Building Apps On Top of Your Genome: OAuth For Your Health Data
There’s an interesting TED talk up right now about the ‘Genomic Revolution’ (link to talk). The idea of the talk is that soon our genome will soon be so cheap to sequence, that it will become a key part of our ‘identity’ data. It will be used in providing personalized health care, but it will also be a core piece of data for other applications as well. Resnick encourages us to reflect on exactly what this means, and how we can participate/influence the oncoming ‘Genomic Revolution’.
In fact, there’s lots of interesting ideas even mentioned in the talk (reminiscent of Gattica):
- Dating apps based on genomes.
- Job applications including your genome (e.g. if you want to go to space, you need to pass certain genome requirements).
- Self-improvement/productivity apps based on your genomic pre-disposition?
What else would you build if you had access to this data? (would love to actually hear ideas)
One of the key things for this to happen is to have something that can store this health data, and can then give other app’s access to that data (in a system similar to what oAuth is for Facebook and Twitter Apps right now). This could be an interesting project to work on, and it asks an important question of how ‘data management’, and granting access to data will work in the future. What data do you want to have access to, that you don’t have access to right now - and what would developers build if they could have access to that data? Will online genome data storage and sharing be something the existing Genome companies like Resnick’s company, and 23andMe do first, or is there room for purely software companies to win in this space?
The Locker Project is a cool idea on this front - that helps you to preserve your data in a local store, and grant access to the data from there. This may be the right approach in spirit, however, I think they’re still missing out on a ‘consumer’ focused application of their technology. It could be that they’re sin the process of building the interface/app that will be a big consumer hit based on this idea right now. I’m excited about this space, and I think there’s some interesting approaches to data management and access that I’ll write about in an upcoming blog post.
So… if you had access to this data - what would you build?
App: Automatically Donate 10% of All Money In Your Bank Account to Charity
What if you could automatically donate 10 percent of all the money that comes into your bank account to charity? Would you do it?
This is one application I’ve been toying with building in my mind, but for the time being its more of a thought experiment than a real intention. Still, I think its interesting to consider this application from both the perspective of technological/institutional feasability, and also from the point of view of moral and economic justifiability. The idea is inspired by one of Peter Singer’s essays: Famine, Affluence, and Morality. For those who aren’t familiar with Peter Singer, he’s a controversial Australian philosopher who has written on several topics including abortion, animal rights and charity.
In the essay Famine, Affluence, and Morality, Singer argues that (in very simple terms):
- there is immense suffering that exists in the world that can be partially alleviated through charity
- we have a moral duty to alleviate suffering when we can.
- there is an abundance of wealth in the developed world, and most people have substantial expendable income
- given 1, 2, and 3, people with expendable income have a moral duty to alleviate suffering in the developing world by donating some portion of their salaries to charity.
It’s a pretty basic argument, and Singer himself claims he donates 25% of his salary to charity (Unicef and Oxfam) [Wikipedia]. The core concept is that the lives of those receiving the 5% of my salary that I donate, would enjoy that 5% far more than I would (this drawing on the idea of decreasing marginal benefit of income). Now, I’m not sure exactly the form in which Singer donates his percentage (e.g. a lump sum he saves at the end of the year, or what?). But, it is easy to imagine building an application that automatically apportions a percentage of any money deposited into your account towards a charity of your choice. Think of it as a feature that could be added into Mint. The siphoning off of the funds could take place whenever money is deposited into the account. This type of application would be based on a subscription charity model: choose to donate to charity once, and be committed until you turn it off type of application. And, by making it a % model, rather than your standard 20$ a month model, this type of application also has some notion of each person donating in proportion to their ability to do so.
In my mind there are two appealing aspects to this application (at first glance anyways):
- Being automated. You don’t have to fatigue yourself with decisions about how much to donate, and when.
- Being proportional to ability. Each person donates a percentage of their income, rather than a fixed sum (there are analogies to taxation here).
So, what do you think about this type of application? Would you be up for donating some percentage of your income to charity? Why or Why Not?
There are lots of different questions that emerge from this thought experiment. Here are some of the different things I’ve been thinking about:
- What are the differences between donating some percentage of your income to charity, versus just paying more taxes to the government instead? Are charities more efficient than government organizations?
- Would we actually be doing the ‘right’ thing by donating our money to charity? Is this the best use of our funds, or would the economy grow faster if we didn’t do this, and in turn potentially benefit everyone more in the long run in different ways?
- To what extent should we focus on ‘growing’ the economic pie, rather than equally dividing it? And, how do varying ‘divisions’ of the economic pie in turn affect growth rates of the economy, and overall happiness?
- Do we really have a duty to alleviate the suffering of others whenever we can? How does this interplay with our duty to bring others happiness? (a recent post on Hacker News brought up the interesting question of whether its more worthwhile to work on projects that bring people happiness, versus working on projects that alleviate injustices [http://news.ycombinator.com/item?id=2882356]]).
