Throwaways 2.0
Every quarter I intend on writing a few ideas which I think are interesting but not ideal for me to work on. Some are not ideal because of the lack of industry knowledge, others are just not exciting enough for me. I will share some of them with you here in hopes that it inspires you. If you do choose to work on these ideas or want to learn more about my thought process, feel free to reach out!
The ideas are not listed in any particular order, and some are not even framed as concrete ideas, just trailing thoughts. See my first post here.
Here they are:
Instagram Meets Credit Card Purchases
For a very long time companies have been trying to match your online and offline behaviour. So far Facebook and Google have been really good at this via targeted ads and lots of data about your usage. If you like a post, or search a product, these sites will find a way to make you purchase from one of their advertisers. When this first started to happen it was very creepy, but its annoying how we simply have accepted this as a way of life and moved on. I’d argue today Google and Facebook know more about me than I do about myself, (even with all the self reflection).
The natural question that they are asking however, is how can they get better. Lets start with Facebook. Facebook wants to know more about what you are doing in your free time and how you are interacting with other humans. To Facebook, it seems every human is just a node, and the more they can understand about node to node interactions, the more they can sell that to advertisers. This is why they have purchased Instagram and Whatsapp, to try and see how we are communicating. Whatsapp messages are not private, and that information is undoubtedly shared with your Facebook account (ever wonder why Facebook messenger asks for your number?) and is used to understand you better. Some have even claimed Facebook calls are recorded and tracked.
The other thing Facebook has done is launched a marketplace. This way they can understand what you are buying and selling, and at what price point you can be targeted at. I am not sure about the marketplace uptake, but make no mistake it is very useful for understanding your offline behaviour. Try searching for a desk and a chair on Facebook marketplace, then messaging your spouse on Whatsapp about the renovation being done, then posting on Instagram with the hashtag #homeoffice. A week later you may receive ads for improving your small business. All you did was provide them the data and they sold you crap you didn’t realize existed. But if you bought the chair from the marketplace and the desk from somewhere else…they have incomplete information. So how can they get that information? Read on.
Now lets think of Google. Google has been trying to get into payments for some time. They launched the notoriously bad Google wallet many years ago, which failed. Then they tried again with Android Pay and are now back with various ways to make payments on Google. You tie your credit card to Google, and anything you buy, Google will atleast obtain the merchant code, if not the exact product you have purchased. Obviously they will use this information to advertise back to you. Google is also getting so good at payments, that it sometimes replaces the checkout flow for existing ecommerce companies. I had a few meetings with them when I was at Zalando Payments, one of Europe’s largest ecommerce companies. Googles pitch? Let us introduce a one click buy button on Zalando for Google customers, and avoid your check out. Great user experience, no additional fees. Oh, and we take all the knowledge about your customer purchases and reduce the offerings of your cheaper payment methods. Thanks but no thanks.
Google is also trying to go local. Traditionally Google avoided this space as its hard to advertise for a small region. Local store owners are not technologically savvy and local data changes all the time, plus local search is harder due to relevancy and sporadic placements. Recently however, Google has changed its mind and launched ‘mini cards’ which not only give information about local businesses, but also popular times and reviews. This gives Yelp a run for their money and forces local owners to adopt new technology. Yet Google doesn’t know if you end up shopping at that local business and what you specifically need. There again lies the problem.
So how do these behemoths understand your spend offline so they can connect that last piece of who you are? One method is payments. If you use your Google wallet at these local stores, or any store for that matter, it is useful insight for Google. However what can Facebook do?
The obvious answer — which was my first idea was create a verified by Instagram or Facebook payments. This will likely be in two steps. One, let people add their credit cards safely to Facebook, Instagram or Whatsapp (already happening in some places). Two, accept payments on third party websites with Facebook (largest store of identity on the internet today). The minute you’re on Nike, paying with Instagram, Facebook already knows you love Nikes and will advertise the next Nike product that you need to match your purchase. Nike wins, Facebook wins, your eventual privacy loses. Building on this for a second, imagine the thousands of ecommerce purchases out there that Facebook Pay or Instagram Pay can cover.
