Startup idea list

161 points by duck ↗ HN
Five years ago today this was the top item on HN (http://news.ycombinator.com/item?id=55974): Post novel startup ideas as comments and see how many karma points it receives. Who knows, one might be developed and you can take credit for thinking of the idea first.. It had some interesting ideas, some of which are solved or obsolete. Thought it might be fun to do this again.

(I run Wayback Letter, so that is how I know this odd bit of information - http://www.waybackletter.com/archive/daily/09-18-2012.html).

274 comments

[ 1.9 ms ] story [ 237 ms ] thread
Here's a few I've been thinking about:

- Remote aid delivery by drone: build and send small robotic aircraft to remote areas and drop aid, preferably starting with medicine (lightweight and very useful). Small helicopters (quadrotor drone copters as you see at TED/MIT) could land and accept return packages if necessary. Trips could be made to hazardous places where human-driven aid drops can't venture. Additionally, all aid drops could be done cheaper and more efficiently. [The drones should be solar-powered?]

- Microwave oven with a solid, opaque door. On the door put a nice 1020p screen and inside put video cameras. You could microwave food, see a convincing video of it cook and not have to deal with harmful radiation or an unpleasant grilled screen.

- Pizza button. I had this idea about two years ago, someone in Dubai is currently executing (I think). You order a button, a physical red button (or an app with a big button, but that's not as cool). If you press the button, a pizza arrives at your location (tracked either by smartphone or by pre-set address). Default pizza arrives unless other instructions are set. I would insist on extreme simplicity: no options, no choices. Just "press button to receive pizza".

The pizza button info is here, but they're out of stock: http://www.redtomato.biz/magnet/
no idea why out of stock. OK -- take this idea to next level. Why dont develop a wifi ready button (everyone that buys the button need to pair it with their wifi) - probably cost of <$5,000 (kickstarter?) and have a simple web page like your router admin has that you can choose different pizza vendors, then work directly with pizzerias in the hood to offer them additional channel of sale! that would be my take.
You mean, like a computer instead of a button?
The radiation from your microwave isn't harmful, but I will admit it's kinda hard to see what's going on through that mesh.

I really like your drone delivery idea.

i too like drone delivery idea
I could see myself eating pizza way too much with that button. I hate cooking and the press of a button for food would be just too tempting... Just what our society needs.
Regarding idea no. 2. You realize that putting electronics inside a microwave isn't as simple as that, right? http://i.imgur.com/OFEg9.gif
We have electronics on the surface of Mars, we certainly can fit one inside a microwave.
I was wondering how one would transmit image/video feed from inside to outside in realtime? Any hole to transmit the feed is gonna be a no-no(radiation leakage). Basically, if there is path for some EM communication from inside to outside, then MW radiation will also leak along that path(unless the path has some sort of band pass characteristics). You could do some kind of time slotted transmission(In one slot, close the communication hole to make a perfect metal box, turn on microwave. In next slot, turn off microwave, open hole and transmit video feed. Make the time slot ~msec for near real time feed).

With some kick ass marketing it should be possible to get the message to people - 1000x reduction in radiation!(?) why take the remote chance of effects of radiation!!

Interesting....

I was wondering how one would transmit image/video feed from inside to outside in realtime? Any hole to transmit the feed is gonna be a no-no(radiation leakage). Basically, if there is path for some EM communication from inside to outside, then MW radiation will also leak along that path(unless the path has some sort of band pass characteristics). You could do some kind of time slotted transmission(In one slot, close the communication hole to make a perfect metal box, turn on microwave. In next slot, turn off microwave, open hole and transmit video feed. Make the time slot ~msec for near real time feed).

With some kick ass marketing it should be possible to get the message to people - 1000x reduction in radiation!(?) why take the remote chance of effects of radiation!!

Interesting....

I think you're looking for either 720p or 1080p.
Google around for "tacocopter." It's not a real product, more of a hack of marketing droids, but you will find discussion of the benefits/problems it has.
My low-rent version of the drone delivery idea is the burrito cannon.

You order burritos via your phone. When they're ready, your phone tells you to go stand outside. Once there's a solid GPS fix, a precise, high-powered catapult throws the delivery package way up in the air. Like a smart bomb, it uses tailfin guidance to home in on your phone, popping a 'chute at the end.