- On the macro scale, are we measuring the success of an economy by its ability to maximize happines, or to minimize suffering? In reality it seems to be neither, it’s GDP we use - there’s a great speech by Robert Kennedy about this.
- How should we choose charities to which to donate given that there are so many different causes? And how do we hold these organizations accountable for how they use our funds?
- If we built this application, how many people would actually use it, and what percentage of their income would they donate?
Concluding Sidenote:
I’ve thought of another variation of this application for overcoming debt - i.e. some fraction of your income is set aside to pay off your debt, and we are able to project when you will have paid off your debts.
10 Amazing SkillShare Classes I Will Teach
For those of you who don’t know me that well, or sadly haven’t had the chance to meet me, I’m really talented and amazing at everything I do. Thankfully, now there’s a new startup here in New York called SkillShare that lets you share your skills with the world, and more people will get to be some small fraction as amazing as I am. Hopefully you’ll get a chance to attend at least one of these fantastic classes - let me know if there’s any classes you’d be particularly interested in having me teach.
Tastebreaking: How to Make Popular Things Unpopular
Remember things like Devil Sticks? Troll Dolls? The Macarena? and Friendster? Those are just some of the popular fads of the past that I had an integral role in crushing.
The 405 Essential Principles of Mime
I’ll be running through only the most basic of the mime principles that you need before you can perform at the competitive mime level.
Eating on a Budget: Preparing Your Own imaginary Food Like in the Movie Hook.
Save money, and burn calories by using your imagination! I’ll be showing you everything from how to prepare your own breakfast, to how to convince an entire dinner party that the food is delicious.
Sword-eating Your First Sword
There’s a first time for everything - and I’ve actually never persay eaten a sword, but I have thought about the technique substantially. We’ll be getting right into the sword eating so please bring your own sword, and don’t eat anything real 3 days before coming to this class (advise you take the imaginary food class mentioned above a week in advance).
Communicate complex thoughts to strangers within 100 meters - just by using your eyes.
If you were within a 100 meters of me you’d know a lot more about this class already.
Defenestration
Partake in the joys of the age old art of throwing someone out of a window
Being arrogant without being arrogant
This class is intended specifically for people who know they’re the best.
Have you been watering a fake plant?
Five tell-tale signs your plant is fake.
Farm Faster: Taking Your Farmville skills to the next level.
The latest in keyboard shortcut technology, hand postures, multi-monitor setups, and crop distribution strategies
#10: Any Other ideas?
I ran out of web appropriate ideas :|
8500 Startups vs Skype
The recent Microsoft purchase of Skype for 8.5 billion dollars left a lot of people unhappy, and calling for Ballmer’s resignation as CEO. So, what could Microsoft have done instead of buying Skype? Well, what would happen if Microsoft invested 1 million dollars in 8500 startups, and gave them access to the distribution channels, expertise and technical resources that Microsoft has? It’s an interesting thought experiment!
Not too long ago a close friend and mentor of our company suggested this potential strategy for Microsoft over breakfast. That comment he made in passing really did make us all think - why the hell isn’t Microsoft investing small sums of money in a large number of startups as their way to establish a pipeline of innovation? They would be playing some pretty good odds of developing multiple hundred million, and billion dollar companies! Especially when considering the network effects that would be gained! So - why isn’t Microsoft Yuri Mcluring it?
Sure Microsoft has their BizSpark program for startups, which tries to connect them with Mircosoft technology, but the offering isn’t overwhelmingly compelling for startups, and they don’t invest in the companies! As a result Microsoft is not doing a great job garnering the mindshare of startups, and this is dangerous for the company’s long term well being. In order to retain that mindshare one thing they could try is to make a TechStars, 500 startups or Y Combinator style startup incubation program where they invest small seed/Series A stage investments in an incredibly large number of startups that fit within Microsoft Research’s vision of the future. Even if the immediate monetary returns were not phenomenal - they’d have mindshare, access to talented developers, and ideas!
It’s true that M$ currently has limited expertise in the realm of startup incubation - being in a large company is VERY different from being in a startup. But, with a couple of pilot programs they could quickly develop the expertise and framework needed, and they would essentially be pumping out 8500 miniature Microsoft Research groups! They could set broad themes in terms of the types of startups they’re looking to fund (e.g. social enterprise, teleconferencing, database technology), and make it ridiculously compelling for entrepreneurs to participate. Some of the benefits they could offer include:
- Easier distribution and partnerships with Microsoft’s existing user base… which…. is a ridiculously large user base.
- Free access to Microsoft technology (There’s a reason so few startups use M$ technology. It’s definitely more heavy duty than what most startups need. But, if more of them did use it, it would put pressure on M$ to improve their underlying technology, and there would be a lot more people who know how to develop products using M$ technology.
