Differences

This shows you the differences between two versions of the page.

--- machine_wilderness_symposium [2015-11-02 12:27] – maja
+++ machine_wilderness_symposium [2015-11-02 21:47] – 213.219.149.208
@@ Line 7: / Line 7: @@
 Machine Wilderness: new ecosystems where environment & technology co-exist, in which humans are less central
-#machinewilderness
+[[https://twitter.com/search?q=%23machinewilderness%20&src=typd|#machinewilderness]]
 === Theun Karelse ===
@@ Line 172: / Line 172: @@
   * Hybridity is a slippery slope - sometimes things should be respected for what they are rather than being forced to 'hybridise'
+=== Kenzo ===
+Attention, movement of water and air in and around the body
+=== Guszti Eiben ===
+Evolving robotic ecosystems - nature inspired robotics/computer science (evolution as inspiration - influences language)
+Takeaway messages
+. Artificial evolution is real, not an emulation of a 'real' evolution, just another form (Darwin evolution, Watson & Crick DNA, Turning & Von Neumann - creating evolution (computers) - programmers set the rules)
+  * link between evolution (biology) & problem solving (engineering): individuals (in a biological framework), natural selection (choosing fitness), reproduction (digital sex)
+  * evolutionary algorithm (evaluation-selection-variation loop)
+  * it can solve hard problems, cope with changes and deliver original solutions
+  * Macroscopic view (after Dennett): if you have variation, heredity and selection you must get evolution. Variation - push towards novelty, selection - push towards quality
+Historical context
+  * (19-20 ct) Wetware (biosphere, we can observe what has happened in the past and present, in vivo)
+  * (20-21 ct) Software (evolutionary computing, a generative concept, in silico)
+  * (21ct) Hardware (evolution of things, in materio)
+. Robots can be evolved
+  * not all humanoid, not all mechatronics (soft robotics)
+  * evolution can create intelligence -> artificial evolution can create artificial intelligence
+  * intelligence and embodiment: environment + body + mind -> behaviour (AI in 20ct. narrow view of only the mind - chess, in now body + mind (and hopefully also environment) - football
+  * Genotypes (variation - mutation & crossover) & phenotypes (selection) - can be done in robotics too
+  * behaviour can evolve in robot populations - we know how to evolve software brains, how to evolve physical bodies -> modules/cells or 3d printer (artificial womb)
+    * ethical dimension - this can get out of hand… (e.g. radiation hazard, biohazard -> robohazard?) - do we need a "kill switch"; we probably don’t want distributed birthing robots, but a centralised birth clinic, with strict control…
+  * challenge: simulations don’t scale up very well
+  * Cambridge: mother robot that produces 'a child' consisting of active and passive parts, that can move on its own
+  * application: breeding farm for service robots or pets, entertainment (robotic parcs…); robot colonies for terraforming or ultra deep mining
+  * science: "cyclotron for evolution", understanding life, evolution of body & brain, robosphere
+=== Ivan Henriques ===
+[[http://ivanhenriques.com|Hybrid forms]]: JAP, PNBM, SM: wet & dry machines
+  * interspecimen communication
+  * environmental robotics
+  * workshop on symbiotic systems, using Amstelpark as a medium, exploring the needs and opportunities of biorobotic systems (abiotic systems - solar, temp, wind, water + biotic systems (plants, animals, bacteria); creating systems to enter a dialogue with the environment - integrated interdependent systems
+  * energy systems
+=== Judith van der Elst ===
+Forest bathing - digital technologies for the enhancement of sensory experience
+  * understanding human spatial intelligence; mapping how native american indians relate to the landscape - our technological system tends to be too flat; flow, relationships, in-between spaces
+  * embodied research - what would an embodied education look like, making use of ubiquitous computing related to landscape?
+  * extending the bodies with digital technologies; how can they help us improve our senses?
+  * understanding processes in the [[https://en.wikipedia.org/wiki/Semiosphere|semiosphere]]
+  * exploration in Amstelpark + University of Urbino - birdsong (sonotopes) connected to scents of landscapes + how do smells and sounds interact
+  * plants smell different when transplanted from their natural environment to a cultivated environment (e.g. a park) - exploring non human languages in the semiosphere (workshop in the spring)
+=== Field robotics ===
+Discussion lead by Theun Karelse
+  * experimentation in landscapes: how do you connect to it, how do we explore it
+  * make an experiment starting from that landscape to create a robotic entity
+  * borrowed scenery: asian gardening technique - including the landscape that isn’t a part of the garden
+  * first experiment in a small village in cornwall connected to five different landscapes - the landscapes change quite quickly (industrial, forest, coast)
+  * how can you connect mind+body+environment - how would the robotic species develop in different environments
+  * the goal is to have an exercise in designing environmental robots with the objective to understand how technology can be more subtle towards our landscapes - creativity?