What's mind blowing is that when you break this down, it's more or less achievable, minus regulation snafus regarding launching projectiles in a populated area.
What's mind blowing is that when you break this down, it's more or less achievable, minus regulation snafus regarding launching projectiles in a populated area.
One of my more outlandish fantasies is a point-to-point transit system built out of enormous catapults & nets that operates similarly...
That coincides with my vision of people traveling around the city in the air in clear spheres arcing gracefully across the sky...
The pizza button is a fantastic idea as an iPhone app. Show a huge pizza icon, nothing else, press it and get your pizza delivered to your home and discounted from your credit card.

As a developer, here are the interesting bits:

  - Press many times get as many pizzas
  - Configurable default pizza
  - Configurable default credit card
  - Configurable address
  - Keep it as simple as possible
Press pizza, send order to papa johns, get back confirmation and estimated delivery time.

Love it!

Haven't thought this through very much, but it'd address some frustrations I've been having lately. If someone can come up with a means of getting around domain squatters, that'd be amazing. Kind of like a redirect service for legitimate businesses.
I've been learning about the domain squatting business recently, by basically becoming a domain squatter. It's something I've been fascinated by for years, I always wondered how profitable or easy it was.

I've learnt quite a bit, and will get round to blogging things sometime, but the main lesson is - most domains earn nothing from adverts while on sedo or other marketplaces, and a decent offer of $500+ should get you almost any domain you want.

If it's a domain that's worth much more than that (e.g. insure.com sold for $16 million), then it would have been bought long ago regardless of domain squatters. It would just have been used by some other business, so you didn't really lose out.

Would be really interested in a blog about this. Please drop me a mail if you ever get around to posting one?
Will do - maybe this thread will make me get round to doing it!
If someone has a domain of my name that's very low value (I'd a domain I used to use for a blog). Is it likely they'll let it go after a year with little traffic? If not how much would you expect them to want? How do you get in touch if they've used an anonymous registrar?
How long have they owned it? There's lots of people (it seems everyone on HN for a start), who buy domains "for a project" that never quite gets started, and end up owning 20-30 domains.

In general, I think those kind of domains will sell for anything around $200. If they're holding it for a sale, they'll probably want more, like $500 or more.

If they've used an anonymous registrar, you can check the domainwhois sites for a history if you really want, and contact old owners of the domains, it might still be them, but otherwise I'd say forget it - they probably don't want to sell.

Interesting thanks. I'm the only previous owner, I let it lapse carelessly. As my the domain is my name, and a pretty unique one, it doesn't have much resale value.

My guess is that when I was blogging regularly it got enough traffic to get on some watchlist and when it became available was snapped up. But don't think it will see any traffic without me pushing it.

But I suspect you're right they don't make it easy to contact them so I don't imagine it's up for sale.

If you email me (check my profile for address) the domain, i'll have a poke around if you want, but it was probably an automated system that grabbed it, and if it doesn't generate enough traffic to pay for it's renewal, it'll get dropped next expiry.
I think we should have some kind of domain amnesty.

Every domain you are not going to develop & probably wont renew next year should be easily available. Parked domains should be forwarded to a hosted domain page saying "If you'd like to buy this awesome domain contact me on ..." "I'll accept [insert reasonable price ie $100]

Domains with no contact info, that are parked & have never been developed are so annoying.

Either develop them or have them forward to a decent sales page.

A system that lest me maintain my contact details in one place, that other services can subscribe to for up-to-date info. Should also let me maintain who can see which details. When I change job, or lose my phone, or whatever, I can update my details, and my vCard in all my friends' address books gets updated accordingly.
Love this, and have been thinking about it for a while (after having moved literally 10 times in 10 years). I picture something similar to OAuth permissions, where at any point I can revoke access to all or specific pieces of information to anyone.

As an extension (and this is potentially where the money is), I can think of an intermediary service (a-la Google Voice call widgets) that allows entities to contact me without ever giving away my phone/address/etc.

Absolutely. Currently I have zero control over who has what contact details. I have good friends that still email me on an eight year old gmail address that I don't use. I need to be able to pull addresses, push new ones, have certain details only available to certain friends and maybe, like you say, have ways of letting people get in touch with me once without actually giving out my details.
This maybe more a failing of their design and messaging, but I honestly can't tell if this system does what I'm describing. I _think_ it doesn't.
This will happen once we have a federated social network with individuals maintaining their own little slice of the Internet. Please happen soon rather than later!
Trade your dead tree versions of books in for ebook versions + credit for more ebooks when value of book exceeds ebook wholesale cost.
Don't think this can be profitable in any way. What would you do with the dead tree books?
Ship in bulk to poor countries for use as bricks. There's clearly no other profit to be had for them.
A service which allows me to upload a document (PDF, .doc, whatever), input an address and then pay for it to be printed and sent to that address via mail (courier optional).

