Random App Ideas
Every now and then you start thinking “wouldn’t it be nice to have an app for X”. When I was in that situation, I took a note. Then that note grew, I cleaned up some absurd ones and added some more. I implemented some of these ideas of mine, the rest formed a “TODO” list.
At some point I realized I won’t be able to implement them, as the time needed is not something I’m likely to have in the near future. But I didn’t like to just delete the notes. So here they are – my random app ideas. Probably useless, but maybe a little interesting.
- Receipts via smartphone NFC – there are apps that let you track your expenses, but they are tedious. There are apps that try to OCR receipts, but they vary too much in different parts of the world for that to be consistent (there are companies like Receipt bank that do something like that, but it’s not the way I’d like this problem solved in the long run). So I thought stores can offer the option for NFC receipts – you just tap your phone to a device (or another phone) and get the receipt in an electronic form. Not the picture, but the raw data – items, prices, taxes, issuer. Then you have to option to print it if you like. Of course I realized at some point that first legislation should allow for that – in many cases you must issue a paper receipt and the digital one is not “legal”. But the idea still remains viable and probably not hard to implement. It will allow much easier tracking of expenses by users and will save a lot of paper.
- Am I drunk? – breathalyzer add-ons for phones are not something many people would buy. My idea was to detect drunkenness based on motion – the accelerometer (and possibly other sensors) could be trained to recognize a drunken walk on the street, thus informing you that you are… drunk. What’s the use of that – maybe if while walking to your car the app alerts you, you may be less likely to drive.
- Mobile scavenger hunt – scavenger hunts have always been interesting. I remember we devised one for a friend’s birthday a few years ago. It would be nice, though, to be able to build such scavenger hunts dynamically, based on multiple configurable building blocks. The app could generate multiple outcomes based on certain rules (mostly using the GPS of the phone). It could be timed, it could include scanning barcodes (including those on actual products), visiting particular addresses, listening to songs and their lyrics, extracting data from Wikipedia articles, etc. “Generate scavenger hunt” (for a friend or randomly for yourself) might be interesting.
- Sentiment analysis prior to commenting – something that somehow plugs to facebook or other social media (or screen-scrapes it?), plus a browser plugin (for desktop), that does sentiment analysis on the facebook comments you are currently typing. And recommending against them if they are too negative, personal attacks, etc. Kind of like a cool-down timer saying “thing twice before destroying the online discussion with that comment”. It is a problem of online communication that too often it gets out of control (and the same people wouldn’t be as rude in real life as online)
- OCR shop signs – this was more of a Google glass app than for regular phone, as it requires capturing one’s surroundings. The idea is to crowdsource the current shops and restaurants and put them on a map. At the moment we rely on old and incomplete data – either data that the owners of a shop or restaurant have put, or that some contributor has added. However, shops and restaurants change often and you may not know that something has moved. It doesn’t sound that useful, but worth a thought.
- Algorithmic impressionist painter – you train an algorithm and it creates a random picture. Now, that has already been done (I’ve been having the idea ever since I created my algorithmic music composer). And I’ve heard critics destroy the algorithmic paintings. But it’s an interesting experiment nonetheless.
- Life game – basically record your actions and get rewarded for good deeds. The game can give particular challenges (e.g. “go help a soup kitchen”), and bonuses. The challenge here lies in data protection – depending on the level of detail, the application may have too much personal data that should not leak. Encrypting everything with a private key stored in the secure storage of your device may be one way to resolve that. I know it sounds a bit dystopian and now, after seeing something like that implemented in China, I’m happy that I can say “oh, I had this idea six years ago” and I’ve grown since then. Anyway, in my rendition it’s a voluntary game, not a state-imposed citizen score. But re-purposing technologies is a specialty of regimes, so maybe it’s a bad idea to build such a system in the first place.
