Monthly Archives: June 2017

Phil 6.30.17

8:30 – 4:30 BRI

Phil 6.29.17

Research

  • Travel from HCIC 2017
  • Continued to read Conflict and Consensus
  • Continued to work on P&R Considered Harmful

Phil 6.28.17

HCIC 2017

  • Thought for the day – board games to explain paper production (Journal and conference) and proposal writing
    • John Walton (Board games, etc) Elizabeth Swenson (Application Crunch)
      • Joe Dumit – Science studies Fracking simulation board games
  • John Tang, Gina Venolia and Kori Inkpen, Microsoft
    • Breaking the Future of Social Media
    • Discussant: Dan Russell
    • Multi stream clusters – how do we view? (Scoop)
    • Interactions among the streamers and between the streamers and the viewers. Also, how does the design scale? Twitch doesn’t scale, BTW – you can’t keep up
    • Likes as a percentage of viewers upvoting affects who sees what Paper at CHI
    • Can be hard to stream under particular circumstances. High crowds, degrated conditions, etc. Leads to short viewing due to frustration
    • SXSW covered with multiple parallel streams
      • Note – stabilized shoulder mount for a streamer with extra batteries? Better mic? Camera?
    • If there is a crash, the preview allows an overview and lets people continue an interrupted stream.
    • Many short glances, most of the time is in long watches/views. Browse/browse/browse/stay
    • Interaction engagement – all deliberate. Lurking is the absence of action, but there is a surprisingly low percentage of that.
    • Sweet spot is high flexibility environments like trade shows, where the remote audience can interact.
    • How do you get effects at at sale with a prototype
      • Recruit crowds from Turk – reflects incentives. Can test usability and engagement at a higher level of interaction, which exercises the platform more
      • Corporate Volunteers – viewed by interest. very linear. More interested in the talks, rather than the wandering.
      • Unknown – power law. many short, few long (and much higher lurking)
    • Breaking existing commercial boundaries to enable the future
      • Many semi-overlapping ecosystems. How to standardize? Standards didn’t work. incentivised to lock-in
      • Maybe a standard for subsets of data to support a commons? Universities may be a good place to take the lead on this?
      • How to unify?
  • Dan Russell discussant
    • We already have a lot of streaming cameras
    • 911 Truthers based on self-selected partial information.
    • Can’t reverse engineer context
    • Reverse effects cause and effect can get confused
    • Google moderator wound up requiring an intermediary to select the questions
    • Only studies with real money matter
    • Is the MCR mechanism of handling multiple feeds the way to handle social media? Possible for ABC to buy feeds from Fox?
  • Questions
    • Sharing PII is also a problem. Not only costs but risks
    • Is the value in the liveness, or in the history? Attention value is value.
    • Value is transparency? Authenticness? David Brin, the transparent society Trustworthiness of the messenger, rather than the anonymous streamer?
    • Social information proximity vs IR gatekeeping?
    • We’re having trouble differentiating between perspectives? There are different small groups that don’t relate to a giant company. A small group may be important.
    • Events of civic interest vs. novel and entertaining live streams
    • Gare Steinberg’s incredible sad true love story
    • There is a difference between a trained, effective producer? What about predators choosing prey from a flock? What does an effective liver performer do?
    • Becoming a temporary media company that produces a curated stream. An improvisational theatre company
    • Tweak the Tweet Palen & Starbird
    • It is very interesting to watch representatives of Google and MSFT discuss research as startups, where acquisition is the large scale corporate research model.
  • Short talk and working session: Charlotte Lee 
    • HCDE Comics – Reflecting on Living in the Present and Designing for the Future
    • Philip Agre – Critical Technical Practice
    • Check out the memex as inspiration for the Research Browser
    • You can draw in pencil all day long at a coffee shop, but when you start inking people will stop and notice.
  • See if ASRC could support HCIC. What’s he cost/value proposition. What does it mean to be part of the consortium?
  • Reworking Design as Critical Fabulations – Daniela Rosner
    • Building core memory as quilts
    • Core rope memory
    • “Valorization of individuals has to recede” in collaborative processes.
    • To believe what is physically impossible might be possible in a different way
  • Discussant – Stacy
    • It’s a book chapter! In what book?
    • Design shift from individualism (includes small teams) to collective
    • Situated, and contextual
    • Not having a solution and moving on, but creating a process.
    • Who gets to tell the stories
    • There is a reviewer 2 meme. I did not know that.
    • Interviews are fabulating?
    • Grounded theory is a way of academically acknowledging and defining path dependency.
    • corememoryshield.com/report.html
    • Creating different physical instantiations that provide different audiences.
    • Inquire into your method’s history. What are the ways of knowing, and Who is knowing?
    • Who are the actual people doing the tasks, and what are the behind-the scenes work to create the product (Read papers by Lilly Irani. Such command of the language!)
    • Identifying devalued labor. How. And what does it mean to track and understand than that. There is a difference between the dialogic practice of group work and the business model of an industry that maximises profit through celebrity creation.
    • There is a difference between a performer and a creator of the source material for that performance.
  • Michael Muller and Q. Vera Liao, IBM
    • Exploring AI Ethics and Values through Participatory Design Fictions
    • The developers of robots believe that they represent the user. THis may be true in that there are no customers yet. Except for LAWS systems?
    • Value Sensitive Design (chlorine water treatment and chlorine gas)
      • Theoretical
      • Contextual
      • Ethnographic
    • We do not have an agreed-upon list of values.
    • Value sensitive inquiry instead? Descriptive ethics. Not at normative ethics yet
    • How do you speak your values about something that you don’t yet have?
    • Using design fictions
    • Problems with current popular fictions (Books – personal view, and media – complex interweavings of corporate interests)
    • Ponder a spectrum from utopia to dystopia and timeframes
    • Exercise that was confusing and too fast. But generally on story completion
    • Slideshare.net slides
  • Discussant: Kristin Eschenfelder
    • Would you kill the fat robot?
    • An echo chamber (like Nazi-ism) might be indicated by the alignment of philosophies around a desired outcome. Because there is no external validity?
    • Moral Machine MIT Media Lab
    • There is a difference in commercial ethics and philosophical ethics.

