Hacker News new | past | comments | ask | show | jobs | submit login
20,000 Startup Ideas (unawaz.github.io)
461 points by overlords on Oct 1, 2019 | hide | past | favorite | 178 comments

You could generate even more startup ideas by running a Markov chain thing on these!

* Check with customers to determine their usefulness.

* Inoculate animals against various diseases, such as internships.

* Review product promotional materials, and physical fitness.

* Monitor computer-controlled test equipment, using germicides and sterilizers.

* Transpose music to alternate keys, or to towers.

* Operate farm machinery, such as espresso machines.

* Document software defects, using hand and horn signals.

* Prepare bank deposits, or performing arts.

* Run errands, such as animal dehorning or castration.

* Forecast staff, equipment, utensils, and silverware.

> Document software defects, using hand and horn signals.

I hope this is the next agile


How do we properly break down our HONKs into subHONKs so they're finished within a WOMP?

This WOMP will never be done in time for the monthly SPOINK

>* Check with customers to determine their usefulness.

Nice, I always say that customers and testers are source of all bugs!

"Demonstrate proper operation of laboratory equipment by inspecting, calibrating, cleaning, oiling, and greasing equipment, and analyze data to evaluate the effects of regulatory limitations on land use projects, such as explaining regulations, answering visitor requests, needs and interests of individuals and groups on basic rules of good nutrition, healthy eating habits and illness prevention and treatment of cardiac or peripheral vascular treatments, such as meteorological, atmospheric, oceanic, paleoclimate, climate, or related structures such as docks, bridges, tunnels, and other specialized material handling equipment such as endoscopes, closed circuit television, still photography, and testing devices, and repair assistive devices or make adaptive changes to equipment or environments."

Seems like this one is destined to fail due to lack of focus.

I generated these with https://github.com/pteichman/cobe - anyone got other markov chain generators to play with? This one seems pretty basic.

Wait, OP's list wasn't generated with a Markov chain?

Some of these could actually be legit:

> Forecast staff, equipment, utensils, and silverware.

Sounds like something a catering business would need.

> Run errands, such as animal dehorning or castration.

> Operate farm machinery, such as espresso machines.

Both of these sound like things a ranch would need (who says a cattle rancher wouldn't appreciate an espresso while turning a bull into a steer?).

There's good business in forecasting staff. Get a CTO to solve the mundane technical details and call SoftBank.

Consider replacing "silverware" with "climate" or "markets" in your deck.

Ah yes, the great music towers of 2020. What a time to be alive.

The Music Towers of the 1960s[1] were pretty intense too

1 - https://news.artnet.com/art-world/taiwan-kinmen-island-sonic...

(cue Leonard Cohen's masterpiece, "Tower of Song")

These are not startup ideas, but activities that people do. A startup-idea might be to find a way to make these activities easier.

To make this list more useful, there should be two additional datapoints: 1. How many people does the activity impact 2. How much does the activity impact each person. The product of these would be the total impact. Sorting by the amount of impact would give the most valuable areas for innovation, and there should already be a plenty of businesses focused on these activities.

Some (most?) ideas might be about how to make these activities disappear completely.

What's missing is a resource that would let you understand each activity, it's tools, processes, problems, etc.

Actually, if you Google them, you'll find the ONET page for the job descriptions that includes the activity and has more information about the task, e.g. https://www.onetonline.org/link/summary/51-8011.00 (tasks are the 'startup ideas')

Even without the list being suitable for mockery, ideas are EASY. Ideas are like opinions, everybody has one. Take a shower, have some ideas. Go out for drinks, hear a bunch of drunk people talking about the great idea they just had. Winnowing out the chaff (not done here) is harder because not everyone agrees on what's chaff, and building a successful company for any of many definitions of success is harder still.

The idea is when an egg is fertilized. Producing a baby is a lot more work, and producing a successful adult is a lot more beyond that.

Having a good idea is at least half of what will determine a successful product.

The problem is that people confuse having an idea with having a good idea. The latter will be something that has things like a justifiable business model, a business plan (seriously, things like path to market and understanding the consumer market are important), something unique that can hook users etc.

