#dpi662

Apple has to care more about the unintended consequences of App Store guidelines. 

The current App Store guidelines threaten the fundamental principle of the ecosystem: they create unequal conditions for apps to be able to change the world and innovate like never before. The current policies and their enforcement make the app review process more challenging for first-time developers and yet are still not 100% efficient in preventing clone and spam apps from spreading across the App Store.

According to the current version of the rule 4.2.6 on banning template-based apps:

and article 4.3. on banning spam:

GUIDELINES VS. IMPLEMENTATION:

What’s wrong with the current implementation of the policy, and what should Apple do about it?

Firstly, Apple should overcome the lack of transparency over the implementation of the guidelines. For example, it is not acceptable, and it goes against the philosophy of the company to allow certain businesses, such as for example IBM, to develop template-based apps for its clients and not being impacted by the new rules.

Apple should be hesitant to implement one-on-one individual implementation tweaks and instead push for more efficient feedback and communication with all stakeholders.  

Apple should also focus on improving the capacity of the human factor in the app review. It should avoid situations when the same developer faces a severe lack of consistency in the rule implementation process.

Additionally, Apple could also change the algorithm that lists apps based on categories and disadvantage “clone” but beneficial apps by making it more difficult for users to find them on the App Store, for example, by having to type in exactly matching names.

What’s wrong with the current policy, and how to solve it?

In order to clean the Apple Store while avoiding killing app-creation tools that allow organizations to have their own apps without developing a unique app in-house, Apple should also make adjustments to the wording of the guidelines. 

New wording should reflect the fact that not all white-label apps should be put into one group, and the process of designing the new rules should take into consideration voices from key stakeholders (developers, representatives of both large corporations and small businesses, non-profits, governments). In particular, Apple should include the opinions of the municipal, state, or federal government bodies, especially in the context of the rising concerns about the lack of regulation of big tech companies. 

Apple should consider extending the 4.2.6 section by also accepting apps based on “unique content and services,” in a similar way as Google Play does. Apple should also focus more on clones that are still present at the App Store despite the change in the guidelines, for example by focusing on companies that acquire their competitors and end up changing their apps into duplicated versions of the main app, which was the case for example for Mail Pix, Inc.

To sum up: 

To reach the goal of further improving user experience, enhancing safety and trust, Apple needs both to update the wording of article 4.2.6 of the App Store guidelines as well as focus on enhancing the implementation of the policy. Apple should focus more on targeting the quality of the app, instead of the way it’s generated.  

word count: 509 

Sources: 

⁃            Kathryn Kosmides, Medium Blog,  https://medium.com/@m_64658/did-apple-just-crush-white-label-apps-4aee14d00b78

⁃            Dan Strutt, Medium Blog, January 10, 2018,  https://medium.com/@danstrutt/apple-updates-guidelines-for-white-label-apps-2018-80d54afdf1fd Apple Updates Guidelines for White Label Apps 2018

⁃             Sarah Perez - Tech Crunch - Apple’s widened ban on templated apps is wiping small businesses from the App Store - https://techcrunch.com/2017/12/08/apples-widened-ban-on-templated-apps-is-wiping-small-businesses-from-the-app-store/

⁃            Apple revises its controversial guidelines on template-based apps - Sarah Perez - December 20, 2017 - https://techcrunch.com/2017/12/20/apple-revises-its-controversial-guidelines-on-template-based-apps/

⁃            App Store Review Guidelines - https://developer.apple.com/app-store/review/guidelines/#minimum-functionality.

⁃            Tech Crunch - Apple removes VoIP app clones from the App Store, February 27, 2019  https://techcrunch.com/2019/02/27/apple-removes-voip-app-clones-from-the-app-store/

⁃            Medium - Apple & White label apps https://medium.com/@andyferrett/apple-white-label-apps-brightec-brighton-uk-60969a7fae8e Andy Ferrett

⁃            TechCrunch - Google follows in Apple’s footsteps by cleaning up its Play Store - https://techcrunch.com/2018/07/27/google-follows-in-apples-footsteps-by-cleaning-up-its-play-store/. Sarah Perez - July 27, 2018

⁃            Apple’s Control over the App Store is No Longer Sustainable, 10/21/2019, https://techcrunch.com/2019/10/21/apples-control-over-the-app-store-is-no-longer-sustainable/.