One great example is tickets to events. Right now Facebook events is missing final numbers of who is going, where and how much were tickets, and sources of sellers/advertisers. If Ticketmaster, Stubhub and Meetup all started allowing Facebook Pay, it would be super easy to not only know more about the users, but know about demand for events — something that is probably a billion dollar industry alone. Side note, demand prediction for companies is also an idea worth exploring — Ticketmaster today has no real good way of predicting demand — just ask any one of the scalpers who bought JayZ tickets and undersold them when the concert hall was empty. And that was JayZ!!
The unobvious answer and the opportunity is for a third party (aka you) to come in and match Instagram accounts (or Facebook accounts) with credit cards. The benefit here is that you obtain that same data and are able to match online user activity with offline purchases. There are a number of ways to do that. You can be a payments company and ask your users to login or connect via Instagram. Offer loyalty points or a product differentiator to make yourself better. Crypto could be useful here. You could charge people money to search up their Facebook and Instagram so they know what Facebook knows about them. It seems far-fetched, but with everything going on now, its not impossible. Or you could be a consulting company and help companies accepting payments solve a problem (like demand prediction, fraud etc) and find a way to get them to invite users to login or add their Instagrams. Even an Instagram follow with a smart crawler would do the trick. It’s a huge technical undertaking but its not impossible.
(I only focused on Google and Instagram, but granted there are other top companies trying to understand your behaviour as well. Facebook has tried a number of times to launch payment related schemes and has failed. Instagram now allows you to shop on their app, which is really scary. Apple and Amazon with their payments is also on the right track. It’s hilarious that banks have been sitting on this data for years and they have not even bothered to identify your online identity. Hey banks, you really think the fastest route to a trillion dollar valuation is selling me cheaper interest rate loans?)
Creating a Data Agency
The antithesis of the idea above is a way to block these websites from getting your data while still enjoying their privileges. Please note, I have no idea if this is possible or not. Many have tried to create new social networks that pay you for your information which is one approach to solving this problem. Forcing user uptake with brand new website might not be the best way. If you look at the histories or Google and Facebook, both started by piggy-backing off of existing leaders (Yahoo and MySpace). A new social media or search engine is very hard to do.
What one can do is try and create a website that logs in and lets you use Facebook and Instagram without giving your data away. Any photo you send could be protected and hard for Facebook’s crawlers to dissect, while any messages you send could be easy for you to read but private to a Facebook agent reading what you are talking about. As far as Facebook will be concerned, you are just a bland identity with no activity, since it is their robots that are checking your messages. However you can still use the social media platforms and do what you normally do, knowing your real activity or identity is blocked. Essentially using machine learning to outsmart other machines. I am sure someone in Russia is reading this smiling and nodding.
Some concrete ways of doing this include:
- A Facebook messenger bot using private keys to block your content or keep it private
- An image protector that allows you to see the images clearly, but machines have no idea what it is
- Web browser extensions that change scan the page for content and make it harder for Facebooks bots to read it
- Or a web browser extension that changes what you type into Facebook, but anyone that has the extension can read exactly what you wrote (think along the lines of peer to peer and privatization)
The goal is to avoid giving Facebook so much data. With search, companies like DuckDuckGo make it possible, but in the end its still super hard to avoid these companies.
Citizen Lobbying Against Monopolies
Governments around the world are struggling with companies like Facebook, Google, Amazon, Apple, and a few others. The struggle is to keep these companies in check and make sure they don’t abuse their powers, or that someone doesn’t abuse their processes. Facebooks recent examples come to mind. I am not in Silicon Valley but I would argue that the resounding fear in the valley is the slow demise of Facebook as a result of building processes and data utilization techniques that have not been checked.
Now the obvious problem here is that when you are ‘moving fast and breaking things’ there is no time to question what your end product will actually do to society. Testers can only assess if the product works, but not all the ways it could mess up. Theoretically this could lead to another idea where you have an AI that understands all the malicious intents of humans, and then tests those intents + actions against people building algorithms on Facebook and other platforms.
