Ethics, Virtues, and SmartCities Tech

Occasionally, at SmartAustin, we like to take a step back from the rapid pace of tech and startups, and to consider some of the ways in which tech is impacting our broader society. We've covered, for instance, the "Blockchain for Impact" and Estonia as a "Digital Republic". More recently, we've been thinking (as have many) a bit more about AI (top AI-focused book we've read is Kai-Fu Lee's AI Superpowers) and came across a quote from Shannon Vallor, a professor of philosophy at Santa Clara University and AI Ethicist/Visiting Researcher at Google. In Wired's 11/13/18 issue, Vallor is quoted as saying that "There are no independent machine values. Machine Values are human values."

After reading a bit more about Vallor's work, we were inspired to pick up her book, Technology and the Virtues and found it a delight. The book deserves a much more thorough treatment than what we'll give it here, but for the purposes of this blog, and thinking a bit about how startups launch themselves at technical problems, with maniacal focus, we'll call out a couple of quotes.

  • Quoting the philosopher Ortega y Gasset: "in the very root of his essence man finds himself call upon to be an engineer" but ". . . technology is . . . not the beginning of things. . . . it does not draw up [the human] project.

  • Vallor earlier quotes David Friedberg, "the CEO of a big data analytics firm . .  : 'What happens when every secret [is exposed] from who really did the work in the office, to sex, to who said what is that we get a more truthful society. . . . Technology is the empowerment of more truth and fewer things taken on faith.'" . . . "Implicit in this statement," Vallor comments, "is the unquestioned privilege of truth over other moral values, including trust, respect, compassion, humility, and flexibility. Better to always have the 'truth' about whom on the team 'really did the work' than to permit the guy who was up all night soothing a colicky child to quietly slack off at work one afternoon without risk of exposure. . . ."

You can definitely argue that all of these high sentiments and moral qualms have very little to do with startups, or that startups can't afford themselves the luxury of thinking too hard about these kinds of things. Founders need to make a living, get their businesses started, and solve problems that other people agree exist. Usually these have to do with efficiency, speed, and customer experience, and not with “drawing up the human project” or with the propagating "compassion, humility, and flexibility." Fair enough.

Fortunately, Vallor's work (and we're sure there are others like her) goes beyond her excellent book, and includes practical tools for helping businesses integrate ethics into their work. The Markkula Center for Applied Ethics at Santa Clara University, where she is on the Steering Committee, includes a list of Best Ethical Practices in Technology, which can guide tech companies looking to instill a culture of ethics (more detail is available at the site):

“1. Keep Ethics in the Spotlight—and Out of the Compliance Box

2. Highlight the Human Lives and Interests behind the Technology

3. Consider Downstream (and Upstream and Lateral) Risks for Technologies

4. Don’t Discount Non-Technical Actors, Interests, and Expectations

5. Envision the Technical Ecosystem

6. Mind the Gap between User Expectations and Reality

7. Avoid Hype and Myths around Technology

8. Establish Chains of Ethical Responsibility and Accountability

9. Treat Technology as a Conditional Good

10. Practice Disaster Planning and Crisis Response

11. Promote the Values of Autonomy, Transparency, and Trustworthiness

12. Consider Disparate Interests, Resources, and Impacts

13. Design for Privacy and Security

14. Invite Diverse Stakeholder Input

15. Make Ethical Reflection & Practice Standard, Pervasive, Iterative, and Rewarding

16. Model and Advocate for Ethical Tech Practice"

Finally, the site offers a series of tools and frameworks to support applying ethics in technology. We'll end with a schematic (source here, screenshot below) that grabbed our attention, provided here for your reference:


We believe that companies that adopt rules and practices around to support ethics in technology will be at an advantage when it comes to recruiting and retention. A systemic approach may be extremely difficult or impossible in the very early survival stages but is well-worth considering as your company gains traction and begins to mature. We'll all benefit.

Protect your 'Smart' Data by Learning the "Language of the City"

Today’s post is the second in our “SmartSpeak” series, which provides a forum for SmartCities thought leaders to provide guest posts.  The “Austin Series”, a bimonthly feature, is graciously authored by thought leaders within our very own city government.  Today’s post is authored by Ted Lehr, City Data Architect.  It’s a detail-oriented post for those interested in supporting, or engaging in, City projects that leverage new data streams and archives, while shedding light on the mindset that it will take to develop the public-private partnerships necessary to achieve a “Smart City” vision.


