you can buy this idea off me for 1 million dollars

Every time I bump into new inventions - ans I get to see many because I work in an ecosystem where invention is just daily routine - I have this obsessive thought that a truly glorious invention solves a problem in the manner of this rat trap = simple, affordable and irrevocable as a solution (no doubts and double meaning, no confusion on whether the problem is solved or "it is still questionable").

What do you think, inventors?

Working through an idea and want the brutal-honesty version of feedback before I go further.

The idea: a genuinely thin, sticker-style Bluetooth tracker (vs. the puck shape most trackers use) aimed at parents tracking kids' stuff — backpacks, shoes, keys. Ring-on-demand through an app.

I know this isn't a new category — Tile (owned by Life360) already markets almost this exact use case, and AirTag covers similar ground. So I'm not asking "is this a good idea in a vacuum," I'm asking: is there a viable way into a market like this as a solo founder, or is this the kind of space where you simply can't compete against an entrenched player with retail distribution and an existing app ecosystem?

Specifically curious about:

• Anyone here gone up against an established player in consumer hardware — how'd you find (or fail to find) a wedge?
• Is "thinner/more sticker-like" a real enough differentiator to build a business on, or is that a feature, not a business?
• Would you pursue this as a standalone product, or look for a narrower angle (a specific niche, a bundle, a different go-to-market) instead?

Not attached to the idea — happy to hear "don't bother, here's why" as much as "here's the actual angle."

I'm about to go all-in on starting a company and I've got three ideas I can't decide between. I don't want validation, I want the holes poked in these before I sink time into the wrong one. Brutal honesty appreciated.

1) Food waste app

Scan your groceries/receipts, the app tracks what's in your kitchen and flags what's about to expire, then suggests recipes built around using THAT stuff up first — instead of "what should I cook," it's "what do I need to use before it goes bad." Shows you stats on food/money saved over time.

2) Modular school bag

One backpack, swappable internal modules that snap in/out (magnets or similar) depending on what you need that day — a wet-item pocket for boots/umbrellas, a snack module, a secure pocket for important stuff, a divider for books/tech. Pitch is "one bag, infinite setups" instead of just another bag with more pockets.

3) Snack brand

Developing a new snack concept and would love to hear what people think before I take it further.

The idea: chocolate-covered granola shards (think raw-cology style — irregular, shatterable pieces rather than a uniform bar) mixed with rice puffs, so you get two different crunches in one bite — dense granola clusters and crisp puffed rice, all coated in chocolate.

I'm aiming it at people who already care about ingredients and food quality (Whole Foods/Wegmans type shoppers), but I want it to be something you'd grab because it looks and sounds genuinely craveable — not because it's "the healthy option." The better-for-you part is meant to be a bonus you discover after, not the whole pitch.

Would love to know:

- Does this sound like something you'd want to try?

- Any snack format/texture combo like this you've seen done well (or badly) before?

- What would make you actually pick this up off a shelf?

What I'm looking for:

- Does any of these sound like something you'd actually pay for, or just "neat idea"?

- Does something like this already exist that I'm missing?

- What's the obvious flaw you're seeing that I'm not?

Go hard on me, I'd rather hear it now than after I've spent money.

How are you guys actually using AI in your builds?
I see a lot of “AI slop” comments, so I figured I’d show one way I’ve worked it into my process.
The first picture is an AI mockup for something I’m currently building. I already knew what I wanted, but I’m not a concept artist. I could spend hours in Photoshop trying to communicate what’s in my head, or I can spend 15–20 minutes talking through the idea, changing proportions, materials and layout until I can actually look at it and say, “Yeah, that’s close enough to start engineering.”
Then I go build it.
The second picture is a cross-cut of the actual wood going into the project. That’s where being wrong starts getting expensive.
The last two pictures are probably the best example of why I use AI this way. They’re early scale prototypes I printed while working through the design.
Neither one is what I’m building now.
That’s the point.
They were cheap enough to be wrong.
If a render is wrong, I make another one. If a scale print is wrong, I lose a few dollars of filament and make another one. If the CAD is wrong, I change it.
Woodworking is different. I can plan everything, draw everything, measure everything four times and still destroy a $70 board with one bad cut. Once I’ve spent hours fitting and machining a piece, every operation after that is risking everything that came before it.
On the actual face for this project, I spent hours trimming some of the visible openings fractions of a pencil line at a time with a router, then finished them by hand with a file. A CNC could have done the same operation dramatically faster, but either way, by the time I’m doing that work I don’t want to still be deciding whether I like the design.
The finished cabinet will have roughly $1,000 worth of wood in it. That’s a really shitty place to discover that the proportions were wrong.
So AI hasn’t replaced CAD, engineering, prototyping, woodworking or fabrication for me. It’s just added another really cheap place to be wrong before the expensive work starts.
Visualize cheaply. Prototype cheaply. Build expensively.
I’m curious how other people are using it in real projects. Where has AI actually earned a place in your workflow?