- inside knowledge of what Microsoft is doing next, and how they can build on top of unreleased technology (this is similar to what Apple does with their app developer ecosystem)
- Ready connection with Microsoft for M&A purposes - i.e. they have a high likelihood of being bought by Microsoft, but they could also pursue other buyers.
- Access to experts who can help the startups with areas that are not their core competencies: e.g. technology experts, marketing experts, sales experts. Who can do regular reviews of the startups and offer advice (essentially Microsoft advisory boards).
- Free office space? Imagine 8500 startups working in close proximity, all near Microsoft’s headquarters to keep their core staff constantly energized, and connected to the cutting edge of technology (Redmond can feel like a bubble at times). The importance of being in close physical proximity, and how that impacts innovation, and execution, should not be underestimated!
- Free legal services up to some amount.
- Help finding additional financing from other sources - I’m sure M$ could put together a strong network of investors, and become a prominent player in the deal flow.
- A free zune.
The list of possible benefits could go on (would love to hear what you think they could offer). I’d certainly seriously consider participating in that program if it was available. Now, all this thinking is at a very high level, and misses out questions like what would happen in terms of follow on investment, etc. But, that stuff could be figured out - and when I imagine what the potential benefits to Microsoft of doing something like this could be - the question resurfaces: why isn’t Microsoft doing this?
10 Ways Tech is Changing Education (Part 2)
This post starts where I left off last week: covering the 10 most important ways I think tech can impact education. I’ve been trying to identify these opportunities, and also brainstorm what types of startups can emerge in each of the areas.
Last week, I covered the first 4:
1. Online courses, video lessons, and lesson plans.
2. Curation
3. Connections
4. Digital Textbooks and Visualization
Given how long each of these are, I’ve decided to split it up into 3 parts to make them more digestible. Now, it’s time for numbers 5 and 6:
5. Online Problem Solving
In my time at McGill University, a large number of the problems we had to solve in math classes were online. This was limited primarily to math classes, and the tools for online homework were at a primitive stage three years ago. I enjoyed the idea of doing my homework online, but not the reality of it. Now, it may be that these tools have improved since I left school, but I think in general online problem solving is an area that is ripe for innovation.
I remember solving countless problems in the backs of my textbooks in preparation for my exams, only to not know at the end whether I had done them right or not. My solutions were scattered across hundreds of scrap pages, that I had a tough time organizing. Many of the textbooks did not even include answers to the problems - and if they did they were usually one line answers.
The solutions in the book were never walkthroughs of the analytic process that goes into solving that type of problem - they were just a numerical solution. Perhaps I was getting the right answer at the end, but I was missing out on more elegant, and effective ways to solve the problem. I was missing out on how to solve the problem. Beyond that, I always found that the problems we had in our textbooks, were far simpler than the problems we were given in our exams: I always wanted access to harder problems to better prepare for exams.
For these reasons, I think there is a need for a larger repository of problems to be available, with nice walkthrough solutions (maybe I can ask for help if I get stuck along the way, and then get a hint?), and ways to input the solutions. I would also like to be able to track how I am doing as I solve these problems. I want to be able to see where I am weak, and where I stand in terms of the problems I can solve. This concept is related to the next areas I will cover: progress tracking, gameficiation, and personalization.
A good approach to innovating in the online problem solving space could just be a giant database of problems, accessible in the cloud, that have nice solutions, and nice ways to work through them online. I am guessing many students would pay for this (or perhaps the educational institutions would pay for licenses). I would like to know which problems are similar to each other, have multiple different solutions to the same problem if they’re available so I can follow different thought processes. Now, this doesn’t apply only to the sciences - in the arts there can be longer written answers (think Quora style answers), that illustrate what makes up a good answer to a particular question. But the entire question of how to handle the different subjects, and their particular ‘types’ of problems, is open to a lot of exploration and experimentation.
Both the types of problems available online, and the ways that people input and work through their solutions, are ripe for innovation.
6. Progress Tracking
Education is all about progress. It’s all about taking what you know to the next level. Sustained exercise at the same level of intensity does not translate into growth - we need to push ourselves. In that respect, there’s many elements of gamefication that are pertinent to education, and I’ll touch on some of them in the next post. But, even beyond gamefication - just being able to know what you’ve learned, and what you have yet to learn, is a very valuable tool.
The university and highschool systems allow us a very cursory level of progress tracking: you’ve passed these classes, and you still have to take these other classes. But, the lack of granularity in progress tracking is a missed opportunity.
Students should be able to track how all their homework is progressing, how difficult they found lessons/materials, what they plan on learning next, and more. All this information combined can give insight for students and the teachers to easily identify which areas they are weaker in. Now, this progress tracking can be largely automated, through the online problem solving and online lesson systems that will likely come into place - if students watch lessons and do the homework, the system can update the student’s progress with that information. This lets students all go at their own pace - which is powerful, and it also gives them a nice history of accomplishment they can refer to later on.