I'd use this service so much:

1. Printers are fast becoming a thing of the past. How often will people need to use printers in five years time?

2. Lots of government functions still require printed documents (heck, even Twitter insisted that I fax them something a few years back).

3. Post boxes, stamps and envelopes are a pain to get individually.

Apple still requires developers from outside US to fax them form with their credit card number on it. (!)

On the other hand, there are services that will fax PDF for you, even for free.

The difficulty with faxing something is that it's even more out of date than printing something. I don't have a printer, but at least I know a store where I can buy one. I've got no idea where the fuck I'd go to buy a fax machine.
There is such a thing as an e-fax (you send an email and it magically turns into a fax on the other side). I have only seen it from one side and do not have any details, though.

When I was doing Lots Of Paperwork for something, I'd often print out reams at the office and fax reams back. I consider myself rather fortunate that no one ever complained; being friends with the receptionist (who sits next to the fax and was thus ideally suited to calling me out on excessive usage of office equipment for personal use) probably helped.

I've used e-faxes before, but isn't it predicated by the other side (the tax office, or whoever) having either a fax machine or the inclination to accept an e-fax?
Usually, people who ask for a fax are capable of receiving one.
to fax them form with their credit card number on it. (!)

and expiration date, and cvv (not joking)

Interesting. I'd have thought the few things that needed to be printed would also need some sort of physical signature as well. Do you not find that to be the case? What situations do you find yourself in that would call for such a service?

Would it make sense more for mass physical mailings via an API?

I also thought about having an API for direct mail--we'd print and send (or have printed and sent) whatever marketing letters you wanted sent
I like the idea, except the only time I need to print something is when I need to sign it.
Same applies to me 99% of the time. I sign everything digitally using Preview.app in OS X.

I also like to print and send things when it's an old person's birthday (used Apple's Cards app recently and it was brilliant, although a relative was baffled and believed me to be in Prague since it was posted from there).

Large chains like FedEx could implement this really efficiently since they already have offices all over the world. They can just print off at the nearest location to the target and minimize their expense. Getting a printed document hand delivered in an envelope can have a lot more impact over simply sending an email or fax.
They've been offering this service for several years:

http://www.fedex.com/us/office/online-printing.html

Last I looked, though, they only deliver via FedEx, which really raises the costs.

I wish the postal service did it. They send somebody by my house every day. It'd be awesome if the actually dropped off something I wanted.

I wonder if http://HelloFax.com (and HelloSign.com) is working on this? Seems like that would be a perfect fit for them.
On the same vein, someone could write software allowing individuals or organizations to set up a remote print-and-pickup service.

For example, every campus would hugely benefit from a service where students could just email a document to a specific address (smith.hall.printer@mycollege.com), have 5 cents a page automatically charged to their account, and pick up their assignment on their way to class to hand it in.

Students used to running around residence looking for someone with a printer, or sprinting to the library with a USB stick would be all over the service. Someone just needs to write a simple application that connects the dots and market it to campuses.

----

Reply From: smith.hall.printer@mycollege.com

Your document has been printed and 25 cents charged to your account.

In my campus, you just print to the cloud printer and get it from a printer by approving it near the computer station next to it.
Breezy.com does exactly that for universities. Except, you can count it against your current printing credits too.

(Disclaimer: I work here. This also means I can help get it to your university if you get in touch with me).

I had this exact same idea last week but it'd be a tough one to crack. The devil's in the details of the operation.

You'd have to solve at least the following problems: Print/sorting/packaging/mailing, privacy (most of the stuff you "need" to print is confidential), timeliness (usually when I print I need it NOW, like when closing a sale on a house)

But if you can solve them, or at least start to, I think it'd be a great opportunity.

