The ultimate goal of work in cognitive architecture is to provide the foundation for a system capable of general intelligent behavior. That is, the goal is to provide the underlymg structure that would enable a system to perform the full range of cognitive tasks, employ the full range of problem solving methods and representations appropriate for the tasks, and learn about all aspects of the tasks and its performance on them.
– from Laird et al., "SOAR: An architecture for general intelligence"
October 4, 2021: This post is a bit of a rush-up recap of the last month or so since this period of my life began. Finding out about the fellowship was a late-August surprise, after an unassuming email notification. I still remember Orthogonal Research and Education Lab PI, Dr Bradly Alicea for his letter of recommendation months ago. August was also my first month after moving to Boston, following a whirlwind of a summer abroad on a bit of travel, new jobs, research, and maybe "a week of vacation in London surrounding a business trip to Oxford." It was also exciting to be lead organizer and co-host for the Discussion Group at CogSci 2021 "Trajectories in Cognitive Science", as well as participate in a slew of other connference activity at OREL. Seeing Boston at the end of summer, and finding I know many colleagues, academic, general nerds, and friends here already, was a treat.
In several papers published in Biological Cybernetics in the 1980s and 1990s, Kawato and colleagues proposed computational models explaining how internal models are acquired in the cerebellum. These models were later supported by neurophysiological experiments using monkeys and neuroimaging experiments involving humans. These early studies influenced neuroscience from basic, sensory-motor control to higher cognitive functions. One of the most perplexing enigmas related to internal models is to understand the neural mechanisms that enable animals to learn large-dimensional problems with so few trials. Consciousness and metacognition -- the ability to monitor one's own thoughts, may be part of the solution to this enigma. Based on literature reviews of the past 20 years, here we propose a computational neuroscience model of metacognition. The model comprises a modular hierarchical reinforcement-learning architecture of parallel and layered, generative-inverse model pairs. In the prefrontal cortex, a distributed executive network called the "cognitive reality monitoring network" (CRMN) orchestrates conscious involvement of generative-inverse model pairs in perception and action. Based on mismatches between computations by generative and inverse models, as well as reward prediction errors, CRMN computes a "responsibility signal" that gates selection and learning of pairs in perception, action, and reinforcement learning. A high responsibility signal is given to the pairs that best capture the external world, that are competent in movements (small mismatch), and that are capable of reinforcement learning (small reward prediction error). CRMN selects pairs with higher responsibility signals as objects of metacognition, and consciousness is determined by the entropy of responsibility signals across all pairs.
Even though the web environment facilitates daily life, emotional problems caused by its incompatibility with human cognition are becoming increasingly serious. To alleviate negative emotions during web use, we developed a browser extension that presents memorized product images to users, in the form of web advertisements. This system utilizes the cognitive architecture Adaptive Control of Thought-Rational (ACT-R) as a model of memory and emotion. A heart rate sensor modulates the ACT-R model parameters: The emotional states of the model are synchronized or counterbalanced with the physiological state of the user. An experiment demonstrates that the counterbalance model suppresses negative ruminative web browsing. The authors claim that this approach is advantageous in terms of explainability.
Xilica, a provider of collaboration products that help unlock the power of human connection, has partnered with global audio specialist Sennheiser to develop a tightly-integrated room solution that minimizes installation time, improves audio quality, and enhances collaboration for end-users. The partnership will leverage premium audio products from both companies, offering integrators a seamless way to deploy adaptable, IT-friendly audio-conferencing systems. The complete solution, which pairs the Sennheiser TeamConnect Ceiling 2 beamforming microphone array with Xilica's ecosystem of DSP, user interfaces and network endpoints, ensures exceptional speech intelligibility for physical and remote participants in hybrid spaces. This gives end-users the flexibility to move throughout their space and present in their own style, while eliminating risk of inaudibility. Integrators and end users benefit from the Sennheiser TeamConnect Ceiling 2's seamless integration with Xilica Solaro processors and Xilica Gio network endpoints, enabling an easy, no-code setup with pre-validated compatibility. As part of the alliance, Xilica has enabled support for the TruVoicelift functionality of Sennheiser's TeamConnect Ceiling 2 (TCC2) across its Solaro Series of DSPs, which simplifies the installation of multiple Sennheiser TCC2 microphones through automatic testing and calibration of audio, saving hours of time during system commissioning.
