Abstract: I have developed a new android called “Alter”. Based on the artificial life principle (using T3 chaos and the new learning principle called stimulus avoidance[1][2]), Alter behaves autonomously and he receives visual and sound information from the environment. Here Alter mimics human behavior coming in front of him spontaneously using the internal simulator. Especially with this version at the Barbican center[3], Alter creates new actions patterns by switching between the action sequence generated by the internal simulator and the stored memories of its past behaviors [4]. I would like to discuss the purpose of this project Alter and the concept of ” offloaded agency”. This work is also based on the series of our psychological experiments called Perceptual Crossing Experiments”[5].
References:
[1] Lana Sinapayen, Atsushi Masumori, Takashi Ikegami: Learning by stimulation avoidance: A principle to control spiking neural networks dynamics., PLoS ONE,12(2): e0170388. 2017.
[2] Doi, I., Ikegami, T., Masumori, A., Kojima, H., Ogawa K., and Ishiguro, H. A new design principle for an autonomous robot, 14th European Conference on Artificial Life(ECAL2017), pp.490-466.
[3] The Barbican major new exhibition 2019 May 15th- : AI: More than Human http://digicult.it/slider/ai-more-than-human/[4] Otto E Rossler, Lisa-Ruth Vial, Frank Kuske, August Nitschke, Takashi Ikegami and Andrei Ujica. Brain Equation and Personogenesis. Clinics in Pediatrics, vol.2 2019, pp.1-11.[5] Hiroki Kojima, Tom Froese, Mizuki Oka, Hiroyuki Iizuka, Takashi Ikegami: A Sensorimotor Signature of the Transition to Conscious Social Perception: Co-regulation of Active and Passive Touch, Frontiers in Psychology, 8.01778, 2017.

The advent of modern technological devices create new opportunities to advance more than one field of enquiry at the scientific and clinical levels on human behaviour and development. Common to these areas are issues with social dynamics and embodied cognition that require new methods for data gathering within naturalistic situations, managing the large amount of data collected from many sensors, and overall introducing new analytical platforms that can handle data streams in near-real time. Up to now, particularly in psychology and psychometric research, many of these aspects have remained largely beneath the observer’s awareness and because we have mostly relied on observation for both science and clinical areas, we have missed important components scaffolding human behaviour and social dynamics. We have also made many a priori assumptions about our data that can now be challenged by empirical estimation of many parameters related to the nervous systems functioning. This has created a tangible gap between the research and the clinical communities that should be better integrated, once we become aware of the issues. In the talk, I will discuss approaches and will provide examples of how we can deal with those issues in multiple settings, ranging from computational areas of AI and ML, to basic psychological research, with a particular focus on autism research. I will advocate for multi-disciplinary approaches to address many of the current problems, to bring multiple communities together, and to develop collaborative efforts to tackle and understand such issues.