79 Matching Annotations
  1. Jun 2019
    1. After all ofour searching,we knew this wasthe perfectplace, and the perfect partner.

    Tags

    Annotators

  2. Feb 2019
    1. tech companies are not so much unilaterally controlling consumers as confusing them about what is under their control.
    2. data scientists found that measuring people’s expressed preferences was a dead end because these were too variable
    3. easier to change the entire infrastructure in which people’s preferences lived, and then, rather than assume people are capable of forming and expressing their tastes independently, you could treat captured behavioral data as revealed preferences
    4. Flow works by disguising its compulsory mechanism in the details of its content, which is nothing more than bait from the system’s perspective.
    5. both want the particular units of content and are indifferent to them. We are both active agents and passive objects.
    6. Coercion and persuasion, then, can’t be cleanly distinguished.
    7. When platforms today deploy recommendation algorithms, their purpose is to produce flow.
    8. “persuasive technology” that can condition prey “to play the role scripted for it in its design.”
    9. what’s being sold is surrender
    10. that users want to consume flow, not particular items of content
  3. Dec 2018
    1. Do not forget that automation will generalize and make employment less necessary. Look at the self-checkouts in supermarkets, automated tolls, but also the software robots that do housekeeping at Wikipedia. I argue that this is a good thing. On one condition: that it enhances the ability of people to develop their social skills, their knowledge, their work in the strict sense of the term, rather than solely their job.
  4. Jan 2018
    1. The self-described “creature of the suburbs” helped remake this city, in some ways, for his own maximum personal comfort.
    2. trust-us self-assurance in their own technocratic moral sense
    3. main thing Doctoroff admits the administration didn’t quite get right was figuring out how to keep New York from becoming too expensive for many of the people who already lived in it.
    4. Doctoroff has little use for “nostalgia,” practically an epithet in the book,
    5. Success is when “more customers want to come and stay. And the ultimate measure is population growth.”
    6. “You have to treat citizens and businesses like customers,”
  5. Dec 2017
  6. Nov 2017
  7. Oct 2017
    1. make data science and machine learning skill development for their staff a part of any ongoing work with Sidewalk Labs

      This is a really wonderful notion!

    2. in larger part than is being imagined now

      +1

    3. perhaps not in their entirety

      my feeling is definitely not in their entirety, and definitely not in their current form

    4. This fabricated story [that the government can’t do innovation]

      I feel this idea, that the story is fabricated, needs more support than is given. I want gov to be better, but do believe it's structurally less capable in its current form.

    5. understanding and incentive

      To play devil's advocate: I feel that many regular folks would disagree that they have the former "understanding". I think this argument needs support -- after all, many would say that they don't understand, because they don't have data collection and analysis capability, and so that's why cooperation with Sidewalk Labs is needed.

      I say this as someone deeply sceptical of our city's ability to navigate this new deal. My scepticism is rooted in the sense that most of the citizenry lacks their own appropriate scepticism.

    6. Data and technology can be helpful when they support, not lead, the city-building that city governments and planners know best.

      I wonder whether this is something that average citizens would agree with. I tend to feel that politicians and civil servants get lumped into the same pile as ineffectual and unreliable. No, it's not fair, but I feel it's how most people feel.

      So perhaps part of the work we need to do, as folks who believe tech/design can be of service to gov, is to help allow regular people to see government (politicians and civil servants) as more competent than Google. Until we do this, it will be very hard to convince people not to give authority to successful companies like google, no matter how reasonable it may be to exercise caution.

      So maybe "help gov be better with tech/design" is the imperative to prevent situations like this where citizens are starting to have more faith in tech companies to lead.

    7. urbansim
    8. panning
    1. Commons-based peer-production relations regularly exhibit three structuralattributes.
  8. Sep 2017
    1. On the other hand, if the structure fails to increase the inhabitant’s ability to dissipate energy, for example by requiring the inhabitant to spend more time on its repair than on obtaining food and resources (free energy), then the structure can be seen as participating in its own destruction (2nd Law).

      Really love this flip of perspective to focus on the intention of the mud hut and it's needs of the more complex "volatile structure" (ie. the inhabitant).

      Makes me wonder what it is about the relationship that makes us as humans feel dominant. Might it simply be that we sense that we have a more encompassing "theory of mind" for the mud hut, than it has of us. (Perhaps the better term would be "theory of structure".)