For a true Scotsman's definition of "good"

It might be closer to say that the ideas are like producing sperm. Easy and fun to do, but getting one to an egg requires some logistical expertise.

So fertilizing the egg is getting funding, and a successful exit is the kid moving out at 18.

I'd prefer a virgin birth, then selling the child a few years later

> ideas are EASY

That's true. I see a few things behind the monitor right now.

1) Trees. Let's do something with them, like planting them at scale, maybe mess with their DNA, WCGR?

2) The window has birdshit on it. Let's build window cleaning drones.

3) Ceiling lights. These are really crap make a really smart, well designed one for office environments.

3) Surprisingly enough (to me); Unifi has done just that. https://store.ui.com/collections/led/products/unifi-led-pane...

Their whole business model seems to be "take quality white box hardware and build a management suite around it". It's not even particularly expensive when compared to white box gear. They already covered the traditional IT MSP model with voice, data and video; so it makes sense they would branch out into facilities management (since a lot of non-tech shops are pulling office IT infrastructure under facilities anyway).

Not fond of their colour temperature. If they had an option to change it, I'd install them all over my house.

Hundred percent agree. Especially if they upgraded it with >95 CRI and adjustable colour temp based on separate warm-white LED chips (not an array of RGBs).

There are multiple companies building and selling window cleaning robots. Drones are... terrible for this use case.

You know, I never realized until I read your comment that "drone" seems to specifically refer to a UAV, and "robot" to a terrestrial type.

Which is strange to me, since when I hear "drone" the first animal that comes to mind is an ant. I thought it'd just be a catch-all term for any kind of bot.

A drone ant is a (fertile, male) flying ant.

I didn't mean a flying thing and I know multiple companies exists. Good ideas doesn't need to sound novel at their core.

I can't recall now where, but heard the word drone used for robots. Maybe the boundary is getting blurrier?

Well, I think the word "idea" covers a vast range, everything from ideas that you come up with in the shower, to ideas you come up with after a decade of specializing in a specific domain, noticing unmet needs and figuring out how those needs can be met.

What is important to understand is that both types of ideas are useless to the layperson. The former is useless obviously because it hasn't been validated. The latter is also useless however because even if the layperson learns it, they won't have the knowledge and expertise to understand its value, much less know how to execute it.

This is why when someone comes to you and says "I have an idea, but you need to sign an NDA first before I can share it", you should just smile, nod and politely change the subject.

You need to execute well, but you also need to be working on the "right idea" for it to become successful.

The simple core ideas behind Dropbox and AirBnB were seemingly disregarded by many devs and investors when first pitched, but have proven themselves to be valuable. Even Google wasn't initially considered to be a good idea where Yahoo rejected acquiring them for $1M, but turns out Search Engines displaying contextual relevant ads has become one of the most valuable businesses in history.

Ideas are easy.

Having valuable, proven insights that nobody knows, and are possible to extract value from , that's harder.

A thoughtful person can have insights that few knows.

but the really hard part is: "and are possible to extract value from"

I think that the point of this list is to say "there are lots of starting points, just pick one and iterate on it until it becomes something with value". It's a pretty well-worn maxim that ideas are worthless and execution is everything, but this is another way to present it.

it's a cute saying, but I've never actually been shown a success were someone just executed perfectly without any idea to execute from.

I've heard multiple times that for a startup to succeed, it needs an idea, a great team to execute it and funding, and that it is enough to have two out of these three. Invariably, the people that say the latter are the ones that have no ideas.

I have lots of ideas, but have very little success. It's super annoying that I'm not able to generate as much interest or momentum on them as they deserve.

Why is it so difficult to move us forward?

> 4678. Develop transactional Web applications, using Web programming software and knowledge of programming languages, such as hypertext markup language (HTML) and extensible markup language (XML).

Where has this list been all my life? I could have been a zillionaire....

I enjoy how some of them run together. 10933 and 10934 as an example:

Clean, shape, and polish fingernails and toenails, using files and nail polish.

Remove bones, and cut meat into standard cuts in preparation for marketing.

I am wondering if you have read all of the 20k rows in order to spot it

I'm wondering the exact same thing!

I'm confused about 10933. Isn't that just a regular nails salon?