⁃            CNBC - Inside Apple’s tam that greenlights iPhone apps for the App Store  https://www.cnbc.com/2019/06/21/how-apples-app-review-process-for-the-app-store-works.html.

⁃            The top 1% of app store publishers drive 80% of new downloads: https://techcrunch.com/2019/11/21/the-top-1-of-app-store-publishers-drive-80-of-new-downloads/.  2019

⁃            Apple launched a dedicated mobile app for its developer community - Apple today is introducing a new resource for the over 23 million registered members of its developer community, with the launch of a dedicated Apple Developer mobile app.  Tech Crunch - November 18, 2019 https://techcrunch.com/2019/11/18/apple-launches-a-dedicated-mobile-app-for-its-developer-community/.

For the purpose of this blog, I assume it’s in the interest of Indian PM Modi to strengthen democratic institutions and the protection of human rights in India. Also, I intentionally narrow down the blog post on the case of WhatsApp (due to the word limit).

Overview of the situation and underlying facts:

Internet in India

In the next three years, over 762 million people will have access to the internet in India. It is the biggest market for Facebook and WhatsApp, with 350 million and over 400 million users, respectively, and a crucial player in influencing the future of regulation of tech companies.

Communal violence

Despite the rise in violence between 2015-2017, the number of incidents was larger in 2008, 2009 and, 2013. Therefore, we are not sure to what extent the higher levels of communication-technology penetration has led to higher levels of violence (it’s also important to emphasize that the government doesn’t publish the complete data related to religious-based hate crimes.)  

Source of the chart: https://www.firstpost.com/india/communal-violence-rose-by-28-from-2014-to-2017-but-2008-remains-year-of-highest-instances-of-religious-violence-4342951.html

Stakeholders:

Keep in mind the end-users of social media, minorities disproportionally affected by communal violence (Muslim communities), civil society groups, tech companies, and regulators (us).

The problem we know how to solve might not be the problem we need to solve:

WhatsApp has become a vehicle for spreading misinformation that, in some cases, led to violence and deaths. As the service is encrypted, it is hard to get information about malicious behavior. How to mitigate the negative consequences without increasing government and/or tech surveillance and alleviate the impact of the current crisis?

DO’s in the short run:  WhatsApp virality regulation and public campaign

On WhatsApp, 90% of all messages are between only two users. Still, the government should temporarily limit the number of messages that can be forwarded and push the company to strengthen its detection system to identify accounts that send messages from automated sources. To temporarily limit virality, WhatsApp could also develop a warning after pressing the “Forward” button, reminding the users of the current misinformation being spread.  

The government should invest in a public campaign similar to the one conducted by WhatsApp in 2018 and get the support of institutions that have the higher trust of Indian citizens (such as the Army or the judiciary).  

Do’s in the long run: address underlying factors causing violence 

The government needs to address the underlying challenges: lack of law enforcement, corrupted police, low trust in the government, and slow justice process. It should invest in increasing news literacy and digital skills.

Don’ts: limit shutting down the internet and restricting encryption

The Indian government should limit the rising trend of shutting down the internet as the downsides of such measures are well demonstrated by the current situation in Kashmir.  

India should also reconsider pursuing the “Intermediary Guidelines,” the government should not limit WhatsApp’s encryption as it would negatively impact the internet privacy. Tech companies should not become the gatekeepers deciding what can be shared online. Lastly, the government should also not force the tech companies to connect the Facebook and WhatsApp accounts with Aadhaar system.

Conclusion:

The government should not use the current panic and public pressure to push for policies that could have a severe negative impact on dissenting and critical voices. I argue that given the lack of data, community violence should not be the main factor driving tech policy reforms in India and that the government should carefully consider any option that could (even unintentionally) crackdown on critical voices in the society. 