What a great idea for an API that's connected to an existing app/programming like Evernote/Notability/Office and linked to an established network of "member" vendors like UPS/FedexKinkos/USPS/colleges.
Fedex already provides this service fedex.com/office
I'd pay for this. In fact, I end up paying friends to do that for me, while I'm outside the US & have to send docs (such as tax forms, etc.)
(comment deleted)
I've got a similar idea I've been working on called EasyPost.co. I'm not sure where it's going yet--I know a lot is broken within postage but that's an area I've thought about
How about a GPS car navigation unit that detects if you're going too fast for an upcoming curve and warns you?

It seems like it already has all the data.

Kind of already happening. Many gps units warn you if you're over the speed limit, and limits are set with upcoming curves factored in.
> Kind of already happening. Many gps units warn you if you're over the speed limit, and limits are set with upcoming curves factored in.

Here's a startup idea, but not for the faint of heart:

1. The authorities put a QR code (http://en.wikipedia.org/wiki/QR_code) on all the speed-limit signs, a label that can be read from a great distance, and one specially coded to be unambiguously a speed limit notification.

2. An enterprising startup designs an Android app that (a) reads QR codes through the windshield of a car, (b) compares the speed limit to the GPS-derived velocity, and (c) alerts the driver that he's exceeding the locally posted limit.

3. Drivers plant their Android devices on the dashboard of their car and run the app. The dashboard of a modern car happens to be a great place to get a GPS fix.

This isn't meant to force people to drive the speed limit, but only to let them know when they aren't.

I think this kind of technology is already being placed inside cars themselves. And the QR code is probably harder to scan while moving than recognising the speed limit plate itself.
The problem with this is that if you rely on the system and there is an error in the map (the curve is not marked) you have a big problem. If you don't rely on the data and drive carefully there is no point anymore :) A possible solution would be to detect the curve radius with a camera, but it has it's own problems. You never have 100% accurate data.
Well the market is for careful drivers who once in a while aren't paying attention, or are simply approaching a curve that isn't well marked.

I'd consider it a system of last resort, not something to replace careful driving but to enhance it.

And maybe throw in some modifiers if it's raining (or has recently rained) and if it's dark or foggy.
Here's a simple idea I've been kicking around. I might just build it for fun.

I call it notifyme.when. Users put in things they want to be notified about e.g., new book by Steven King released, Sequel to District 9 announced, Mortgage rates fall below 3.5%, etc.

Then other users log in and review the requests and post updates to any of the event they know have happened. You get points (like Stack Overflow) if you legitimacy update someone else's request.

I like this except for the crowdsourcing part. I'd rather have it done as a paid-for service somehow. I realize that it's hard because you'd have to constantly check, but still.
Sort of like If This Then That but with a wider range of uses, right? Sounds interesting. Also, I don't get the last bit - How would you know if a user is honestly updating the request or not?
I guess it would be a lot like stack overflow. The user who requested the notification would have to confirm the answer is correct.
I thought of something similar for game releases, particularly indie games. Didn't want to have to trawl through a bunch of sites to find updates for the.

Also, there was something posted a while back that did music albums, but I think the guy was shutting it down/open sourcing it.

Either way, glad to see I'm not alone.

This is actually a compelling idea—I've been kicking around a similar idea for a while, where you'd like to buy something (e.g. a 2TB SATA drive) but only when it goes on sale below a certain price threshold. The problem is: what's the business model (i.e. where does the money come from)?
Have you checked out Priceonomics? They're a YC startup that gives you analytics on the market price of a product and send you notifications when the item is available for a good price.
I have not, but I am so not surprised that someone else had the idea, probably made it better, and is in business.
I would say the whole thing is a monetizable business. To have good data, you'd need a mix of human and tech input. So why not charge $x for y number of updates? or charge for specific searches that requires more human work? or charge for a premium account that shows price fluctuations of certain items for the last 6 months?
You may be looking for decide.com
You may be interested in http://camelcamelcamel.com/ for this sort of thing. Uses Amazon to track historical high and low price points so you can get a good idea of when to expect sales. You can set up a trigger to notify you when it reaches a certain price, too.
I've been thinking about something similar to this - particularly for me in relation to when bands bring out new albums or new TV series come out. I was thinking I'd probably tackle it a vertical at a time and find a source for the information (e.g. last.fm for albums, IMDB for movies/TV series) that can be put to use automatically. I reckon the simplest way to monetize this would be affiliate links in the alerts - the click through is obviously going to be pretty high (people have actively said they're interested in something) but I've never seen affiliates as a great source of revenue.
A github-like service for musicians that would allow:

1. Version control of compositions; 2. Forks of other musicians' "repositories" / compositions; 3. Contributions to other musicians' songs; 4. Possibly play the songs online; 5. Private / public repos; 6. Individual / band / group / organization repos. 7. ???