Isnt wework just a regular real estate company

Yes, but then you remove the bones from the freshly polished fingers and toes, and cut the meat into standard cuts in preparation for marketing.

I've heard somewhere the way to find start up ideas is to take something like this list and add AI to it. How do you add AI to a nail salon start-up?

How about: build an augmented reality display for customers who are waiting to get their nails done. It's got a camera, and it will show users what their nails would look like with any of the styles offered by the salon applied.

Looks like some of the large retailers are already doing this: https://mobile-ar.reality.news/news/modiface-applies-its-aug...

Nails as a service - monthly fee, your nail technician comes to you once a week

Nails as a community, you get to live and work in a space exclusive to people with wonderful nails.

Isn't 737 just a fast food restaurant?

When I was a child, I really hated having to cut my nails. One of the reasons why I developed an interest in technology, is that I wanted to build a device that cod take care of this horrible job for me. Insert hands, and bam, nails trimmed to the ideal length.

To this day, I've neither seen nor built such a device. Devices to print patterns on your finger nails do exist, however.

Well, considering ideas are a dime a dozen this site is worth 20,000 / 12 * $0.1 = $167.

Maybe less actually :)


there aren't even 20k ideas...

18514 so they are worth $154.28

I'm sure there's a method we could use to find overlap and reduce the worth of items that overlap ...

"Maybe less" indeed ;-)

Or ideas are worth nothing, it's execution that matters... so $0?

Looks a bit random.

If you give the start of this list to transformer


the generated ones look as good as those, example:

5.Evaluate a vehicle's operation and use of the system and record information on the system, such as the number of miles driven, total mileage, speed, and odometer reading. 6.Prepare and administer reports to provide data on vehicle performance, and to support vehicle performance maintenance and repair.

Maye someone should interface this to the patent applications system.

The idea is that if a bot can generate patent applications they aren't very novel.

From this we should backtrack and look at existing patents and throw a large percentage of them out : )

Could a massive list like this help stop frivolous patents from being viable, or granted? Perhaps if all people could just add their ideas to this list, over time, we could prevent the ability for anyone to patent anything if the idea already existed.

Or do I still not understand patents?

Could this list exist at a stop-patents type website where even the patent office could do a quick search to see if an idea is unique and original, or has been thought of easily already and put on big list?

You dont understand patents. You cant patent an idea. You patent a specific way to do something. For example, you cant patent a mop, but you can patent a specific design of a mop, a way to build it. You cant patent encryption but you can patent an encryption algorithm. You cant patent "compressed video" but can patent a compression format. A patent requires a detailed writeup of how to do something.

> 10799. Align wheels.


> 11031. Transport the deceased to the funeral home.

I think we've found another use for robots.

You do the align wheels thing and I'll get right on #660.

> Apply and release hand brakes.

I’m pretty sure your mocking. But the funeral industry needs to be disrupted. Went through the process a couple weeks ago and it felt pretty scummy.

I don't really have a point here, but my wife and I recently looked at burial options and there's a beautiful cemetery near us but the only really affordable option is cremation and storage in a wall of tiny notches with little doors with an indoor/outdoor viewing area and garden for visitors.

I have started referring to it as, "Our Cubby" and my wife hates it.

A lot of it is legal, but there is some change happening. WA recently made it legal to compost* human remains, perhaps the first new option outside of burial and cremation?

*in a very specific way https://www.cnn.com/2019/05/22/us/washington-human-compostin...

Very good idea, as big plague is coming...

> 7384. Prepare customers' nails in soapy water, using swabs, files, and orange sticks.

Some are oddly specific

> 12399. Give enemas and perform catheterizations, ear flushes, intravenous feedings, or gavages.

Very specific, and somewhat disturbing.


How did you went so far while reading the list?

I ctrl-f'd for something else and had to do a double take when I saw that. The worrying part is presumable someone actually does that job - which seems to only involve sticking tubes into people, never mind which hole.

Someone's gotta do the job.

I've thought that inserting IVs is something that nurses struggle with a lot and maybe could be automated. It would be an absolute nightmare to develop that robot though, as you'd have to go through all of the medical device certification on top of dealing with all of the edge cases of human anatomy.