+  * borrowed landscape - also an english invention - landscape outside separated by a fence, so the wild animals cannot enter; a robotic creature - should do something else respond to a range of different environments
+  * how do the machines/robots perceive/experience the different landscapes?
+  * begin with observing and mapping (on cards) - what you see (treetops, soil, sky…), actions (migrating, decomposing…), textures (crunchy, sticky, slimy, fluffy)
+  * how does the robot live in this system, how does it interact and die?
+  * beyond functionality and utilitarianism; starting point: it has no purpose; the systems exist for their own purpose -> to exist for some time, it’s a stakeholder in the environment, so it wants the environment to keep existing.
+  * defined capabilities - the cards could be a way to explore capabilities of the hypothetic machines; SICS experiment in expression of emotion through facial recognition with masks on people’s faces to understand what the computer might see; try to have the participants explore the environment with limitations and capabilities of hypothetic machines
+  * how could the humans explore the environment the way a limited machine/robot might do it? e.g. immersion, distractions, bodily/sensory constraints - even before the robot exists, you try to experience what it might be like for that being to exist
+  * What would the creatures respond to their habitat? What would they feed on? Where would they exist (in the earth, in the sky…)
+  * Beyond mimicry of existing biological movement - different set of responses (heat, humidity…) - what kind of patterns would the robots make?
+  * What vulnerabilities could they have? Can robots be suicidal? Can you have survival without purpose? What is survival from the POV of the robot - self-preservation, learning from the environment, dissolution into the environment? Is there reproduction?
+  * The experiment should include the robot AND the habitat - and how they might change through their interactions?
+  * How does the robot learn (procedures)? (design question) How can you make a system which is autonomous for as long as possible?
+  * Longer timeframes (Gerrit van Bakel - machines that slowly walk in the landscape), something happens once in 20 years, or so fast/ so slowly that it isn’t humanly perceptible. making links to things that technologies aren’t usually designed to do
+  * The imaginary dimension - people will make stories about it, at which point can you say whether it works or not?
+  * How will the robot have/experience a sense of agency and meaning?
+  * Machines that are sensitive and sensible
+  * How are the humans involved, if at all? How do the robots affect or interact with them?
+  * Define what you mean by co-habitation, participation, interaction…
+  * Can the robot help to overcome deficiencies in the landscape? How does it contribute to the landscape?
+  * How to avoid negative effects? Watch out not to introduce an invasive species which puts too much strain into the habitat?
+  * It must be pinpointed and defined what the goal is. The question is to rephrase the role of robotics as part of a much larger discussion of the role of humans and technologies in the landscape. Very important to make this clear before starting to work.
+  * You could make machines that can sense one thing and do one thing, then experiment.
+  * You might have a community of small robots that behave like one organism, instead of one big one; the simpler the robot individuals, the easier it might be to adapt to the environment
+  * How can the robots become a part of a larger living landscape? A whole range of processes happen - cultural, cultivated, wild, industrial, rural, tidal, cyclical… but everything is also always a part of a larger whole
+  * How do you treat the whole landscape as a robotic entity?
+  * Do you want to intervene in the entropy that is a part of that landscape or do you want to intervene and change it?
+  * What kinds of questions do we ask in the design process of an artificial organism that would co-exist in a landscape… a design science that starts from something that is as complex and changing as an ecological habitat? What is that process like? What the robot actually ends up doing is secondary
+  * connecting with other intelligencies in the landscape; collect information from plants, animals, pollution, air… (analogy of 'smart cities' that collect information from humans); how might plants and animals react to pollution, for example?
+  * the landscape might need technology so that humans might be more aware; to find out what has been hidden from our view (long timeframes, different layers and rhythms)
+  * look at ritual behaviour across different species
+  * how distinct do you want the robots to be from the environment? what do you want to find out from the environment? how will you introduce it into the environment?
+  * robots to redevelop landscapes after disasters? terraforming
+  * it isn’t so obvious to understand what is missing from our landscapes (e.g. missing elephants in EU forests; indigenous farming in America - to Europeans it looked like wilderness)
+  * First: find out how you as a person connect with a landscape; feeling the sweat and pain of the landscape (observe, then interact)