Meeting Notes 2015-02-06 (@fcc_cio)

This document is CC BY 4.0

Related post: Cyber-Civic Endeavors in Taiwan

Location: Ex-Handlino Galactic Headquarters, Taipei

Participants: david, albert, tonyq, clkao, au, diane, joseph, jonna, nchild


這是由 Albert Tzeng 發起的聚會,與 @fcc_cio in a personal capacity as an Eisenhower Fellow 聊天以促進社群相互瞭解,秉持公開透明原則,製作全程文字紀錄並將 hackpad 放在 g0v 的 workspace 底下,供社群參與者參考。


david> The world is rapidly changing. Are there ways we can continue the checks and balances intentional to multi-party systems of democracy, and at the same time adapt with speed? Are there new models in which we can keep the checks and balances? One possibility is increasing public <-> private partnerships and other one is giving back to people roles what traditionally the government had to do because of the lack of internet  connectivity. We can empower the public to participate more in the work of government/public service. For example, at FCC we did a GitHub open source speedtest app; "crowd-sourcing" signal strength measurement by consenting participants with anonymization of test reporting, was #4 on the iOS app store for a while, right behind Google Chrome, a 1st for any government app.

david> Might be worth considering to develop a similar approach Earthquake / Tsunami scenario, giving the public the option of using their phones to ping for a specified period of time of their choosing to say "I’m here", yet at the same time what would be interesting would be those areas without signals, that would form a picture of where connectivity would need to be re-established. is also a good example of a platform using SMS for disaster response. 

david> We had a successful case with the development of and ZenDesk; SaaS is disruptive because it also puts existing on-premise IT contractors out of what they traditionally have done. We all need to adapt with evolving times. The ratio of IT government civil servants to contracts can be 20% to 80% in some places. At the same time SaaS saved the taxpayers about 85% of the costs vs. if it had been done on-premise instead. 

clkao> Procurement regulations in .tw is not very friendly to this kind of SaaS.

david> Same in .us ... yet if we are to be good leaders, someone has to assume the risk. Multi-party systems, while they are good at check and balance, is not working well to catch up with accelerating technology. At the same time, single-party systems has its traditional pitfalls; am not advocating for a dictatorship or oligarchy approach, just raising the concern about how can democracies keep up with the speed of technology change? 

joseph> Although there exists a certain new organization, we have a long history before we’re formed that dates back. We don’t have the "legacy systems" which has its own benefits and disadvantages. Here we have a problem of gathering and dissemination of personal information. Also especially changing the internal culture of the administration is a challenge.

david> Generally in any situation where you’ll do change: 1/3 early adopters, 1/3 neutral, 1/3 resisting the change. Need to form coalitions first: start piloting visible wins to bring the neutral "not sure yet if this change is good or not" people along. Do not try to do change as "separate enclaves" separate from the existing participants, as any separate effort will not last and doing change in an enclave alienates both groups. Have to do change even within and amongst those who do not fully believe the change is needed yet. Important to be a non-anxious presence. Good leaders demonstrate openness and calmness. Also need to remember that individuals almost always join public service because they cared, so if you can hoop back to why they joined in the first place, asking "what brought you here? what’s your motivation?" as well as "what brings you joy?" you can help bring them along.

There’s management and there’s leadership. Management is when you meet expectation. Leadership is when you step outside expectations — but doing either just management to leadership to the extreme too much scares/alienates people. So one has to learn to tune the dial appropriately. 

tonyq> Maybe 20% people in my experience inside the administration want to change things; they brought in collaboration from the community to try to win over the 80%. When we campaigned for Mayor Ko last year, we started with some showcases to convince people that this (collaborative) way is working.

Within [Minister Jaclyn Tsai’s] office I’m working on 4-5 new projects, and more outside the office  capacity. These are projects that deliberately strive to change the existing public<->government relationship, especially literacy and infrastructure projects around "Open Data".