In talking about this to a few people, we actually came up with a simpler idea. Why not give power to the users of Facebook or any of these large companies. As a user of Facebook today I really only have 3 options — continue giving my data to Facebook, limit my use of Facebook as much as possible, or delete FB entirely. What if I could pay a small amount every month or every year to a non profit entity that tracks the data that I have given away, and the minute a company mishandles my information, they step in and fight FB on my behalf. Its like having a technical lawyer on retainer in case my information is misused or the company does something I disagree with. (Or it could be the holder of my information and can destroy it if its misused, but thats an even bigger dream).
Extending this example, users today have no power or say in the products they use, especially when it comes to technology. The mishandling of data is one thing, but boycotting shouldn’t be the only solution against large companies. Imagine Google returned search results that were racist in nature every time, or Amazon gave you only products that were made in China. It would be difficult to boycott these companies as Bing and other online stores just aren’t as great. However if we come together as users, pool our money and use it effectively against these monopolies in terrible systems, we will keep them in check. Or atleast do a better job of keeping them in check.
Aggregated Visual Symptom Checker
Health checks in the 90s used to be:
- I have a problem.
- Maybe I’ll call a friend and chat about it
- If it Doesn’t seem major, I’ll let it heal
- If it gets worse I’ll see a doctor
Today its more like:
- I have a problem
- Lets google it
- Google confused me, let me call a government hotline or a friend
- Still not sure whats going on, lets open up Babylon or another skype for doctors tool
- Babylon can’t really help, I need a real doctor
I think there has never been a time where patients have been more armed with knowledge about their conditions, but also more confused and more scared. It’s almost like these ‘Skype for doctor’ tools are being used as preventative solutions, instead of real doctor alternatives. My personal fear is that as access to medical information becomes easier and more accurate, we will constantly ping these services for queries because unfortunately our biology fears for our mortality all the time.
In thinking about this, there is one area that I think computer vision could tackle, and that is a visual symptom checker. Web MD is useful when you can type out your problems, but even then its not always able to nail it down. Skype for doctors is not bad, but it costs money and time. There has to be a way for me to take a picture or a video and get a doctor (initially) or a machine to tell me what it thinks I may have. It doesn’t have to be 100% accurate, but high probability scoring is key.
Some companies have done that already when it comes to skin and STDs, but there is no way to aggregate their data just yet. Imagine being able to take the app that can detect eye problems, with the one that can detect skin problems, hair problems, rashes, and other things you can see with the eye. Now you have one tool that can take the best of knowledge from all these apps, and combine the diagnosis. Since images are more powerful than words, this could be a better way to diagnose patients.
Questions to Ponder About
Why should large corporations have data on the people I care about? What if I want to access that information to ‘monitor’ my parents? Kids? Employees? Students? Basically using social data to better care take.
Why can’t there be a tool that starts to learn about things you are interested in or with you knew about and then answers your specific questions. Think of a smarter version of google trends — so instead of daily mentions of key words, it learns the topic holistically and can dive deep into specifics. Imagine it can even predict the future of certain markets based on what it knows.
Can you get insights onto what scientists and leading field researchers are working on without having to go through full research papers or patents? Kind of like snapchat like stories but for new fields of information?
Why is information on your receipt different than the one in your bank or card statement? Don’t banks want to know exactly what you are spending your money on so they can get a better idea of who you are? Sensibill is in this space, but not sure how smart their AI is.
Could you use blockchain for creating fluctuating equities or stock options so if I make a bad hiring decision or someone loses motivation after the probation period/cliff period, I am not stuck with a poor worker?
Why is it not possible to index video based on time intervals so you can quickly jump to the most exciting parts of different videos?
Why did Amazon just create over a dozen patents to do with video? Most of the video patents have little to do with their movie business?
Is there a way to make Google search more equal, so the people with all the money and better SEO capabilities don’t dominate search results?
If you ever watch a millennial work, you’ll notice they have multiple screens open. Social apps (WhatsApp), music (Spotify), sports (various) and their main work screen. How can you leverage that for either better focus, or better management?
Further Reading and Acknoweldgements:
- Facebook tracking more data than you know
- Why Facebook will never change
- Latest patents by the biggest companies
- Facebooks recent patents (scary)
- Robert Rip for thinking about lobbying against Facebook
- Nilum Panesar for helping me think about a world with a more equal Google
Don’t hesitate in reaching out to me over twitter or email! And check out my other work here.