Sources:

First Draft NGO, Telling fact from fiction inside Kashmir’s internet blackout https://firstdraftnews.org/latest/telling-fact-from-fiction-inside-kashmirs-internet-blackout/ 

Human Rights Watch: Violent Cow Protection in India https://www.hrw.org/report/2019/02/18/violent-cow-protection-india/vigilante-groups-attack-minorities 

Los Angeles Times, How WhatsApp is battling misinformation in India, where “fake news is part of our culture”: https://www.latimes.com/world/la-fg-india-whatsapp-2019-story.html

The Economist, India’s election campaign is being fought in voters’ pockets https://www.economist.com/asia/2019/04/11/indias-election-campaign-is-being-fought-in-voters-pockets

Washington Post, On WhatsApp, fake news is fast — and can be fatal https://www.washingtonpost.com/business/economy/on-whatsapp-fake-news-is-fast--and-can-be-fatal/2018/07/23/a2dd7112-8ebf-11e8-bcd5-9d911c784c38_story.html

New York Times: In India, Release of Hate Crime Data Depends on Who the Haters Are https://www.nytimes.com/2019/10/24/world/asia/india-modi-hindu-violence.html

Financial Times: WhatsApp says Indian rules on encryption ‘not possible’ to meet https://www.ft.com/content/9fcfa604-2a0d-11e9-88a4-c32129756dd8

The Hindu: WhatsApp says Indian rules on encryption ‘not possible’ to meet https://www.thehindu.com/news/national/political-parties-most-distrusted-army-judiciary-win-peoples-trust-says-study/article26647338.ece

Washington Post: Reports of hate crimes Cases Have Spiked in India https://www.washingtonpost.com/graphics/2018/world/reports-of-hate-crime-cases-have-spiked-in-india/

TechCrunch: WhatsApp hits India’s Jio feature phones amidst fake news violence https://techcrunch.com/2018/09/10/whatsapp-for-jiophone/  

WhatsApp suggests a cure for virality - The Economist - https://www.economist.com/leaders/2018/07/26/whatsapp-suggests-a-cure-for-virality

The Czech government is already behind several European countries in the adoption of the "ask once" policy, and currently, a proposal for legislation that would enforce it is debated in the Parliament. In this blog, I would like to pose several questions related to user experience, accessibility, but mostly focused on privacy and security challenges. I find these questions crucial to sufficiently answer before considering the policy implementation.

Challenge #1: Trust, Privacy and User Experience  

Is the government able to ensure that citizens will have enough awareness about what kind of data they share with the authorities? How to formulate consent so that it's not too broad or too specific? And similarly, how to ensure that the system is not offensive and only collects the information it needs? How many purposes should every collected info fulfill?

Challenge #2: Accessibility & Usefulness vs. De-identification

What kind of data should the government provide for reuse? How to balance the generalization of the information with maintaining their value to draw conclusions? Are we able to avoid profiling of citizens from de-identified data while doing so?

Challenge #3: Sufficient Political Leadership for Implementation  

Is the Czech government committed to not only talk about the issue but invest significant resources in putting the principle into reality, including all additional measures that are necessary for its successful implementation (such as e-ID)? Does it have enough political leadership to push through the legacy network and people who profit from the status quo? It the government able to effectively enforce the principle across all levels of government from municipal, regional to national and cooperate on the European level?

Challenge #4: Flipside of Government Analytics

Are we able to eliminate the risk that insiders will misuse the system? What measures can we take to prevent a situation when public officers would abuse their authority to access the shared database? Are citizens able to control their data? Is the system ready to protect the citizens in case if the political leadership has extreme views? How do we protect the most vulnerable groups in society from potential targeting? How does the current political situation in Estonia test the principle, and how do we put in place similar checks and balances that would protect it from potential misuse?  

Challenge #5: (Legal) Data Protection

Who would be responsible for sharing the data and maintaining the servers? How to prevent vendor lock-in or potential security breaches? What's the cost of data protection for these databases? Is the legislative proposal aligned with the European General Data Protection Regulation?

Conclusion:

My understanding of the political situation in the Czech Republic makes me believe that the government should currently not attempt to adopt the "ask once" policy as it is not able to sufficiently answer many of the questions. I believe that as of now, the risks of poor implementation, especially in the privacy and security realm, are too high.