I believe that this type of granular progress tracking is far more encouraging to students than one-off, checkpoint exams. It can let them see how all the things they are learning are connected to each other - giving them a sense of purpose. It can also let them feel like they are doing ‘sustained’ learning, rather than short bursts of preparations for exams, followed by brain dumps of all they learned immediately after.
— Stay tuned for next week’s post :).
Subsidizing The Internet
I’ve been thinking a lot about whether or not there will be a shift in the way consumer content companies will make money on the Internet in the future. What services are a large number of consumers actually paying for on the Internet - Netflix, Pandora, Evernote, Rental Listings, maybe some task management tools, games? What else is there really? It seems most people are unwilling to pay for ‘content’ alone on the Internet. However, now there are new challenges to the idea of not paying for content such as the New York Times pay wall, The Daily, Arc 90’s Readability and more. The question is, given the ruling paradigm of monetizing consumer content companies through advertising - will this paradigm be uprooted?
Most of the web content that we enjoy on a daily basis is supported entirely through ad-based revenues. Just to give you an idea, click through rates on the Internet average around 0.2 to 0.3 % (Wikipedia factuals) - i.e. about 1 person clicking an add for every 499 who aren’t. Now, when I first started thinking about this, my initial thought was that I get to browse the web for free, and it has all this amazing content, all because some other group of people are clicking on ads to subsidize it for me.
But, thinking about it a bit more - clicking on ads may not be all that important. That’s in part why often people pay for ads based on CPM ( Cost Per Mille is advertising you buy based on 1000s of ad impressions shown, rather than per click). The issue is that advertising has a profound effect on people that can not be captured in click-through-rates, and actions after clicks. After all, television ads don’t receive my click feedback - but the advertisers do have the knowledge that the people watching the ads are less likely to ignore them than internet banners on the sides of pages. Television is a good tool for advertisers, because it forces us to watch the ads, whereas on the Internet many of us have visual-blinders for ads, automatically removing them from our conscious interaction with a page. But, even if the ads are only on the side of the page, and I don’t explicitly click on them, I’m guessing that doesn’t matter too much; the effects of advertising are almost always of a second order. For example, I know lots of brand names that I wouldn’t know otherwise, simply because I briefly glanced at their name on some page’s sidebar. An advertisement or endorsement on the page of a person for whom I have considerable respect is likely to impact me more than that same ad on some random page I stumbled on - even though I may not click on the ad in either case. Another example is the Old Spice viral video campaign on Youtube which I saw a while back. Now, I was already using Old Spice - granted. And, I did not buy anything immediately after seeing the advertisements. But, that video did substantially strengthen my brand loyalty (I enjoyed the brand even more, which made the product even better). The issue is that the impact of that ad is difficult, if not impossible to measure. It is so directly related to perception, that nothing can really gauge the true scope of the effect that it made on me.
Advertising’s effects are often immesurable just because they are so pervading - it has the potential to sway me from not buying something, to buying it continously for several years of my life. Advertising is the most powerful instrument any company has in its arsenal, at least when the term advertising is broadly understood. I can think of no other thing that can take something that has some market value X, and then double it’s value, or multiply it’s value by orders of magnitude, without making any ‘physical’, ‘objective’ changes to the actual product. From the perspective of amount of ‘effort’ put in, I suspect the returns from good advertising make it the best investment possible. It’s true that some things, e.g. an IPad, are obviously simpler to advertise than other things, e.g. a bag of pebbles - but examples like bottled water, and the wonders advertising has done there, suggest that even a bag of pebbles could be worth a considerable amount soon. Image is everything in the world of products. Knockoffs are the perfect example of this. While they are lower quality, even if they were of comprable quality, the issue would still be that the image attached to being an original is lost with knockoffs. Value, and the economy, like all things, are virtual.
So what’s the point of all this? Well, I guess the point is that natively human preference, demand and advertising are absurdly murky domains. Like Hume said:
‘Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.’
Though at times it may seem like the people who are clicking on ads are supporting the monetization of most of the ‘content’ on the Internet - they’re actually not doing this alone. I too am playing my part by partaking in the subconscious manipulation of preferences (my own). And, all that’s happening, is that whoever is making the decisions to invest in further advertising, is just admitting their blindness, and naieveity, and taking a leap of faith on that unknowable value of changing perceptions. That leap of faith however - is one I would take as well, and I do believe it is well warranted. The things that are most worth measuring, can’t be measured. Maybe if we could develop better methodologies for quantifying all the second order effects of advertising, we wouldn’t be so eager to pursue more clearly measurable, and hence more clearly monetizable approches such as pay walls.