It's 1216 and 2399 actually.

Goes nicely with 18497 "Wash and reset wigs" provided that the wigs are made of ginger hair.

“Dump sugar dust from collectors into melting tanks and add water to reclaim sugar lost during processing.”

"524. Solicit new business." how did i never think of this before...

Then when done:

7024. Collect payment upon job completion.


> 1600. Collect payments from customers.

All of these are huge businesses. Debt collection. Factoring. Etc.

done recursively, it is the koan for MLM.

Nah this can't be right. Blockchain, AI and Machine Learning need to be huge. No wonder no startups are getting funded nowadays /s

Seems like the best start up idea is one where you USE stuff to create stuff people USE.

I've cracked the code.

Such use, much wow.

Idea 20100, make a filtering and full text search service for all of these ideas for better reading experience.

This is hilarious. The first one that caught my eye was helping passengers leave and airplane.

Was this an ML project? This looks like the output of a very basic machine learning project.

The commentary here is the best part of this. As (un)valuable as this list may be, it definitely gets the wheels turning in your mind.

"883: Design computers and the software that runs them."

Seems to check out

> 12064. Manage leasing of facility space.

Sounds profitable.

> Regular Brain - Lease an Office

> Nebula Brain - Virtual Offices like Regus Offices

> Galaxy Brain - Coworking spaces

> Exploding Galaxy Brain - WeWork

> Super Exploding Galaxy Brain - OurWork a per second billed coworking/living space

OurWork ?

More like Amazon Workplace Services, if Amazon does it.

WeLive.com is kinda getting there.

Just prepend "Uber for " or append " for developers" and this is indeed a startup idea list.


> 20. Carry messages or documents between departments.

"I have people skills! I am good at dealing with people! Can't you understand that? What the hell is wrong with you people?"

Unsurprisingly, contains some duplicates:

#11755: Separate scrap waste and related materials for reuse, recycling, or disposal.

#11784: Separate scrap or waste materials for recycling, reuse, or environmentally sound disposal.

#11997: Identify or separate waste products or materials for recycling or reuse.

  2948. Deliver babies.

Here's an expired patent you can take advantage of: https://patents.google.com/patent/US3216423A/en

Succinctly named "Apparatus for facilitating the birth of a child by centrifugal force"

I raise you "2603. [...] deliver babies safely"

Sounds like something to worry about once we've achieved market domination.

Can you patent motherhood?

holy shit, some of these are making me actually cry. this list is something else

Uber for Babies™

To clarify, is this meant as a joke of some sort?

Just not sure if it's just a satirical thing -- like, "Hah, look at the silly word jumbles my script generated!" -- or if it's meant as a tool for something.

This combined with the WeWork news is not doing anything to counter my opinion that startup culture is both practically vacuous and morally bankrupt.

If I upload the right HTML to github, can I get you to move to Portugal and become a cod fisherman?

>1556. Enter data into computers for use in analyses or reports

Literally scale.ai

I was thinking about Power BI.


A few of these are totally ripe for disruption!

>51. Conduct raids and order detention of witnesses and suspects for questioning.

Definitely my favourite economic disruption!

Is this what people mean when they say that start up ideas are worthless? Or is this a joke?

I subscribed to a few newsletters that would send me startup ideas and problems to solve on a daily basis.

But somehow that didn't do anything for me.

Some even would offer to connect me to people with the problem, but I had the feeling even with people who would say "I have problem X and I'd pay for a solution" the whole process was still too abstract for me.

I think this has more to do with the business side of things than with talking to potential customers and implementing a solution.

How do I start? What are good first steps if you don't know anyone, don't have a network etc.

Flip it on its head: don’t start with an idea. Start by identifying a group of people you belong to (e.g. “Rails developers”), because it gives you an instant leg up in knowing their struggles. Then research their struggles. Write about them, and solve them. Grow an audience (read: email list). After all that research you’ll have a solid idea of what they already want and buy, and you can create that for them. Plus you’ll have built up an audience to sell it to!