(483 words) 

Sources: European Data Protection Supervisor, Opinion 8/2018 On The Proposal For A Regulation Establishing A Single Digital Gateway And The “Once-Only” Principle, August 2017,  https://edps.europa.eu/sites/edp/files/publication/17-08-01_sdg_opinion_en.pdfDeliverable 

4.1: Gap Analysis Report Of Challenges, Needs And Benefits Of The Once-Only Principle Submitted, Kess Meyer, July 2018, https://scoop4c.eu/blog/deliverable-41-gap-analysis-report-challenges-needs-and-benefits-once-only-principle-submitted 

EDRi Report, A Truly Digital Single Market?,  June 2015 https://edri.org/files/DSM_Analysis_EDRi_20150617.pdf

In this blog post, I'm going to focus on ways how to improve the HKS course registration interface. I will briefly describe seven things that work well from a design thinking perspective and nine "rooms for improvement," including hypotheses I have regarding what could be changed and why.

Let's begin on a positive note (please, see picture below). The interface is relatively easy to understand, even for new students, which is excellent. It provides you with most of the necessary information about the given course quickly (no. 1), and it has a decent graphic design that is pleasant to interact with. The interface doesn't overburden you with too many details (no. 2). Instead, it provides you with hyperlinks to other pages where you can go if you want to know more. It is considerate of non-HKS students because it includes information on cross-registration (no. 7). Also, it contains a detailed map of the Harvard campus that helps students decide faster whether we can make it on time from a different class (no. 6). By enabling the users to save a course into favorites (no. 5), it makes it easy to make non-binding choices and come back to pre-selected courses before making a final decision. 

So far, so good! 

That being said, there is also a lot of room for improvement (see visual demonstration below). When you are registering for courses, you need to know several things. Will the class clash with the others? How did other students like it in the past? Are there any prerequisites? Will the class count as part of my specialization? Any many others. During course registration, the HKS Services could relieve students' pain by limiting the wasted time we spent on switching between many different spreadsheets, all of which include indispensable information for course registration. Specifically, I believe it should be easy to include information regarding shopping days schedule (see no. 1), course evaluation (no. 2), dates of mid-term and final exams (no. 3), course's specialization eligibility (no. 6) or bidding history (no. 8). More efficient data usage could enable students to advance and improve their course selection. Consequently, we could discover courses we might not have been aware of before (no. 5). The interface could also warn us about potential schedule conflicts (no. 4) and add hyperlinks to several key terms that are not very straight-forward, especially for non-HKS students (no. 9). 

I believe that all of these changes are desirable because they would save students' time and improve our user experience and decision-making. On top of that, they should be relatively easy and low-cost to implement, since HKS already has most of the information at its disposal. 

Hmm. You might be asking why hasn't that happened yet? I believe it's because the HKS Student Services might not regularly collect feedback on students' experience with the interface. 

This is a great opportunity to change that! 

I propose to get inspired by Finland and focus on increasing flexible working practices and the proportion of full-time employees who can decide when and where they work. In doing so, I assume that if more people work remotely, it could significantly decrease the number of daily commuters coming to Boston. Therefore we could reduce the amount of traffic coming to the city. The proposal focuses on people who use cars to commute to work. Our goal is to see if flexible working hours are possible and attractive enough to change employees' behavior and make them use cars less often.

To assist with analysis, draft, implementation, and roll-out of the solution, I propose to partner up with the Harvard Kennedy School and get second-year Master in Public Policy candidates involved. We should give preference to those who took David Eaves' class on Digital Government. The students would oversee the management and operations of the test, which would significantly reduce labor costs during the trial period.  

Figuring out the root causes of the Boston traffic challenge

We should begin by examining the structure of Boston traffic in more detail by using data provided by the Massachusetts government. We want to see the detailed classification of traffic to find out the % of all personal vehicles that are used by individual people for purposes other than ride-sharing. Once we find out % of individuals and the frequency of using their cars to commute to Boston, we can proceed with our analysis and focus on the reasons why people drive to Boston. If we find out that one of the main reasons why people drive is to get to work or if we don't find enough evidence to prove or disprove it, we will continue with our test. If not, we will shift the scope of the study and focus on the most pressing issue instead.

Do people mostly drive to get to work? And do they have to?

Let's assume that one of the most significant factors in people's decision to drive to Boston is to get to work. Then we can do brief desk-research to find out who are the biggest employers in Boston by industry and size. Next, we need to test our other assumption that by giving people more flexibility regarding when and where they work, they actually reduce the frequency of driving to Boston.

To do that, we should conduct a survey about what kind of workers and from which industries are most likely to drive to downtown Boston daily. We would also like to know whether there is a correlation between low flexibility of working hours and the percentage of employees who drive to work. If our assumptions hold, then we would proceed and implement more flexible working hours in a representative sample of employers. If not, we would have to change the scope of our solution.