- Random hugs – you know those “free hugs” signs. You can have an app that detects other users of the same app via WiFi direct (or other GPS or non-GPS proximity/distance detection) and offers both of you to do a random hug. Yes, yes, I know it is awkward to hug strangers and it may be abused for sexual misconduct. In my defense, when I had the idea six years ago, the public attention and knowledge about sexual abuse was not on the levels it is today. Still, with a proper rating system and “report abuse”, this may not be an issue. And as hugs are considered a good thing for psychological health, it might not be such the dumb idea that it sounds.
- Startup idea generator – good ideas for products often rely on existing knowledge and skills of the founders. However, the “serendipitous” combination of the skills and knowledge may not occur automatically (and it will look obvious in hindsight). That’s why there could be an app where you list you skills and knowledge and it tries to connect them into meaningful products. It won’t be trivial, as it would have to be some sort of AI that understand what’s behind given words.
- Semantic adapter for websites – many websites offer simple services (e.g. hairdressers, car repair, accommodation) but finding such services nearby and comparing them is hard. So an adapter for pages can be created that turns each website into a semantic version of it, and then a marketplace-like service can crawl them and allow for comparison. When “semantified”, the sites will contain all necessary information, including location and price, and so they will be able to be meaningfully parsed. Another feature of semantifying will be the ability to sell the data on your site, since it will be well structured (there do exist websites that scrape existing sites to offer that, but making the site itself present the data in a meaningful way is better)
In all cases above, there are interesting technical challenges – devising a standard receipt format and optimizing the UX of receipt exchange, training a model to detect drunkenness based on accelerometer readings, training a painting algorithm, doing sentiment analysis on small pieces of text, using niche technologies like WiFi direct for proximity, data protection and cryptography, OCR-ing pictures of surroundings. This is why I’m sharing them in this blog.
This is in line with my previous posts about making side projects with technologies that are new for you. The app may end up useless, or not used by many people, but you’ve still learned an important skill and applied it in practice.
Every now and then you start thinking “wouldn’t it be nice to have an app for X”. When I was in that situation, I took a note. Then that note grew, I cleaned up some absurd ones and added some more. I implemented some of these ideas of mine, the rest formed a “TODO” list.
At some point I realized I won’t be able to implement them, as the time needed is not something I’m likely to have in the near future. But I didn’t like to just delete the notes. So here they are – my random app ideas. Probably useless, but maybe a little interesting.
- Receipts via smartphone NFC – there are apps that let you track your expenses, but they are tedious. There are apps that try to OCR receipts, but they vary too much in different parts of the world for that to be consistent (there are companies like Receipt bank that do something like that, but it’s not the way I’d like this problem solved in the long run). So I thought stores can offer the option for NFC receipts – you just tap your phone to a device (or another phone) and get the receipt in an electronic form. Not the picture, but the raw data – items, prices, taxes, issuer. Then you have to option to print it if you like. Of course I realized at some point that first legislation should allow for that – in many cases you must issue a paper receipt and the digital one is not “legal”. But the idea still remains viable and probably not hard to implement. It will allow much easier tracking of expenses by users and will save a lot of paper.
- Am I drunk? – breathalyzer add-ons for phones are not something many people would buy. My idea was to detect drunkenness based on motion – the accelerometer (and possibly other sensors) could be trained to recognize a drunken walk on the street, thus informing you that you are… drunk. What’s the use of that – maybe if while walking to your car the app alerts you, you may be less likely to drive.
- Mobile scavenger hunt – scavenger hunts have always been interesting. I remember we devised one for a friend’s birthday a few years ago. It would be nice, though, to be able to build such scavenger hunts dynamically, based on multiple configurable building blocks. The app could generate multiple outcomes based on certain rules (mostly using the GPS of the phone). It could be timed, it could include scanning barcodes (including those on actual products), visiting particular addresses, listening to songs and their lyrics, extracting data from Wikipedia articles, etc. “Generate scavenger hunt” (for a friend or randomly for yourself) might be interesting.