Those designing autonomous systems that interact with humans will invariably face questions about how humans think and make decisions. Fortunately, computational cognitive science offers insight into human decision-making using tools that will be familiar to those with backgrounds in optimization and control (e.g., probability theory, statistical machine learning, and reinforcement learning). Here, we review some of this work, focusing on how cognitive science can provide forward models of human decision-making and inverse models of how humans think about others' decision-making. We highlight relevant recent developments, including approaches that synthesize blackbox and theory-driven modeling, accounts that recast heuristics and biases as forms of bounded optimality, and models that characterize human theory of mind and communication in decision-theoretic terms. In doing so, we aim to provide readers with a glimpse of the range of frameworks, methodologies, and actionable insights that lie at the intersection of cognitive science and control research.
Cognitive Design for Artificial Minds (Routledge/Taylor & Francis, 2021) explains the crucial role that human cognition research plays in the design and realization of artificial intelligence systems, illustrating the steps necessary for the design of artificial models of cognition. It bridges the gap between the theoretical, experimental, and technological issues addressed in the context of AI of cognitive inspiration and computational cognitive science. The event is moderated by Antonio Chella (Prof. of Robotics at the University of Palermo) The event is free (but the registration is mandatory) and will be held on Gather Town (you will receive the link once registered). The book "Cognitive Design for Artificial Minds" (with related editorial reviews) can be found at: Antonio Lieto is a researcher in Artificial Intelligence at the Department of Computer Science of the University of Turin, Italy, and a research associate at the ICAR-CNR in Palermo, Italy. He is the current Vice-President of the Italian Association of Cognitive Science (2017–2022) and an ACM Distinguished Speaker on the topics of cognitively inspired AI and artificial models of cognition.
Now, a complex AI-driven automation initiative like cognitive RPA is majorly anchored by bots. They need extensive strategy, support, maintenance and trouble-shooting arsenal around it – in a continuous manner. Process changes, app updates, security patches etc. have a huge impact on bots and involve human intervention. Bot maintenance and governance should be a conscious and well-thought out strategy from day one. Having a Robotic Operations Center (ROC) in place makes the job easier for enterprises betting big on cognitive RPA.
Care Angel, the leader in AI and voice-enabled nurse assistant technology, announced that Mike Hahn has recently joined the company as the new vice president of business development. In this role he will be responsible for leading the company's strategy for aggressive revenue growth moving forward. The appointment follows the recent news that prominent healthcare executive Bud Flagstad, will now serve as the company's new CEO. Mr. Hahn, holds an MBA and is a 15 year healthcare veteran who brings to the senior business development role a commitment to both innovation and collaboration as well as deep experience with large payer and provider relationships. He previously led the Innovation team for UnitedHealth Group's OptumCare unit where he was responsible for the strategy and implementation of new business opportunities and technology-enabled resources to serve various health care stakeholders spanning consumers, health care providers and practices.
Ali Boyle is currently a Research Fellow in Kinds of Intelligence at University of Cambridge and University of Bonn. Her main research field is philosophy of mind and psychology, focusing particularly on nonhuman minds and the methods used to study them. She holds a PhD in Philosophy from University of Cambridge. In this interview, we talk about artificial intelligence, cognitive sciences and philosophy of mind. My research focuses on theoretical questions about nonhuman minds: what are nonhuman minds like, and how can we learn about them?
Modern machine learning techniques based on artificial neural networks (ANNs) are implemented through algebraic manipulations of vectors, matrices, and tensors in high-dimensional spaces. While ANNs have an impressive ability to process data to find patterns, they do not typically model high-level cognition. Furthermore, ANNs are usually models of only a single task. Otherwise, when an ANN is trained to learn a series of tasks, catastrophic interference occurs, with each new task causing the ANN to forget all previously learned tasks [8, 21, 22]. On the other hand, symbolic cognitive architectures, such as the widely used ACT-R [1, 31], can capture the complexities of high-level cognition but scale poorly to the naturalistic, non-symbolic data of sensory perception, e.g., images, or to big data sets necessary for modelling learning over a lifetime, e.g., corpora with hundreds of millions of words.