      So then what of corporations? They seem to operate at a scale above us, and be composed of us, but they also contain our stupid AI enmeshments (distinct from upcoming general AIs). And these corporate structures don't seem to have the same sort of "theory of structure" for understanding humans, in the way that we have of them. Perhaps they are more like the wind or the mud huts in their relation to us, but great lumbering beasts at a scale above us, entangled with the primitive automation we've built so far.

      So what relevance might this angle have in consideration of future AI? Is the process that will unfold that AI will develop a better "theory of structure" for us, and our role will be less the "inhabitant" and more like the mud hut. And then, what will happen as the AI's needs move from mud hut to split-level and skyscraper? Where do we factor in there?

    1. Yet when you “criticize in private” for behavior that occurred in a team meeting or affects the team, you undermine team members’ accountability to each other. You send the message that team members are accountable only to you, not to the team. You also send the entire team the message that they don’t need to hold each other accountable — you’ll do it for them. In short, you shift accountability from the team to you.
  9. Jun 2017
    1. Done regularly whenever she addresses the viewer, though it's justified as the show originally was supposed to be set in a computer game. This then leads into a ten second pause during which she stares directly at you waiting for a "response from the viewer."
  10. Apr 2017
  11. Mar 2017
    1. 10,000 subscribers to the vTaiwan mailing list and a total of around 1,000 individual participants contributing to the forums

      Seems really important to note that an whole order of magnitude more people want passive notification of activity, but have never signed up to participate in forums.

      Definitely bangs home the important of mailing lists for engagement.

    2. four stages to the vTaiwan process

      These is actually the established ORID methodology that is well-established in the public consultation community: http://www.betterevaluation.org/en/evaluation-options/orid

    3. vTaiwan tackled this by introducing a mechanism in Pol.is that measures the opinion groups before and after. In order for an idea to reach the successful threshold of agreement, it had to gain the support of “all of the majority plus half of the initial minority”. This subsequently made mobilisation by a majority group less of a problem

      We need to figure out what is meant by this, as it seems quite important in terms of methodology.

      A couple possible interpretations:

      1. Might be that two polis conversations were run, the second involving duplicates of comments from the first. This would allow monitoring of changing opinions. But we haven't seen any evidence or mention of this up to now. Derek rightfully brought up that this would be asking a lot of participants, and so seems unlikely.
      2. Could involve monitoring opinions groups over time in a single polis convo. But this seems unlikely, as people can't change their individual votes on comments. This methodology would seem very unsound.
      3. Perhaps it's something lost in translation in communicating to the Nesta report writers. Perhaps the mention of "initial minority" is adding a time element that is not actually in play. If this is the case, it's likely referring to a way that vTaiwan allowed non-Majority Opinions into the final agenda, so long as they were consensus statements for a large opinion group, and at least half of the other opinion groups agreed.

      The third option seems most sensible. We were planning to compare the linked docs in the article on the Uber consultation. Comparing the raw data, the polis convo, and the PDF agenda should allow us to validate whether this was the case.

  12. Feb 2017
    1. set of evaluation criteria from the outset

      This is where the gaunlet is thrown down. We should take this challenge seriously.

    2. limiting data collection

      also ethical considerations for limiting data collection. even if we can collect it for free (ie. via Facebook), should we still do it?

    3. who participated and why? Did the process inform citizens about important political issues? Did it succeed in improving public trust, or propensity to engage in the future?

      really important questions

    4. Honest discussion around failures

      Would be SO powerful to kick off our initiative with a culture of HONEST discussion around failure

    5. Your Priorities

      investigate this

    6. some criticism, particularly where conversations on Facebook appear chaotic or difficult to follow

      Using Facebook is meeting people where they are, but comes with compromises at scale.

      Further complicating this is that Facebook makes metadata collection simple, which would help with the measurement goal suggested at end of report.

    7. Pol.is, opinions have to win over the approval of both groups (a supermajority), which reduces the negative effects of mass mobilisation by one opinion group

      Ah! Didn't think of this as a hack to prevent gaming and hijacking, but works really well for that. (ie. would have to fake a whole persona in order to affect, which takes concerted effort)

    8. disallowing the direct responses

      this is a really interesting hack that pol.is seems to embrace as well, according to CEO

    9. Ensure that sufficient staff and resources are available

      Consider scheduling days/weeks for "shifts". We didn't do this for MyDem0cracy, and so sometimes no one was watching.