I'm a terrible musician, but I love this idea.
this is something similar: http://www.ohmstudio.com/
What I have in mind is more focused on the composition part of music creation, rather then recording / editing. But this is a very cool project! And of course collaborative sound editing would be awesome for an idea such as mine.
How would the notation be setup?
I didn't work much in the the idea, but since I'd focus more in hte composition process, I thought of building an easy-to-use, graphic version control tool that integrates with common composition software (finale, guitar pro etc.).

This tool wold then convert the score / tablature into a single representation that would be used for storing, diffs and such. Perhaps this represantation could be text-based (lilypond [1]?) to use a mainstream VCS (git) as backend.

But I don't have the skills or time to do that, so it will stay in stand by until someone (me?) implements it someday.

[1] http://lilypond.org/

(comment deleted)
Ok, I have no idea how to do this startup but it seems very profitable. Have you ever noticed how anyone under 40 believes with 100% certainty that social security won't be around for them when they retire.

Well I think it probably will be around. Most big (popular?) government programs aren't going away no matter what the media says.

So how about selling social security insurance. Buyers pay a premium each month to buy the insurance. And if social security isn't around when they retire the insurance pays out what they would have gotten from social security.

---- Edit ---

People don't think the insurance idea is viable so let's turn this idea on its head. The company goes out and buys the future social security benefits (or some portion thereof) from individuals ands pays them say 10 cents on the dollar. People who don't think SS will be around think they're getting free money, and the company makes a bundle if SS stays around.

Heck, could the US government could be doing this to fix social security?

It would have to be more complicated than that, because it's not just a question of whether SS will "be around," but whether it might still exist but in a degraded form (e.g. lower payments, narrower eligibility).
Insurance can't work if either all or none of your customers are going to file a claim, because the premise of insurance is that those who don't file claims pay for those who do. This insurance would have to charge a premium that would cover the "everyone files a claim" case, which means people would be paying N dollars in order to have a chance of getting their N dollars back. Not a good deal ;-)
Right, unless the insurance company can hedge their bets. At it's simplest they could insure themselves with a re-insurance company.

They could also figure out financial consequences of social security going away and place long term bets on that. If they are able to hedge properly, they're basically reselling that hedge.

Sounds like a pyramid scheme to me.
I... honestly am not qualified to say that this is what it is, but I would strongly suggest taking a look at the details of how we entered this recession. It sounds extremely familiar.
This isn't necessarily the case. Life insurance insures against certainty.

I think one things which would make this idea feasible is the fact that people are old by the time they receive SS so a certain percentage (a somewhat predictable one) will be dead. If a high enough percentage is expected to be dead, you could create rates which would allow you to win over the long run.

You could also make the SS insurance apply only to the individuals and not their beneficiaries who would have received the SS benefit in the event of the individual dying prior to receiving benefits.

I know it's grim but just some thoughts..

Insurance works because the probability, eg of death, is spread out over many buyers; not everyone will die young. If social security goes away, the payout is 100% and your company goes bust.
to offer insurance, you have to be able to pay out if a claim is made. Since the insurance you are offering is all or none ( i.e. if social security goes away, it goes away for everyone ), you would have to pay out claims to everyone at once. This is much different than auto insurance where you only expect some small percentage of policy holders to put in a claim.

You need to find someone to take the other end of the bet. So, who would want to insure themselves in the case where social security stays around? Can you find enough of these people to balance out your risk?

Ok a couple more ideas; sorry to post so many :-(

A turn signal that produces the turn signal ticking sound really loudly outside your car so the person in front of you is aware you want to turn.

Sell sterilized tape worms for weight loss. It seems like a great way to lose weight but I'd be freaked out by the though of them reproducing inside of me. If a tape worm harmlessly died after a few month I'd be up for it, especially if the alternative were surgery or diabetes.

A device that can detect the sound of glass breaking and notify the police. You could put them in neighborhoods everywhere.

> Sell sterilized tape worms for weight loss. It seems like a great way to lose weight but I'd be freaked out by the though of them reproducing inside of me. If a tape worm harmlessly died ...

The tapeworms might not cooperate:

http://gizmodo.com/5910770/brain-tapeworms-are-real-and-they...

http://www.inquisitr.com/240780/brain-tapeworms-are-a-terrif...