It’s not necessarily quick but it works. Amy Hoy & Alex Hillman were the ones who taught me this strategy and I heartily recommend reading everything they’ve written at https://stackingthebricks.com :)

It is a lot easier from a technical standpoint if the group you pick is not already in tech. Replace "rails developers" with something like "antiques collectors" or "massage therapists" or "arborists" any other profession that isn't saturated with software options already. These people all have smartphones and computers but probably have only crappy options for how those tools specifically accelerate their work.

I read their books, somehow I felt lost rather quick after starting to pursuite their ways.

Can you share the links to those newsletters? Sounds interesting, thanks!


The best link is: https://nugget.one/ideas

We have 4,000+ useful ideas manually sourced from people with problems to solve.

> 7819. Participate or assist in raids and arrests.

Yeah, well, not sure about this one.

Lots of money to be made selling equipment for those.

The five shortest ones (11, 13, 13, 13 and 13 characters long):

> 874. Carve meat.

> 5221. Dig trenches.

> 8361. Graft plants.

> 10799. Align wheels.

> 16441. Shampoo hair.

And the five longest ones (291, 305, 308, 317 and 332 characters long):

> 6231. Keep abreast of government regulations and emerging Web technology to ensure regulatory compliance by reviewing current literature, talking with colleagues, participating in educational programs, attending meetings or workshops, or participating in professional organizations or conferences.

> 12863. Design, develop, select, test, implement, and evaluate new or modified informatics solutions, data structures, and decision-support mechanisms to support patients, health care professionals, and their information management and human-computer and human-technology interactions within health care contexts.

> 7530. Keep abreast of game design technology and techniques, industry trends, or audience interests, reactions, and needs by reviewing current literature, talking with colleagues, participating in educational programs, attending meetings or workshops, or participating in professional organizations or conferences.

> 4806. Direct environmental programs, such as air or water compliance, aboveground or underground storage tanks, spill prevention or control, hazardous waste or materials management, solid waste recycling, medical waste management, indoor air quality, integrated pest management, employee training, or disaster preparedness.

> 5895. Compute, retrace, or adjust existing surveys of features such as highway alignments, property boundaries, utilities, control and other surveys to match the ground elevation-dependent grids, geodetic grids, or property boundaries and to ensure accuracy and continuity of data used in engineering, surveying, or construction projects.


My method for determining these things (which may be of interest to some), using just my favourite text editor, rather than slurping it into a REPL for some programming language and manipulating it thus:

1. Copy the whole document to the clipboard;

2. Paste in Vim (and note at this point that lists lose their markers in favour of just a tab character, so the numbers disappear);

3. Manually remove everything that’s not the list;

4. Manually remove the tab at the start of each line with `:%s/^\t//` or block selection (this could also be done in the next step with slight modification, matching `^\t\(.\+)` and using `submatch(1)`);

5. Prefix each line with its length and the line number using `:%s/.\+/\=len(submatch(0)) . "\t" . line(".") . ". " . submatch(0)` (I actually used asterisk instead of \+, but HN formatting hates unmatched asterisks; see `:help sub-replace-\=`; another approach for the line number part alone would have been to block-insert `1.\t` on every line, then block-select all but the first, and use `g<CTRL-A>` to increment them, see `:help v_g_CTRL-A`);

6. Sort by line length with `:sort n` (since I had put it at the start of the line for convenience);

7. Optionally strip the character counts out again with `:%s/^\d\+\t//` or similar.

sub-replace-\= and v_g_CTRL-A are among my favourite not-so-well-known Vim features. They’re not useful very often, but when they are, they’re really great.

My first chance ever to be that guy!

My solution:

  curl "https://unawaz.github.io/stochastic-hill-climbing/tasks/" \
    | awk '{ print length, $0 }' \
    | sort -n -s \
    | sed 's/<li>//' \
    | (tail -n +11 | head; tail)
Updated to remove noise.

The problem with such a pipeline is that it’s not interactive. Now if you had a REPL that would memoize the steps so that it only ever did one `curl` request, that’d be different and useful.


  curl [address] > /tmp/foo
Then, build up the pipeline starting with:

  cat /tmp/foo |
(And I know this is probably "useless abuse of cat", but whatever.)

Congrats on being that guy!