Phil 6.27.17

HCIC 2017

  • What if Robots designed your next homes (Maya  Cakmak)
    • PR2 robot
    • Homes are designed for people, home tasks are performed for people
    • Are homes of the rich different from the homes of the lower classes?
    • Affordances for robots. Actually means that the home is becoming part of the robot. Like an automated factory. Structure that is needed for robots
    • Hotels and warehouses are intermediate, semi-structured environments that afford robots better
    • What about home gantry robots? With it’s own storage for non=-human tools?
  • Technology- centered design – Amanda Lazar
    • Solutionism – Solves problems before they exist
    • The MAYA principle – Most Advanced yet Acceptable
    • What happens when the technology is explicitly at the center of the design process? Making this explicit opens up some opportunities. Variant of the flying shiny logos problem (Overton window)
    • How things affect you, your family, and people you know in that order
  • Daniel Russel. Unintended consequences. Ramps are one thing, people and cats and dogs are another
  • Lana – what about the schedule of change. Like wiring up houses for networks and then wifi comes along
  • National Robotics Initiative
  • Should robots be designed around current accessibility standards. THen robots can grow out of their disabilities over time?
  • Are we too tied to practices now? Washing machines lead to clothing that survive the washing machines. Ecosystems. What are all the dependencies that need to simultaneously shift?
  • What about friction? And how we deal with slaves?
  • Futures Beyond Design Panel (Brent Hecht, Lilly Irani, Silvia Lindtner, Scott Mainwaring)
    • Brent – Design as an optimization process
      • But externalities! How do we internalize these ideas so that unintended consequences are minimized
      • The great decoupling – computers displace workers
      • Importance of data sets (google’s powers of search are heavily dependent on Wikipedia user sourced data
      • Designing technologies for a world where data creators are compensated in a meaningful way
        • It would be neat if the research browser could do this. Maybe show a ticker that shows a ticker with ‘you would have made $xx’
    • Scott
      • How technological infrastructures are produced?
      • Three-ring-binder project in Maine. Fiber to rural areas. But how does this affect the users? Who are the users and how do we know them? Many misconceptions. Culture can change rapidly but what’s the difference between minority rights, women’s rights and LGBT rights?
    • Sylvia Lindtner
      • Making Cruel Machine Dreams
      • The maker movement (Is this human value add?) References Clay Shirky’s Here Comes Everybody
      • The means of production need to be placed back in the hands of the proletariat?
      • Precarious work as opposed to techno utopianism? Who profits? The 3D printer manufacturers
      • Shenzhen is really a policy experiment. Which concentrated the means of production
    • Lilly
      • When impact requires more than design. The case for crowds.
      • Turkopticon – Turkers reviewing employers in the system.
      • Turker organization – Turker Nation, CloudMeBaby, etc
      • Dynamo: Safe space for turk worker activism (forum with upvoting) – but for precarious labor in general
      • Need to support the understanding of labor laws and have a legal understanding
      • Platform Cooperativism
    • John Zimmerman – Carnage Mellon
      • Won’t the ‘computers will take our jobs’ problem fix itself?
      • Shouldn’t stop us from thinking about how to minimize disruption and pain
      • Lilly – Culture matters in how these technologies are integrated
    • What is an example of a project that embodies these ides
      • Sylvia – civilla
      • Scott – HatchTheFuture.org Facebook’s communities of action??
      • Brent – Machine Learning can be tuned to populations so that it does not have disparate impacts. A chrome plugin that supports selective boycotts.
      • Lilly. Platform Cooperativism. Stockse is a photo collective made by ex Getty people. But what does participatory management look like when there are 10,000 members?
    • How do we do this? How do we engage as academics?
      • Lilly: Who represents the crowdsourcing workers?
      • Brent: Personas vs. Sociatals?
      • How to communicate complex political and social ideas in a easily consumed way