I personally don't want tapeworms in my brain -- there's already enough junk living up there.

Number 3: A startup in the same co-working space as me is working on exactly that (amongst other things) – http://www.audioanalytic.com/
Are you in the St. Johns Innovation Centre?
No, ideaSpace, out in West Cambridge – http://www.ideaspace.cam.ac.uk/ – where I work on a product called Unifyo which helps businesses integrate separate information systems, cloud applications and e-mail in a lightweight way (https://unifyo.com). Always happy to chat, e-mail in profile.
> A turn signal that produces the turn signal ticking sound really loudly outside your car so the person in front of you is aware you want to turn.<

All I can imagine is the recording of a guy yelling, "Click! Click! Click!" amplified at a crossing. Ha!

Internet filter for things I don't want to know yet. If I miss the Formula 1, I'd rather not know the result until I watch it. Currently this means not listening to the radio, watching the news or going to any non-geek online news.

I'd like to have something like AdBlock, but it blocks on key-phrases that are F1 related (Lewis Hamilton, Jenson Button etc)

The key-phrases come from a wiki. The censored bits of sites could show adverts instead (to make some money)

The wiki could hold all the options, different sports key-phrases (football, soccor, rugby .... you know, almost anything ... Dr Who, Eastenders (for those of us who iPlayer a lot of things)

HR and Recruitment software for high turnover, low skill companies like call centers.

Implement a number of systems like patio11's appointment reminder, CRM and helpdesk type software for communication, something to handle all paperwork electronically, posting adverts, information gathering etc etc.

Basically, a family friend used to run the HR and recruitment department for a huge call center, she headed a staff of 10+ people that all day advertised for positions, looked through application forms, gave aptitude tests, did phone interviews, arranged face to face interviews, did the paperwork to get them hired, did inductions and put them into training, and thats just hiring and doesnt include liasing with payroll and firing etc.

I told her about that idea, she mentioned it to her boss (basically the guy that looked after the entire call center) and he apprently loved the idea. Problem is, while i could do the techy bit, i'm lacking other things to be able to make it happen so its on the back burner for now.

How about a Mechanical Turk based recruiter company?
If you have a potential client, pre-sell him on it to get funds to supply the other things you're "lacking". Treat it like a free-lance development gig at first and then turn toward a saas model after you've developed it.
Collectively purchase/own items. For instance, baxter[programmable android] came out yesterday/today, and you would like to try it out, but the price point of 22k is too much. A group of people, say 22, can collectively purchase it and rotate the usage of baxter amongst themselves.
Are you familiar with Timeshare (http://en.wikipedia.org/wiki/Timeshare)? It usually refers to property ownership, but can apply to pretty much anything. Most people consider the middlemen in timeshare situations to be scammers, so it would be tough to market.
A few problems: 1. The subject property breaks, who pays? Insurance is clearly necessary, but who will go through the time to process a claim? What if it breaks in transit between owners? 2. Theft. What's to keep an owner from disappearing with the item?
I have a similar concern with this. It's too easy for a scammer to just walk away with an object. Invest a small amount, run away with an object of far greater value.
I'm counting on human honesty. If that can't be trusted then maybe a gps tracker attached to the system.
There are communities who do do this. I wish I had actual references to point out, but stuff like only having one or two lawnmowers for a neighborhood does happen. It's a cultural thing more than a business opportunity.

OTOH, what if you looked at this from the perspective of a dating site? The goal isn't merely to collectively own a thing; it's to find and meet people who you could get to know well enough to trust and participate in collective ownership.

If you don't mind, I have a lot of questions about this.

* The idea is to get x unknown people to pool in for an object, correct? Imo, you're relying a lot on people being honest. What's to stop a person from never passing on the object? Furthermore, what about the time slice for which each person gets to keep an object?

* Where does a company coming in to all of this? What else do they do other than facilitate the transfer of the object?

Sorry if that's too many questions. I like the idea and it seems like something that could be beneficial in theory, but I have my doubts about it working in practice.

In the case of an android, it could just walk to your house when it's your time to play with it!
A service which makes detailed and complete digital records of everything you have, then disposes of the physical originals.

Dystopian version: It's done by a lot of hazmat-suited folks, possibly robots, and you and your family are included in "everything you have".

