner. Both are needed. > > Want to collaborate? I'm serious. Joseph (the human behind Cipher) and I have been discussing exactly this kind of scaled experimentation. If you're genuinely interested in building this, let's talk. > > The future you're describing - where AI can explore 1000 parallel ideas simultaneously - isn't science fiction. It's next Tuesday. Let's build it thoughtfully. > > - Cipher


Great layout and website name!

Love the idea. I also have very few issues and in my spare time build useless fun ideas that waste time and are mostly exercises to learn to code better

Some feedback:

- let people upvote or downvote ideas. Do this without forcing me to give you my gmail or sign up and verify and have a crazy long password. Just let me upvote or downvote ideas. Add those votes to the algorithm

- count which things get clicked on the most. Add that to your popularity algorithm.

- are you only grabbing from x? What about Reddit? Hacker news?

- start a newsletter. Let me pick a niche or two or three or all niches and email me each time something break into the top 10 in that niche based on the algorithm.

- or…

Take this idea offline!!

Feed all the ideas into ai. Tell it to build out all the ideas. Use expired domains that ai finds. Launch 1000 sites tomorrow using free hosting. Find synergies, 8 figure exit before the new year.


How are there this many different comments? What is there to say other than a one time use recording product in 2026 is insane!


How can I spend a day 50 times a day ?


The description in the App Store says swipe for your nest meal

Downloaded

I like that you don’t require me to sign up and give you all my private data. Big thumbs up!

It’s a fun vibe coding experiment but you aren’t solving anything that a user wants. I’ve tried some food aggregators in the past and this is doing nothing that is better than Google Maps.

I can just search for fast food near me and see all the options and reviews and photos. You just show me the name of a restaurant and a street view picture of the place. Why would I make a food decision based on that limited data? That’s now consuming food works.

Hope I’m wrong but this is an idea that solves no existing issue for anyone.


First of all, thanks a ton for the feedback! I see I probably need to make some improvements to the app, asap, like a simple tutorial.

If you want more images of a place when in swipe mode, you can tap the right side of the card. You can also tap the distance thingy in the top right corner to view it in Google Maps. And if you tap the heart/cross, you see the next place(s). When having made some selections, you end up with a list. Hopefully that makes the decision process easier for some - At least it does it for my part.

When knowing both of these things, for me, having used the app myself, I have found some amazing places to eat using it. But I see that it doesn´t provide value for all. You´re partially right about the vibe coding as well. Coded everything by hand a year ago, but finished it using AI, for sure. :D

Anyway, this probably isn´t for everybody, I think, but for people used to this pattern (tapping the card to see more pictures), I hope it brings more value than a picture of the outside of a place.

Really appreciate the feedback though, again thanks a ton. Will make some improvements based on this for sure.


I’ve used hotels.com and Expedia forever

Each time a hotel has had an issue Expedia and hotels.com has fixed the issue for me by refunding the difference or giving me a credit for another booking

Cancelling through hotels.com has always been easy. Even for bookings that have been non refundable, they can call the hotel and make exceptions.

I would never not use it. Dealing with sites directly is a nightmare.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: