ed

> To prevent abuse we ask you to sign in before chatting with AI.

I want to try this, but, likewise, "to prevent abuse of my Github I ask for some idea what the service does before signing in"

This is not a new face swapping technique, it’s a wrapper around inswapper (aka InstantID, an IP adapter): https://github.com/haofanwang/inswapper

Relevant source https://github.com/hacksider/Deep-Live-Cam/blob/main/modules...

You chose an insult over encouragement. Why do that? Be kinder

Ah yes, the easily memorable room-genius-xi.vercel.app ;)

Any thoughts on how to get distribution? These products are kind of commodity services at this point (assuming you’re doing a straight line controlnet) so marketing is like 95% of the product. Best of luck!

Since these are also really inexpensive to generate (15 seconds of GPU time ≈ 1/4 of a cent), there’s probably some room for innovation on the product side around that?

Grounded SAM[1] is extremely useful for segmenting novel classes. The model is larger and not as accurate as specialized models (e.g. any YOLO segmenter), but it's extremely useful for prototyping ideas in ComfyUI. Very excited to try SAM2.

[1] - https://github.com/IDEA-Research/Grounded-Segment-Anything

Grounded SAM has become an essential tool in my toolbox (for others: it lets you mask any image using a text prompt, only). HUGE thank you to the team at Meta, I can't wait to try SAM2!

> dozens of wheelchair-bound passengers (who get up and run off the plane at the destination)

That is definitely not a normal (or even believable) occurrence

That’s actually right (cost per FTE), according to the article:

> The company now employs about 1,500 people, which could cost $1.5 billion as it continues to grow

Yes, in progress (in mice) https://www.levf.org/projects/robust-mouse-rejuvenation-stud...

They're in roughly the same class but totally different architectures

Deepseek uses a 4k sliding window compared to Codestral Mamba's 256k+ tokens

iOS has a built-in app called “magnifier” which might be more handy than a lens!

First time seeing Cerambot - cool!

I was taught that compression of clay is a big factor in ceramic strength.

I’m curious, how fragile are the pieces? Will a cerambot mug survive a small (eg 4”) drop?

Man, commenters here are... grumpy. Yes it’s true you can do diffusion on a pi, if you’re okay waiting 4 years for your 50 steps to complete lol.

I want something like this to work, and am sure someone will eventually make the “photoshop of stable diffusion.” Personally my generations use custom fine-tunes and syncing 7gb checkpoints to a remote server can be a pain.

The bigger issue for Rubberband, I think, is that Civit weights are the killer app, and all Civit needs to do to win is make their generator a little more visible. Same with Civit’s trainer. They have an easy to use pipeline to go from training images -> generations, and they seem to run it close to cost, but it doesn’t even appear in the navigation until you sign in.

That doesn’t mean they’ll win, of course. Plenty of companies screw up, and maybe it’s not worth “winning” to begin with. See also: HuggingFace, which has Spaces, but is mostly used for the free bandwidth.

Anyway, best of luck!

> While Internet followers may have discovered the location of the old signal, the purpose of UVB-76 remains a mystery.

Betteridge strikes agains.

For an ad-free and concise explanation see https://en.m.wikipedia.org/wiki/UVB-76

If the service opportunity is to resell AI skill to the people who lack it, consider productizing and selling Education instead. Otherwise the clock is ticking for a new model to put you out of business. Most folks will need to learn “AI for ___” eventually, and product businesses are way better to own. Service businesses are a lot like being employed. (Having run both.)

This is specifically about web browsing agents, which is a pretty specialized subset of “AI agents.”

Congrats and I am sure a lot of work went into this.

That said, here's some stranger-on-the-internet feedback: I don't know what your product does and the landing page lacks fit and finish.

Is this a form builder that saves responses to notion? Or it's just inspired by notion? (If I say I am uber for kitchens, my service has nothing to do with cars.)

Your video needs to be storyboarded and could use some title cards. It probably needs to be split up into multiple videos. You can cheaply hire an editor to help with this. Landing pages are a lot of work, almost as much as the product itself. This is hard to accept if you're an engineer.

"Create beautiful forms and surveys with NdbForms" – this hero copy needs to be really, really tight. Currently it sounds like just another form builder. I assume you do something different. Don't repeat the name of your product since I'm already on your landing page. A human copywriter will knock your socks off and they're even cheaper now that everyone uses AI for this stuff.

Consider some user testing. Get on zoom with a potential customer and send them to your landing page. Just listen as they try to figure it out. You'll learn a lot.

Does notion have an app store? the big advantage of building an integration with an existing platform is they help to get you distribution. when i google "form builder for notion" i get a lot of results. it's a crowded space. if you can't easily break into the top 3 for your niche you're facing an uphill battle. If you're a solo dev, it's not possible to compete by building a better product. It's much better to "win by default" by being early to a less-competitive niche.

the best feedback hurts the most! i hope this hurts!

This is done by your email client, not the sender. Here are Gmail's rules.

Emails are grouped if each message meets the following:

- The same recipients, senders, or subject as a previous message

- A reference header with the same IDs as a previous message

- Sent within one week of a previous message

https://support.google.com/mail/answer/5900?hl=en&co=GENIE.P...

Personally I try to finish a feature, but leave QA for the morning. This way my first task is “copy and paste this command into terminal” or “use X feature in the app.” Because there’s always something that needs tweaking before it’s really done, this naturally pulls me into flow the next day. It’s a great trick!

this feels apocryphal and appears to be one theory of many[1] (also - just take out an extra penny during the next transaction). to quote cecil adams "trying to find out who invented [psychological pricing] is like trying to find out who invented the hat."[2]

1 - https://en.wikipedia.org/wiki/Psychological_pricing

2 - https://www.straightdope.com/21341889/why-do-prices-end-in-9...

1% of 8 billion = 80 million developers

80 million developers * 100 websites * $20 per month * 12 months

= $1.9 trillion revenue per year

Not a bad idea! Congrats on the launch.

One of my first projects (about 20 years ago) was a free photo uploader like Imgur but before Imgur.

I was so concerned with “profitability” that I watermarked every image people uploaded. People fucking hated it. What I learned is don’t be so commercial when you’re starting out. Services usually cost a lot less to run than you imagine.

Yes you’ll get a burst of traffic when eg posting to HN, but it dies down after 2 days, then you’re back to 0 traffic (and no costs) or, you have traffic, and non-zero costs, and can tell your users “I’m going broke, time to pay”, and they’ll be mostly sympathetic and understanding.

I have these! They really do glow in the dark — but it’s much easier to capture in a photo than see with the naked eye. (Dim) “green tinted moonlight” is a good description of it. Or imagine a barely charged glow in the dark object. Still, cool technology!

Agents and tooling around LLM’s can probably make some small number of applications viable, but for the most part we need better foundation models to deliver on the hype.

We’re definitely in the “wait” phase of the wait calculation. Everyone is expecting GPT5/q* to change things but really we won’t know until we see it.

Hey, you're that guy! Purchased, I'll give it a look.

The Mom Test in easily digestible podcast form: https://podcasts.apple.com/us/podcast/indie-hackers/id120616...

Start Small, Stay Small has some good ideas but is out of date.

X is where I find the most helpful and relevant SaaS tips.

It's Sora. https://openai.com/blog/sora-first-impressions

A lot more social good with this one though. Hard to cheer for Stability.ai’s failure.

You’ve now posted 4 meta comments about comments. If you have something substantive to say about the study, say it.

about 7 weeks from idea to launch –

Idea: jan 6th, after a long road trip listening to danny postma and pieter levels interviews

I bought a cheap but relevant domain, put up an AI-generated landing page and bought some ad traffic to make sure I could get a decent CTR and advertise against the keywords without issue, and to confirm ahrefs is correct about it being a high-volume phrase. (Google is particular about health-related keywords.)

that worked, got a CTR of 10% and good volume, so i paid $1k to acquire the "real" domain, hired 5 models from Craigslist to experiment with different training techniques and to generate real examples for the landing page, spent about a month automating a pipeline for training, then, for the landing page, paid $400 for a human copywriter (which I really, really recommend – it turned out so much better than AI copy), $1500 for the design and implementation on Webflow, and a few hundred for misc expenses.

For implementation, signup, payment, image upload and order submission is all implemented. Using phoenix liveview which I love. But I am waiting for actual purchases to connect the training pipeline to the webserver.

launch: feb 27th with paid traffic via google, instagram and reddit (all performing poorly!)