- Sentiment analysis prior to commenting – something that somehow plugs to facebook or other social media (or screen-scrapes it?), plus a browser plugin (for desktop), that does sentiment analysis on the facebook comments you are currently typing. And recommending against them if they are too negative, personal attacks, etc. Kind of like a cool-down timer saying “thing twice before destroying the online discussion with that comment”. It is a problem of online communication that too often it gets out of control (and the same people wouldn’t be as rude in real life as online)
- OCR shop signs – this was more of a Google glass app than for regular phone, as it requires capturing one’s surroundings. The idea is to crowdsource the current shops and restaurants and put them on a map. At the moment we rely on old and incomplete data – either data that the owners of a shop or restaurant have put, or that some contributor has added. However, shops and restaurants change often and you may not know that something has moved. It doesn’t sound that useful, but worth a thought.
- Algorithmic impressionist painter – you train an algorithm and it creates a random picture. Now, that has already been done (I’ve been having the idea ever since I created my algorithmic music composer). And I’ve heard critics destroy the algorithmic paintings. But it’s an interesting experiment nonetheless.
- Life game – basically record your actions and get rewarded for good deeds. The game can give particular challenges (e.g. “go help a soup kitchen”), and bonuses. The challenge here lies in data protection – depending on the level of detail, the application may have too much personal data that should not leak. Encrypting everything with a private key stored in the secure storage of your device may be one way to resolve that. I know it sounds a bit dystopian and now, after seeing something like that implemented in China, I’m happy that I can say “oh, I had this idea six years ago” and I’ve grown since then. Anyway, in my rendition it’s a voluntary game, not a state-imposed citizen score. But re-purposing technologies is a specialty of regimes, so maybe it’s a bad idea to build such a system in the first place.
- Random hugs – you know those “free hugs” signs. You can have an app that detects other users of the same app via WiFi direct (or other GPS or non-GPS proximity/distance detection) and offers both of you to do a random hug. Yes, yes, I know it is awkward to hug strangers and it may be abused for sexual misconduct. In my defense, when I had the idea six years ago, the public attention and knowledge about sexual abuse was not on the levels it is today. Still, with a proper rating system and “report abuse”, this may not be an issue. And as hugs are considered a good thing for psychological health, it might not be such the dumb idea that it sounds.
- Startup idea generator – good ideas for products often rely on existing knowledge and skills of the founders. However, the “serendipitous” combination of the skills and knowledge may not occur automatically (and it will look obvious in hindsight). That’s why there could be an app where you list you skills and knowledge and it tries to connect them into meaningful products. It won’t be trivial, as it would have to be some sort of AI that understand what’s behind given words.
- Semantic adapter for websites – many websites offer simple services (e.g. hairdressers, car repair, accommodation) but finding such services nearby and comparing them is hard. So an adapter for pages can be created that turns each website into a semantic version of it, and then a marketplace-like service can crawl them and allow for comparison. When “semantified”, the sites will contain all necessary information, including location and price, and so they will be able to be meaningfully parsed. Another feature of semantifying will be the ability to sell the data on your site, since it will be well structured (there do exist websites that scrape existing sites to offer that, but making the site itself present the data in a meaningful way is better)
In all cases above, there are interesting technical challenges – devising a standard receipt format and optimizing the UX of receipt exchange, training a model to detect drunkenness based on accelerometer readings, training a painting algorithm, doing sentiment analysis on small pieces of text, using niche technologies like WiFi direct for proximity, data protection and cryptography, OCR-ing pictures of surroundings. This is why I’m sharing them in this blog.
This is in line with my previous posts about making side projects with technologies that are new for you. The app may end up useless, or not used by many people, but you’ve still learned an important skill and applied it in practice.
The receipt app is something that I thought about myself, too.
I see a particular application in calorie counting. There are some apps that recognise food expenses (e.g. Revolut identifies receipts from restaurants, coffee shops, etc.) but they don’t go very far.
The missing link is connecting the individual receipt items to products in that store and then counting the calories. Other apps like MyFitnessApp have an enormous database of products and their calories but require a manual entry. Basically, a mix of the two can (almost) do the job even without a NFC interface – it may try to decompose the receipt amount into plausible product mixes. But it is still not as good as having a NFC interface that would transmit the receipt content digitally. And, yes, legislation requiring the technical capability to issue BOTH paper and digital receipts may help.