SmartSpeak: The Austin Series

Protect your 'Smart' Data by Learning the "Language of the City"

How's this for a nightmare?  You've just spent the last year helping your community and city imagine, design, plan and commence implementation of a great smart city project only to learn that no one budgeted for archiving all that data the project is generating:  "We thought this was for operations?" you are told.  Then,word comes that, "We can only keep six month revolving windows of data, " and that the City will delete data once it ages beyond the window.

That primal scream you're roaring at these irregular and inopportune moments is wholly avoidable.  With foresight, you can institute new approaches to data that support the development of profound and corroborating insights,  behavioral models that lead to improved operations processes, and curated data that can support policy discussions.  But first you need to learn the Secret Language of the City.

Let’s break it down: There are four basic steps to ensuring that a city will maintain the right kind of data archives, and develop the right kind of analysis, to support a “Smart City” vision.  And the core principle that supports all of these steps is understanding a few words of “Cityspeak.”  That is, you need to understand how Cities define and use concepts like “assets”, “valuations”, “costs” and “projects”.   If you don't take the time to express the thing you care about in these terms, you will vastly increase the probability  that the organism we call a "City" will not even detect its existence.

And here are the steps:

  1. Make data an asset

  2. Assign a cost to the asset

  3. Designate, and plan to measure, the value returned for paying that cost

  4. Attach the asset, its cost, and, the appropriate metrics to an existing project

Now, let's consider each of these in order.

Data as an Asset

You can begin to make data exist in the eyes of the City by identifying the physical assets that generate, carry, use or store it. “Assets” as known by cities, and certainly for the City of Austin, are tangible things: signs, roads, computers, swing sets are all assets. But, data? You can't touch data! As a result, it will be difficult to assign it an asset code. The devices that generate or store the data, however, can be treated as assets. So sensors, networks and storage are assets. Even non-touchable cloud storage can be an asset because it has what are known as “physical equivalents”.

Assigning Costs to the Data Assets

The second step is to assign costs to data assets. This step is more straightforward. Costs can be expressed in terms like:  

  • dollars per Gigabyte of storage  

  • dollars per Gigabit per second of network demand

  • dollars per unit of data processing capacity

These costs can be tiered in such a way that helps us budget and evaluate projects more efficiently. For example, raw data from sensors on a city's streets might have one cost. Data combined and curated to produce neighborhood or corridor specific data might have another, larger cost.

Designating the Value Returned from the Data

We next need to answer the questions of, “Why is the city going to spend public dollars on these data assets?” and “What value will the community derive?”  Importantly, the valuation of the return on a city asset does not have to be expressed in financial terms.  For example “values” might include:

  • Reducing pedestrian injuries at intersections

  • Increasing or maintaining neighborhood satisfaction with their local parks

  • Being able to measure, understand, describe and eventually control greenhouse gas emissions

  • Understanding whether City parks and services are used equitably

It is best to describe such values in the context of stated city objectives.  These might be objectives like “reducing traffic fatalities,” “increasing access to affordable health care,” “developing walkable communities,” etc.

Now that you have defined the “value” of your data, make sure that, these valuations are measurable. You will need to define value metrics the City can monitor as well as associated target values, so that the city can assess the return it is getting. For example baselining and tracking pedestrian traffic injuries at a set of intersections with a goal of reducing them by 30% would constitute a value metric and goal.

Attaching the Data Assets to Projects

Congratulations!  Your City can now ‘see’ your assets. But it can’t pay for it yet!  The last thing you need to do is to attach your data assets to an existing, or planned and budgeted, project. Then, your data can live, and you can pursue your most imaginative data science dreams!  

Conclusion:  Don't forget to consider the Data Market

You’ve completed my course today. But there’s a bonus opportunity, for the ambitious: you might ponder developing a cost recovery plan for the data assets. Instead of drawing on traditional revenue like the tax base, perhaps there are fees or other charges you can suggest the City impose on the external use of the data. For example, raw data could be free to the community, while curated or analyzed data could incur a charge. Or, if a business would like to make the data mission critical and therefore requires SLAs on throughput and response time, you could propose a tiered pricing mechanism for these additional premium services.

These are just a few ideas that can help to make the foundational element of the Smart City of the future, data, better available, better curated, and better analyzed. We will need the creativity of thousands of entrepreneurs and officials to effectively leverage data and technology to serve urban residents. But, we cannot get there if we don’t attend to the sometimes boring work of learning to speak the City’s language.