Drafting the solution

We would pick and approach a representative sample of employers in the Boston area to cooperate with us and pilot the project. We would ask them to provide us with two groups of employees which we could monitor for eight weeks. One of the groups would get absolute flexibility in terms of time and space, and the other one would work under the same conditions as before. It would be crucial to maintain communication with the "users" of our new flexible approach. We would want to make sure we get continuous feedback from them via short online surveys and weekly track the changes in their behavior. We would do the same with the other team that wasn't provided with any additional flexibility to see the difference. To increase the motivation of getting employers on board, we could also offer to track the impact on worker's results and work efficiency.

Implementation – test and evaluate every week  

We would install low-cost applications in phones of the employees (such as  Waze) to help them track their cars for the upcoming eight weeks. We would also monitor other alternative means of transport to see whether there was an actual change in the patterns of their daily commuting habits. Throughout the eight weeks of the trial period, we would analyze and compare the data of those two groups to see if our policy had any impact on the behavior of the employees.

Our goal would be to see whether the collected data validate our hypothesis. In other words, if we provide more people with flexible working hours, will they prefer not to commute to Boston and therefore limit the frequency of using a car? We would also want to understand the reasons why people who got more flexibility still travel to central Boston to come up with an appropriate policy to complement the increase in flexibility. For example, they might still have to drive to Boston to drop their children to school. How do we solve that? 

Is our solution effective in tackling the problem? If so, how to scale it?

In case the data confirm our hypothesis, we could design a more sophisticated policy that would focus on how to make more companies flexible about their working hours. We could do that by reaching out to city governments that already implemented a similar solution in the past and learn from their mistakes. The goal would be to figure out what are the main barriers for them to apply the change and what support the government can provide to help them overcome it. Then we would test these steps again before we would consider scaling up the implementation and actual roll-out of the program.  

Source of the picture: http://bikeyface.com/2012/08/23/boston-beware/ 

My interest and work in the field of Asian affairs provided me with the opportunity to live in India, China, and Myanmar. Each of these countries proved the importance and impact of technology on public policy and government in a different context.

In 2012, I volunteered at an Indian NGO called Janaagraha, which fostered active citizenship by reporting and tackling bribery, documenting corruption, and connecting citizens to deliver change. All of that by using modern technologies. With its work, Janaagraha demonstrated tremendous opportunities that technologies offer to improve the quality of life in and governance of cities.

Living in China showed me the power of digital government from a different angle. The level of censorship, surveillance, and oppression the government was able to build with its pro-active approach to technologies made me more aware of the necessity of having a more profound understanding of the challenges and risks for the public.

Lastly, Myanmar was fascinating for its lack of legacy network caused by decades of military rule during which the country was isolated from the rest of the world. Only recently, it started to leverage the opportunities stemming from this situation. The case of Myanmar is essential also for another reason. The failure of Facebook to prevent inciting violence against the Rohingya Muslim minority group in Myanmar demonstrated both the challenge of fighting hate speech as well as a global incapacity to regulate big tech companies and hold them responsible.

Since I returned to Europe in 2017, I experienced different professional settings. From NGOs, the European Parliament, to a political campaign. They were quite different but had one thing in common. I was always aware of the fact that there are opportunities to make our work more efficient by applying more technologies. But I often found ourselves in a situation well described by the picture Professor Eaves showed us during the first lecture (see below) when we lacked understanding of what are the current possibilities and what on the other hand, is impossible to do or request.  

I’m keen to learn about ways in which technologies can be used to increase trust in governments and make them better at delivering to their citizens. The Czech Republic is in the process of executing the Digital Czech Republic strategy, and I aim to be able to better evaluate this plan and its implementation. I also hope to deepen my understanding of contexts and issues related to EU-Asia relations. The rising influence of China in the field of technology is an excellent example, recently demonstrated by the debate regarding the implementation of 5G networks by Huawei.

To sum up, I’m taking this class to learn how to think about digital governments in its complexity, to understand the current state of affairs and deepen my understanding of the tradeoffs between security and accessibility. Hopefully, I will learn enough to leave behind the level of knowledge of technologies represented by thinking that cookies only stand for sweet round flat cakes.