We’re setting a new KPI, and a new culture that doesn’t demand perfectness out of government-released data; the usefulness of "release early and often" is what I’m trying to get across.

tonyq> De-identification process, we’re trying to get mature process in place, so more data can be made open as part of an automatic process, like electronic invoice record as indicators of price fluctuations of commodities, etc.

Also trying to get the 1999 (311) process more open in all levels of governments, by displaying questions-and-responses in a public forum after getting opt-in from participants.

david> Focus on winning over the neutral/uncertain group not completely sure by your new approach by coming up with infrastructures that saves them time. For every 4 efforts you do to help the public, also include 1 effort that helps civil servants too -- we’re all stretched thin and pressed for time. Such an effort can win over people within the existing system that strains their attention to address so many pressing needs. 

clkao> Figuring out their pain points…

david> Am excited about the future of public service. Perhaps one day: Siri-like personal assistants for the public services could connect people who are doing overlapping/related things and connect people with people. A combination of humans and machines paired together, each providing strengths. 

clkao> That was the short version of what we do, and a sample of our more interesting projects:


clkao> Unlike traditional "civic" groups, we have a more vibrant community that encourages the "release early & often" way for existing NGOs. We hold a major hackathon of ~120 people every couple month, there’s a minor weekend working-group hackathon of 30~50 people every weekend too — it’s organic, we have hackathons centered on laws , and design, etc. in just the last few weeks.

clkao> This year our main focus is collaboration with NGOs, even the largest NPO (the administration of .tw...), and the academia, etc. The idea is setting up a platform that’s neutral and efficient in getting people’s ideas across to each other.

The plan is for internet-related regulations, in which all regulations are deliberated on it a month before the official hearing process, then in parallel with it. If it works then we are thinking about extending that model to all regulation-making.

Because the project is open source and developed by community, there are other parties (constitution reform, political parties) interested in using this platform as well. Technically, it’s just a GitBook -> Discourse integration system, which should be quite portable for reuse.

david> Your efforts are impressive. I’d be happy to connect you with potential developers & users. 

david> Of note, sometimes I am concerned that increasing transparency, while good, can also sometimes make public individuals more risk-averse. We face a challenge of: How can we let people claim victories while retaining the flexibility to "compromise", since compromise requires you to sometimes let other goals come before your own so that others will equally compromise their goals to advance yours. If everything is transparent, you might lose the ability to make compromises and it can become very polarized. 

david> Also, a question worth asking: How do we educate people to be more accepting of people who make mistakes earlier in their career, instead of the risk-averse expectation. There’s a challenge of everything being reduced to sound-bites.

tonyq> One simple solution is that we elect someone who’s always making mistakes! Mistake-making then becomes the new norm.

Mayor Ko is making a lot of change in just a month. He makes mistakes very publicly, corrects himself many times a day in public, so we are starting to discuss everything in-context.

david> There’s a HBR article "In praise of the incomplete leader", the best kind of leader is one who opens themselves up and admit their blind spots. Bringing data to decision makers who point to blind spot.

au> presentation:


david> Great presentation, impressed that what you’ve developed is a confusion-reducing framework. Also that you were equally helping the anti-protest protesters to show that you’re not taking a specific stance, rather you’re being a facilitator of all views. Having that balanced approach is commendatory. Would like to connect you with the Next Generation of Government Service in the U.S.

david> On the topic of how to improve connectivity in areas where connectivity is reduced, maybe consider a Kickstarter campaign on rapid connecting satellite connection?

tonyq> I’m starting to rethink "what’s government" now, we try to mediate between the big-gov and the minimal-gov lines.

david> When I meet with folks, I intentionally reference "public service" because the internet has been blurring the lines between gov and non-gov; and maybe 15-20 years into the future maybe nation-states won’t be the only form of organizing.

tonyq> AirBnB or Uber, they claim they could organize in a more efficient way...