Phil 6.26.17

HCIC 2017

  • Carl DiSalvo – Alternate Histories of Design Futures
    • What else might design be – different interpretations of the world
    • Speculation (Speculative Design?)
    • Design as critique to critiquing design
    • Designers domesticate technologies
    • Speculative design is to trigger thoughts, not to become a product per-se. THis is about nudging in a direction. It’s more like art harnessed to thinking about interaction
    • Fantastic futures, extraordinary objects intended to be provocative
    • Archigram 
    • Global Tools.
    • 21st Century Speculative Design MOMA curates and legitimates Speculative Design
    • Design and Violence
    • Core77
    • From representation to events
    • Vernacular Futures -> cheap futures. Cheap stuff that we buy. Fidget spinners???
    • Speculative Enactments (SIGCHI paper)
    • prefiguration
    • Mundane work of performing on the world (Signage)
    • Implications for design
      • How can computation become integral to our design and experiences (Bill Gaver – Your Data Catcher)
      • Turkopticon
  • Katherine Isbister – Reflections on Carl’s presentation
    • Interdependent wearables
    • Magic circle of play (new games movement, whole earth catalog, etc)
    • Global Game Jam
    • Feynman diagrams spreading through interpersonal transmission. In other words, they do not get used unless someone shows you. The difference from personal contact vs reading.
  • Legitimate peripheral participation
  • Brave New Alps
  • Prefiguration in Contemporary Activism
  • Analog game movement – write rulesets that reflect the vision of the architect. No, that didn’t work. Then creating the experience locally. Now a co-creation process that’s half product/half event? Practice being transmitted through objects. De-centering the designer, recentering the practice.
    • Douglas Wilson (Folk Games) No Autour.
  • Repair maintenance vs event. Events become objectified and bounded. It can be treated as a product.
  • Star Trek & CMC (Lana Yarosh)
    • Technology constraints and technologies
  • Discussion as speculative fiction Wendy Kellogg – Microsoft Research
    • Apologetics
    • Research through design
    • Artifacts embody myriad choices that can’t be captured in a written document
    • Generalizing through juxtaposition? Dourish 2014 – LSI is a theory of meaning

 

Phil 6.23.17

7:00 – 8:00 Research

  • Thinking that I need address Starbird’s idea of false triangulation in the context of belief space. I think there is something akin to a strange attractor for certain topics, particularly around fear and authoritarianism.
  • Found some good stuff on path dependency breaking, particularly this dissertation, which includes a simulator: Path dependence in two-sided markets
  • Ebook manager for handling things like pdfs? Caliber

8:30 – 4:00 Research

  • Got the last run. What a mess: Oct_2016_vs_June_2017_local_clusters what’s going on? Need to look more in depth at the source data. That’s going to require more code, so I’m going to write the report on what I’ve got now.
  • Writing report. Done.

Phil 6.22.17

Research 6:45 – 7:45

BRI 8:30 – 3:30

  • Working on reading in files, then running permutations on cluster membership
  • Because I can never remember how ass-backwards pandas.Dataframes are:
    def initDictSeries(rows = 3, cols = 5, offset=1, prefix ="doc_"):
        dict = {}
        row_names = []
        for j in range(cols):
            row_names.append("row_{0}".format(j))
        for i in range(rows):
            name = prefix+'{0}'.format(i)
            array = []
            for j in range(cols):
                array.append ((i+offset)*10 + j)
            #dict[name] = tf.Variable(np.random.rand(cols), tf.float32)
            dict[name] = array
        return pd.DataFrame(dict, index=row_names)
    
    df = initDictSeries()
    print("df = \n{0}".format(df))
    s = df.loc['row_1']
    print("\ns = df.loc['row_1'] = \n{0}".format(s))
    s['doc_1'] = 99
    print("modified = \n{0}".format(df))
  • Yay! It looks like it’s working. Next step is to run it. Done. This is the compare of the two runs on the older data as it sits on the servers: Oct_vs_Mar_2016_clusters
  • Here’s the same runs locally, using t-sne. This is more like what I was expecting to see: Oct_vs_Mar_local_2016_clusters
  • Here’s the old vs. new (Left this cooking. These charts take forever to calculate) Oct_2016_vs_June_2017_local_clusters