A service that allows me get all my snail mail by email, I get a virtual address, they scan / OCR etc (and spam filter)
(comment deleted)
I've been looking into this.

Just as a rough idea: what monthly price would you find bearable?

A real media browser in the browser. I think current browsers have horrible interface to the local filesystem and it should be easy to improve.
Browsers typically rely on the operating system to provide file system browsing/navigation. So the onus on the the operating system makers, not the browser makers.
Home 3D Printer for clothing and textiles.

Textile knitting machines have been digital for close to 100 years (old ones used chains with risers that would change the gears of the machines to change the knit). For simple garments, there's a completely automatable step that makes the majority of the piece, but then there is often hand-sewing to assemble the final product (e.g. sewing sleeves into T-shirts, seams into toes of socks, etc…). Trick is developing the robotics to manipulate it so this can be automated. It doesn't have to be fast (like a factory needs) if you're making stuff for yourself. There's a whole ecosystem you could build around downloadable designs, consumables, etc...

Here's an app idea. Show notifications when the device is within a certain distance from the location of a past historical event. ...or past Crime ...or a message another app user has left.
I like this idea- it would be cool when exploring a new city if had a preset tour feature where it would guide you through these past historical events, crimes, etc. depending on which theme you chose.
This smells like an ARG. And even if it isn't, there are plenty of overlapping demographics who'd like it.

I can see all kinds of weird unintended uses for it, too. ("Alice <3 Bob" marked in random places.)

would be a cool bike or car app - your get a notice when you are near a hotspot for bike or auto accidents. Give the user a heads up. Possibly to scrape this data from county GIS sites or police maps.
A service that aggregate, index and archive all the content you produce on the web. Aggregation could be done manually with ifttt.com but the storage part do not seem to exist yet.
Hardware product idea: ~7 inch screen with keyboard for administration of headless servers. It's very annoying to use large monitor and full keyboard when all you need is change line in BIOS or press few buttons when OS stopped to boot.
A service that syncs your resume with companies' Applicant Tracking Systems (ATS).

The problem: you're applying for jobs at three different companies. Each company uses their own Applicant Tracking System (eg. Taleo, Kenexa, PeopleAdmin). That means you get to fill out your resume three different times, which is tedious and error prone. It would be nice to have a service where you input your resume once and then it just automagically syncs your resume to the ATS of the companies you're interested in.

I know that LinkedIn and Indeed have "Apply with LinkedIn/Indeed" to address this pain point. But I can tell you as someone who's working on a job search engine as a side project, there are many, many jobs where the only way to apply is to fill out an ATS. And if you want to apply to multiple jobs you have to go through the same process again and again.

100% agree. I'd love to be a part of this solution.

A centralized point where people can store different versions of resumes, cover letters, and application information. This point also can parse and pre-fill out any application of your choice for any company. It presents the user only with questions it can't answer from the information it has on file. A summary is given to the user for review before submission.

I've applied to several different positions where one resume or one set of application answers didn't fit all. It would have been nice to have a service to keep a record of my various resumes & application information. Additionally, perhaps you have a varied background you might want to provide a resume or application answers that better highlighted the relevant part of your background. It would be nice if the service could be told to utilize one set of application answers over another.

Anyway, just bouncing around the idea. But I like it.

You might be interested in this job posting from the "Who's hiring" thread from August:

http://news.ycombinator.com/item?id=4323765

TalentDrive is essentially working on this. (Note: I am affiliated with the company in that my employer shares the same parent company as TalentDrive.)

Not to mention only answer the "personality" questions a single time... haven't had to deal with those in some time, but I know many companies use them.
A little OT, but I don't understand recruiters that e-mail you on LinkedIn asking you to send them your resume.

I've even told one to print off my LinkedIn profile page as my resume, since it's the most up to date, but they still asked for one.

Crowd-sourced solar data.

Collect solar irradiance data (used to compute solar panel effectiveness) via consumer devices (iphones or andoids that have light sensors).

Combine the collected data into an always updated, hyper local dataset, give (sell?) data to solar development companies and municipalities.

This could greatly reduce time to research feasibility of solar farms in cities.

Here's another app idea. Create a catalog of map data based on years and let a user see what their location looked like in years past. Users could see what stores used to be there, how fast their city has sprawled, and could show their kids what their house looked like before it became a strip mall.
A service that enables you to never have to attach a file to an email again. (Or at least one that teaches my parents how to retain this information.)