david> Back in U.S. history, Ben Franklin’s been said to have observed, after the signing of the U.S. Constutition -- that he could can rest well knowing the great American experiment will last another 50 years. Of course those 50 years have been surpassed, yet the most important element is the importance of "experiments" to gain expertise. We need to make sure there’s a safe place for public service to experiment with better approaches to engage and work together. 

albert> nchild is known for creating one-glance infographics, he does an "explanation for the lazy" (懶人包, "LazyPack") for in-context communication.

albert> I’m currently working on UDN Debate, probably the first site for deliberation on public, controversial topics. In UDN I also work with international journalism — I think the term is too narrow — we’re still in the process of defining it. My own field was in sociology of knowledge. The job market is quite terrible, reducing 170 universities to 100 — so I gradually shifted my field to media innovation. In this field, I don’t have much to say re Open Source / Open Data, but I’m an accidental observer in this field.

On the night of March 18, I was paying a visit to the parliament — where people were shouting "there’s a parkway, hold that fort" — then I accidentally became a occupier, and have given a few talks on that topic. In 2013 I’ve been aware of g0v... and got introduced to tonyq and other friends in the NECTW.

david> It’s good news because we do need experiments on how people organize. The IoT/Internet of Everything will likely change this landscape even more. 

albert> In the past year we see the gov trying to do cyber-participation, there’s another one last month called — although they did have several channels where they could solicit participation, the actual turnout is non-optimal. I think people still feel a general powerlessness over their opinions could hold much sway. I think holding a referendum after the participation is still key — because it attracts so much national attention — my main hypothesis is that the cognitive scarcity is the key factor. Sometimes tech people forget people in general have a very limited attention span.

So my main interest in civic society is how people can organize them into Interest Groups, who can dedicate more quality time into specific topics. After Sunflower a keyphrase is "公民好忙" (citizen is too busy.)

david> Our individual human attention span is finite, yet any member of the public should always have the opportunity to participate. For most modern representative systems of government -- we have opted as for the most part that you are not required to participate in order to be a citizen. Periodic Jury and Voting are really the two asks of the public. 

david> Per a professor at Harvard, there was one experiment some years back in Idaho, where they actually mailed out letters saying, whether you would like to spend two weeks preparing a brief for the state legislature. More than 80,000 people volunteered. A subset were randomly chosen to participate in preparing the brief for the legislature. They didn’t have to reach consensus, and it turns out the the state legislature said it was one of the best briefings they received. Similar to the development of the new Icelandic-constitution. So experimenting with a randomized, jury-like, ballot of Interest Group, is an interesting idea for the future? That said, you also need to be aware of the are risks of pseudonyms in cyber-participation, where people could set up fake accounts [sock puppets], etc...

tonyq> This year I believe people are getting the key point, which is a quicker turnaround from governments. It’s getting more rapid in emergency responses. For example one of the ministry of justice’s site were compromised, but this time they took only an hour to respond on social media before it escalated into a media crisis.

david> DARPA has made a open-source AI available. It’s early days, but I think once could involve the AI in the job application process and the review process of application screening, etc, could be automated eventually. Will be interesting to see what applications are developed

jonna> Q for the three speakers to senior public servants (tonyq clkao au): What would you suggest them do next?

au> I have no agenda...

david> I wonder if you can pair someone with them, a more prolonged fellowship-like relationship. Each mentoring and sharing insights with the other. 

tonyq> I’d like to create a space for public servants where developers can participate, and for both side to feel safe doing so. I get flak just for "joining the government", but I’m happy to provide cover — while surviving within the system — to other fellows.

clkao> Getting institutions and programs to fund such fellows, bringing outside experts into the agency, and triage the problems they’re facing. We are also looking to reformulate the issues themselves, encouraging them to bring their issues to our hackathons — my idea is that each agency should make public their #1 pain-point w.r.t. information systems every month, "open challenge" style, to invite public participation from civic hackers.

david> You could do this by sponsoring awards, recognizing agencies doing better cooperating with collaborative ways. What gets highlighted can encourage others to do more. 

