What is 3Ai homework? In 2020, 3Ai is excited to be back in the classroom with a new cohort of Masters students, teaching a new branch of engineering into existence.
How would you like to play along at home? We are trying something different this year and giving you a little glimpse of life as a 3Ai student. And that means… homework! (disclaimer: don’t worry, you’re not being assessed). During the semester we will share with you a sample of what we are reading, a snapshot of the influences and discussions shaping 3Ai. This will match up with the fortnightly themes of our program.
In Semester One, students are undertaking two courses:
1. Question Framing, within which students engage with perspectives from a wide range of disciplines to frame critical and constructive questions about cyber-physical systems; and
2. Build, where students learn to create cyber-physical systems in collaborative teams with an explicit awareness of the environmental, social, and technological contexts that their systems could eventually exist within.
Like you, our students are now learning from home. 3Ai has officially gone diasporic, and our students have been transitioned to digital and remote classrooms for the remainder of the Semester. Staying connected but keeping apart!
Your resources for Theme Five:
In Question Framing this fortnight, we learn that when building blocks of artificial intelligence start to connect, we begin to encounter networks, internets, the World Wide Web and IoT. But networks and systems are not new, and are not unique to technology. From trade routes to weaving, from ecosystems to cities, we encounter systems all the time. How do we understand the way human, environmental and technical parts of systems interact and shape each other? Where are system boundaries? How do we learn about technologies going to scale from those that already have? How do we start thinking about systems as connected at scale? What are some key concepts that a focus on networks and systems leads us to think about?
This was a fascinating topic for the #3Ai2020cohort to tackle, as there are all sorts of ways you could cluster and make sense of this fortnight’s readings. One way you could go about it:
explore the different perspectives about and from within cyber-physical systems,
consider the different ways to study, understand and theorise these systems, and
check out the models, maps and visualisations of these systems for a window into the relationships, dynamics, flows, key nodes or features.
Here’s a small selection of resources for Question Framing this fortnight:
Kipp Teague (1968–75) Project Apollo Drawings and Technical Diagrams. Curated by Steve Garber. NASA HQ. Technical diagrams. [Have a look at some diagrams and consider what’s depicted and what’s missing]
CERN (1992) The birth of the web. Website. [Browse the first website]
Susan Leigh Star (2010) This is not a boundary object: reflections on the origin of a concept. Science, Technology and Human Values 35(5): 601–617. Paper. [This is an explainer which followed her original paper. If you want to extend your thinking further, here it is: Susan Leigh Star (1989) Chapter 2: The structure of ill-structured solutions: Boundary objects and heterogeneous distributed problem solving. In: Les Gasser & Michael N Huhns (eds) Distributed Artificial Intelligence (Volume 2). San Francisco: Morgan Kaufmann Publishers. Pp 37–54. Book chapter.]
First Dog on the Moon (2015) The NBN is bad. Featured in the Guardian, December 16. Cartoon. [This is a reminder that there are vastly different perspectives in a system — consider how this juxtaposes against other stories of the internet]
Nora Bateson (2015) An ecology of mind: A daughter’s portrait of Gregory Bateson. Film.
Antonio de Luca & Sasha Portis (2019) New York’s subway map like you’ve never seen it before. NY Times. Interactive graphics.
It has been a BIG couple of weeks in the Build course. Perhaps you have seen the buzz and the outpouring of love on Twitter after the Semester One Demo Day? Each student has been building a system throughout the semester. This was their opportunity to show off their ‘maker project’ to each other. In the video chat, there was a constant stream of statements expressing amazement and encouragement. The 3Ai faculty and staff are very proud of what the #3Ai2020cohort have achieved so far!
This fortnight the students also delved deeper into AI research, while acquiring hands on skills by continuing to develop tic tac toe agents using collaborative coding skills. Students worked together to define a metric that measures the “success” of a tic tac toe agent and then used it to guide the design of their group’s agent. They practiced using a version control system, git, to merge their code, and we reflected on the kinds of norms and practices that guide distributed software development teams and collaborative coding projects in general.
The homework this fortnight was to take on the role of “AI Archaeologist”, applying research methods to trace threads running through the history of AI research and development. We provided students with a collection of 35 research papers, patents, and articles, some of which are below. From these documents, each student created a cataloguing system guided by what features of the paper they deemed significant for answering their research questions. Using this information, they then created a visual “knowledge map” of AI. Comparing and critiquing different mapping techniques furthered student’s contextualisation of work so far undertaken this semester and look out to future horizons.
This is a taste of what our students were reading:
McCulloch, W.S., Pitts, W. A logical calculus of the ideas immanent in nervous activity. Bulletin of Mathematical Biophysics 5, 115–133 (1943). https://doi.org/10.1007/BF02478259
Rosenblatt, Frank F.. “The perceptron: a probabilistic model for information storage and organization in the brain.” Psychological review 65 6 (1958): 386–408. https://doi.org/10.1037/h0042519
Samuel, Arthur L.. “Some Studies in Machine Learning Using the Game of Checkers.” IBM J. Res. Dev. 3 (1959): 210–229. https://doi.org/10.1147/rd.33.0210
Minsky, Marvin. “Steps toward Artificial Intelligence.” Proceedings of the IRE 49 (1961): 8–30. https://doi.org/10.1109/JRPROC.1961.287775
Turk, Matthew A. and Alex Sandy Pentland. “Eigenfaces for Recognition.” Journal of Cognitive Neuroscience 3 (1991): 71–86. https://doi.org/10.1162/jocn.19220.127.116.11
Goodfellow, Ian J., Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron C. Courville and Yoshua Bengio. “Generative Adversarial Nets.” NIPS (2014). https://papers.nips.cc/paper/5423-generative-adversarial-nets
All of our materials strive to reflect a variety of voices and perspectives, to in turn reflect our diverse cohort. Our 2020 cohort come from a range of countries, including Nigeria, the United States, Nepal, Mexico, India, Iran and Australia; a range of disciplines, including law and policy, economics, computer science and machine learning, biology, music, you name it; and are diverse in gender and ethnicity.
Some of the #3Ai2020cohort have been blogging their progress, reflecting on the resources and their learnings in the 3Ai Masters Program. This fortnight, you can check out Lorenn’s insightful reflection: Mapping my understanding of systems mapping [verb] and systems maps [noun]