Phil 6.21.17

6:45 – 7:45

  • Got the travel info from HCIC. Looks like they got tickets just for the conference, and I don’t have the energy to try to get that fixed. I’ll just bring my running shoes.
  • Multiple movement modes by large herbivores at multiple spatiotemporal scales
    • Interestingly, the rate of movement by individuals depended strongly on the amount of time they spent in groups, with highly gregarious individuals being much more sedentary than more solitary individuals (19). Far-roaming, solitary individuals had a higher risk of mortality than did more sedentary, gregarious individuals (19). Hence, sociality triggered changes in movement modes that had important demographic consequences.
  • Socially informed random walks: incorporating group dynamics into models of population spread and growth (Ref 19 from above)
  • Behavioral Change Point Analysis
    • in Python (and a bridge to [R])
    • Change-Point Analysis: A Powerful New Tool For Detecting Changes
      • Change-point analysis is a powerful new tool for determining whether a change has taken place. It is capable of detecting subtle changes missed by control charts. Further, it better characterizes the changes detected by providing confidence levels and confidence intervals. When collecting online data, a change-point analysis is not a replacement for control charting. But, because a change-point analysis can provide further information, the two methods can be used in a complementary fashion. When analyzing historical data, especially when dealing with large data sets, change-point analysis is preferable to control charting. A change-point analysis is more powerful, better characterizes the changes, controls the overall error rate, is robust to outliers, is more flexible and is simpler to use. This article describes how to perform a change-point analysis and demonstrates its capabilities through a number of examples.
  • The Similarity Metric
    • We propose a new “normalized information distance”, based on the noncomputable notion of Kolmogorov complexity, and show that it is in this class and it minorizes every computable distance in the class (that is, it is universal in that it discovers all computable similarities). We demonstrate that it is a metric and call it the similarity metric. This theory forms the foundation for a new practical tool.

9:00 – 5:00 BRI

  • Got FoxyProxy working and downloaded the data. The clustering seems non-optimal. Rerunning locally to see what’s going on
  • Need to write an adaptor to build a common excel files from subsequent clustering runs
  • Need to get the previous run. Got the input and output files for both. They are the same length, which is odd. Pinging Heath. These are both older data, clustered differently. What I’ve pulled is the newest data.
  • Meeting on remote dev environments. Lots of security discussions. My main concerns are on usability WRT bandwidth and network quality.
  • Dug up the CSE billing info and updated the credit cards
  • Adding method in membership_history_builder.py that will organize multiple cluster files into a single dataframe.
    • Built the set containing the common elements
    • Need to build  Dataframe of the correct dimensions and populate it
    • Need to run the membership code over that and plot the results

Phil 6.20.17

7:00 – 8:00 Research

8:30 – 4:30 BRI

  • Didn’t think to download the NVIDIA libs while at home, so doing that now
  • Tensorflow now requires 64 bit Python and I had 32 bit. Reinstalling
  • Setting up access to the AWS instance. Connecting!
  • Having trouble with Firefox… Rebooted
  • Installed the FoxyProxy plugin but am unable to get it to read new config file. Asked Matt for assistance. He’s going to send me his config file.

Phil 6.19.17

7:00 – 8:00 Research

  • Nice trip back from the conference. I think my take home messages are:
    • What I’m working on is non-obvious
    • Agent-based simulation is becoming a norm in this field
    • Creating a useful knowledge map makes sense to everyone(?) working in the field
    • The concept of explore, flock, and stampede have good resonance
    • Gate-keeping in agent design/distribution and information retrieval are SERIOUS problems
  • Brought in the code that I wrote on the laptop. Compiles and runs nicely
  • Creating a poster (24×36) version of the PosterPage. Also creating a dual-sided handout. Done, and sent out for printing.
  • Ordered a poster tube with strap

8:30 – 3:30 BRC

  • Expense Report. Done. Maybe? The app doesn’t take street address without barfing. May have to go back in and re-edit? Or is that only for mileage?
  • Continue to work on setting up Linux Python/Tensorflow env
  • Got IntelliJ up, but had to point at the ‘Development’ files for windows due to the crippling download rates.
  • Apparently we have new data on CI. Recluster?
  • Then I let Windows update and everything stopped. Went home to continue
  • Installing python
  • Installing Java
  • Installing xampp

Misc

  • Register van – done!