How long did this take you? I don't know vim wizardry and I was curious. Took me 4 mins to do the same thing in js console:

1. copy the text 2. split by \n 3. sort by length 4. you got sorted list

I know all of this off the top of my head, so less than a minute all up. Well, except the line(".") bit; that didn’t occur to me until I’d done it manually with the help of v_g_CTRL-A, and I had forgotten about the "." argument until I looked up its docs—haven’t used it in a while.

Great stuff - however, one thing that is missing is the ability to discuss these ideas in a forum-like medium. This is why I created the GH repo: https://github.com/marcelosousa/million-dollar-ideas. I might code a script that pushes each item as an issue to the repo to at least allow for that.


^^ I built this more or less on a bet about a year ago. Give it a dollar and it will give you a startup business idea.

(No Markov chains here, it's a couple of arrays of industries and keywords and a JS function that uses Math.random()... )

> Shape shoe heels with a knife, and sand them on a buffing wheel for smoothness.

> Adjust felt hammers on pianos to increase tonal mellowness or brilliance, using sanding paddles, lacquer, or needles.

> Review filed tax returns to determine whether claimed tax credits and deductions are allowed by law.

Three adjacent ones there, could be rewritten as:

* become a cobbler

* become a piano technician

* become an accountant

"Implement dynamic relational" is not listed. Fooey! In fact, it doesn't even contain the word "relational" or "RDBMS".

In general the IT-oriented items seem like variations of "fix my goofed up spreadsheet". Did they scrape the tech support logs?

>It maybe is even the listing of approximately all startup ideas.

I wasn't able to find cloudflare in there

What would make this for more useful is to have a list of even mildly _validated_ ideas, even by showing that there are existing competitors or a market for it. Or pointing to a gap where it exists.

Validation is the first step to product market fit.

There are only 18514 ideas there

I wrote an app that turns those into proper startup ideas. Source here: https://news.ycombinator.com/item?id=21126901.

A startup idea: an app to generate a startup idea (based on this list).

Here you go:

  (function () {
    var els = document.getElementsByTagName("li");
    var startups = ["A web-based platform", "An AI-powered platform", "Uber for",
                    "Subscription-based service", "Blockchain", "A folder that syncs"]
    for(var i = 4 ; i < els.length ; ++i) {
      els[i].innerText = startups[Math.floor(Math.random() * startups.length)] + " for people who " + els[i].innerText;
Paste it into dev console on that page to receive an instant list of proper startup ideas.

The code is GPLv3, and can be trivially modified to generate further interesting startup ideas.

EDIT: Modified to add a greater variety of instant startup ideas!

Cool. But it should be in React Native and AI to raise some money.

Don't know about React Native, but updated the app to generate AI-related ideas (and some others).

Replace xylophone bars and wheels.

Watch out guys we've got a unicorn over here

I like it. I created this mini API. Throws random ideas: https://startup-ideas-api.btl.es

20 001: collect, organize and rank a library of startup ideas

> Select sizes and types of molds according to instructions.

Fair enough

Title should be: "20,000 incremental startup ideas that will waste years of your life and generate mediocre returns for your investors".

>524. Solicit new business.


6835. Provide testimony as a witness in court.

“Perform oral practice periodontal surgery on the jaw or mouth.” Dude! I even have a spare room, we can totally do this one!

"Process or book convicted individuals into prison." AirBnB and Uber for prisoners. Sounds legitimate :)

768. Measure and mark cutting lines on materials, using a ruler, pencil, chalk, and marking gauge.

Ideas are a dime a dozen; Everyone has them.

What is rare is the hard work needed to make an idea a reality.

I never have any that are even half-decent. Plenty for things that aren't businesses. Sometimes I act on those but they don't make me money. None for businesses. When I think I have one it always turns out there are a few players in the space already, one or more of them appear to be doing a good job with no clear way to outdo them, but nonetheless all appear to be struggling (so it might have been an OK idea when no-one was doing it but trying now would pretty much be doomed, spending weeks to months to get off the ground to take 10th place eating from a small pie). Or it's something you'd have to have FAANG resources and/or a huge pile of cash available to do (owning/collecting relevant machine learning datasets, mostly).