    10. maximise the quality of interaction and contributions, based on the task at hand and the target participant group
    11. micro-tasks

      micro-tasks will need to be balanced to avoid soliciting really shallow interactions (ie. "upvoting")

    12. Consul

      investigate this

    13. Sli.do

      investigate this

    14. engagements that involve multiple phases, users should be guided through the whole process from a single ‘hub

      This seems really relevant, given that we aspire to have multiple consultations, which in turn each may have multiple phases down the road, if we take cues from from Swerhun book (3 cycles) or vTaiwan (4 stages)

    15. Evaluating Digital Citizen Engagement: A Practical Guide.

      we should check this out, as measurement seems really important and missing so far.

    16. fluctuating or insufficient numbers of volunteers
    17. grow the skills of the team
    18. ongoing development of the digital tools
    19. create and manage projects
    20. demands on civil servants were acute

      In order to avoid making life harder for civil servants:

      1. Have enough civil servant allies to spread load, and
      2. ensure that tools support them VERY well.
    21. financial and human resources
    22. translating

      This might be really interesting for more "common interest" consultations down the road, given Toronto's diversity. We should help non-English communities engage more and learn about City processes.

    23. without the requisite support, there is the risk of failing to deliver on promises and ambitions, as a result alienating citizens who have contributed
    24. support from decision-makers

      We have support of civictech community and the CIO branch of staff. Should we aim to get support of elected officials as well? Particularly as their power is most easily understood by citizens, and so would perhaps give more legitimacy.

      If we get support of politicians, might be interesting to secure from progressive and conservative.

      Also, how important is it to get support from other branches of staff, if at all?

    25. interactive live-streaming tools

      Might be interesting to give people autonomy to look around and be more present with YouTube Live 3D via Ricoh Theta 3D camera (Ricoh Canada has printer service oontracts with city, and might be interested in donating hardware for pilot?)

    26. higher investment in mass communications and PR may also be required
    27. work closely with civil society groups to help reach audiences not accustomed to using the internet
    28. local or experiential knowledge
    29. specific knowledge, expertise or information
    30. restructured the budget to reserve a proportion exclusively for the most deprived areas of the city

      This is a little more obvious how to solve when money at stake (ie. just earmark funds), but important for us to consider how we could compensate for under-represented voices in consultation. How can we avoid excluding, or re-weight outcomes to compensate for privileged folks who have more [time] resources to shape conversation.

    31. risks of conflict or capture will be higher when engaging people widely on a topic that is highly contentious

      this is something we need to be very careful about. If we make the tent too big, interest groups might take over.

    32. set of questions that are based on beliefs or values which do not have an objectively right answer

      This complements the other types of knowledge very well:

      1. specific knowledge / expertise (education for dyslexic kids)
      2. experiential knowledge (community needs)
      3. beliefs of values without objective "right" answer (abortion)
    33. being clear about the issue you are seeking to address, who you need to engage, and how you plan to use the outcomes.

      Really like this clear list:

      1. what issue
      2. who needs to be engaged
      3. how will outcomes be used
    34. with the representative making a final decision on the recommendations

      Seems like this project was very low-maintenance for reps: they kicked it off with vid, then had everything summarized and presented to them. This sounds low-burden.

      Would be great it we could aim for low staff burden for this process. (Although first run might be a little higher burden.)

    35. begins with a video from the representative presenting the issue

      Could we do something like this, with a staff member or elected official presenting the issue, to create legitimacy from the get-go?

  13. Jul 2016
    1. That City Council request the Minister of Transportation to consider making a regulation under Section 185(1) of the Highway Traffic Act prohibiting pedestrians from texting actively using a handheld wireless communication device or handheld electronic entertainment device while using on any travelled portion of a roadway.

      Watch out Pokemon Go, fans...

  14. May 2016
    1. OPSEU is expecting its own independent report tracking continuing problems with SAMS, now being undertaken by Prof. Wayne Lewchuk, of the School of Labour Studies and Department of Economics at Hamilton's McMaster University. Lewchuk will be visiting job sites next month, but CBC News got a sneak peek at the results of a survey undertaken in February and March which suggest 95 per cent of caseworkers reported that they felt the software change made the job harder. Approximately one-quarter of caseworkers took part in the survey.

      Lewchuk will have insight into a large procurement failure, and how it affects people in the civil service and the public. He might also have ideas on opportunities for citizen initiatives in light of these failures.