albert> The underlying logic of g0v contradicts, especially with the legal system, the existing structures of the government and agencies. It’s very difficult. In Taiwan, our legal system is from .de and .jp, where everything is strict and clearly stated, as opposed to the case law systems.

david> Maybe I’m speaking too much as a programmer, it would be helpful to "defragment" the legal code — e.g. SaaS in the U.S. with auto-renew clauses can be interpreted as an "infinite contract" that makes procurement difficult. Need to find lawyers who are willing to identify the key (re-)interpretation one would need to introduce this into the administration.

au> Jaclyn has been working with IBM’s legal department for many years and is quite an expert in this regard.

albert> It’s also interesting that the three people here are all drop-outs of various degrees...

david> We do need to think about encouraging more people to join Public Service who do things differently, to not feel frustrated to the system. I see my biggest responsibility as a Senior Executive to provide the "top cover" so that others who think differently are encouraged to join, experiment, and help make a positive difference. Others earlier in my career did that for me and it inspires me to do the same now. 

albert> I have another meeting, got to go.

tonyq> I’d like to bring up the topic of Digital Divide — lots of people in Taiwan, who technically have internet (mobile) connectivity, are not yet full "endpoints" in the online space of discourse. When I was in Tainan, I talked with the city government, saying maybe you have 1000 service centers in your city, but with an online presence you can serve a lot more people.

clkao> Mobile internet definitely helped to expand outreach — evidently in the next decade or so we’ll put internet-enabled devices on the entire population. To me, more than outreach of connectivity, it’s more like care-taking for minorities...

david> At FCC, before the launch of the new Consumer Help Desk , we had 18 different paper-based forms. We went with online (508-compliant of course), we could streamline this by walking individuals through the nature of their concern. For those without internet access, we still provide telephone numbers where people can relay their concerns into the internet-based solution and human operators will route people through properly. 

tonyq> We have a PKI card system (MOICA Personal Certificate)...

david> In .us not sure if a PKI system is on the immediate horizon, it’s important to protect privacy. Much discussion will need to be had at local, state, and the national level before this. 

tonyq> For example to buy medicine or alcohol over the internet where one needs proof of being 18+... I’ve been thinking about this issue, but there’s no clear consensus yet.

david> A challenge is TCP/IP was not originally designed with security in mind, instead security was to be controlled through physical access to the network. For every solution you’re trying to do, risk that someone will be able to spoof it. Best you can do is reduce the risks of such vulnerabilities through consistent approaches. 

tonyq> Maybe we can have two- or three-factor authentications?

david> The way one sits, gesture-print, biometrics, etc. — can be identifying of course, but ensuring complete confidence can be hard — 100% certainty probably not possible. Agreed multi-factor helps to "buy down" the risk of a compromise. 

au> On rule-making: We’ve been experimenting with a topic-setting process and the discourse ("Civilized Discourse Construction Kit")-based community. For example, getting badges to ministries who participated in vTaiwan rulemaking ( ) worked well as a social incentive for public service people to participate — both as representatives, and sometimes with a separate individual account.

david> In a past role, tried to semantically analyze the verbs and nouns in one executive order, it eventually became a 2 meter x 2 meter poster. This mapping process also identifies potential gaps and redundancies in laws. Am hopeful we might develop better ways to visualize the law in the future, to help both the public and policymakers/decision-makers. 

au> The main objective of building this platform is getting a safe space for agency people to speak both as themselves and also as representatives. As we know, Facebook commenters can be cruel, and PTT (the .tw equivalent of Reddit) doubly so.

clkao> Maybe next week’s g0v underground podcast can have a screen-cast demo of vTaiwan in action!

david> Thank you all for a great meeting, I really enjoyed meeting each of you and your efforts are impressive. :sunny: 

au> Glad to work alongside you on this hackpad as well! :+1:  

joseph> off-record discussions on GFW, MAC, and the history of GW in China etc.