When other people say "hey I'd love an app that let me do X, you should build it, you'd make so much money" it usually turns out there are already 20 in the app store, and they "love" the idea so much that they'd never bothered to look for one. This is true whether it's for themselves, or for their job.

Right, but these don't even sound like ideas, they sound like sequences that fell out of somebody's intro to AI project.

Everyone has ideas, but not everyone has good ones.

How old are some of these? For instance, is #402 even relevant today?

"Splice or solder cables together or to overhead transmission lines, customer service lines, or street light lines, using hand tools, epoxies, or specialized equipment."

No...no no don't do that you will stand a good chance of being electrocuted.

Replace xylophone bars and wheels.

Watch out you guys we've got a unicorn

"117. Buff defective areas of inner tubes, using scrapers."


Part of the job description of Tire Repairers and Changers: https://www.mappingyourfuture.org/planyourcareer/careership/...

Oh no. There is a banner ad now at the top of the list.

uBlock Origin got my back. No ads to be seen.

2413. Google maps

Well, if you can out-Google Google Maps you'd definitely have a huge business.

15728 would be an organised crime startup

Or the police becoming privatized?

this is an outrage: there are only 18,514 startup ideas in there.

haha > 5131 Move and arrange furniture and turn mattresses


ideas are cheap

These are so useless


i have a backend tech stack that i believe is scalable and efficient (REST or websockets). i'd like to prove this out with a demo or app. i can do a pure-js SPA frontend (eg fitno.us) but i'm terrible at UI so it's not gonna be pretty

is there a free service that i could provide that people would use at high scale ? there doesn't need to be a way to eventually monetize it

required: store data to database optional: rest calls to 3rd parties

anyone have an idea they're willing to share ?

My FBI guy: a moderated chat app.

Instead of just a 2-way chat program, there is a 3rd participant "the FBI guy" who also joins the chat. Every message sent must get approved by the FBI guy before it goes through to the recipient.

It should also be possible to send messages to the FBI guy directly. e.g. "Does she/he like me?"

i like the concept (and it's sort of the inversion of another chat app idea i've been tossing around along the lines of cyrano de bergerac) but i don't see what's going to drive adoption

what's the motivation for a user to use this app vs un-moderated chat ? is it driven by the moderator, eg a parent wanting to limit what their kids can say or see ? or is it one of the 2 chatting parties that prefers to have someone else sanitize content before they have to see it, eg to ensure that they don't get d*ck pics ?

in the latter case, what motivates the moderator to participate ?

I think it would have to grow (or die) virally. Like GMail in the early days, moderators would get accounts by being invited.

Although there probably is a way to pitch it as a parental control, I think other demographics could enjoy it too. As you say, the moderation basically ensures no creepers or d*ck picks. That will probably encourage more female users to trust the platform, which would certainly help to boost overall numbers.

It could be a means of chatting to strangers, where the moderator chooses 2 people to pair up. That could get interesting. The idea is still pretty half-baked, but I'd like to explore it more, and maybe use it as a testing ground for decentralised chat technology.

For people who are unsure about social norms of chatting to people, such as those with Aspergers, autism spectrum, or who are just plain shy, I think being a moderator would be a great way to learn how to have an ordinary conversation. I'd also want to make the website fully accessible, with text readers for the blind, and automatic language translation so that people can make friends in other countries.

+ the agent can retract words or make edits. retracts are visible, but edits are just counted.

this sounds fun :)

and chain the fbi-guys together into a game of telephone

How much do users pay for the privilege to use this app?

No idea, I'm a programmer not a businessman. I don't think OP actually minds about income: "there doesn't need to be a way to eventually monetize it"

I guess some money could be made by using the chat logs to train AI chat bots, and selling those bots. I have no idea how that would work in practice though.

The traditional answer for people wanting to get free load is to run an image-hosting service:

* Accept uploads.

* Push them to S3 / similar.

* Marvel as you have to filter out nasty images.

* Struggling under the load of hotlinked images.

* Start performing rewrites to serve ads.

* Add chat/comments to try to build a community.

* Get more and more slow and ad-heavy over time.

The process repeats every few years.

Applications are open for YC Winter 2022

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