Archive

Archive for January, 2017

Weekly QuEST Discussion Topics and News, 27 Jan

January 26, 2017 Leave a comment

QuEST 20 Jan 2017

We will start this week with our colleague Igor:

Applying QuEST inspired Self-structuring Data Learning to aerial infrared and visual images.

The Update

Previously, we proposed and implemented a Self-structuring Data Learning approach for autonomous exploitation of multimodal data based on synthetic data. The fundamental aspect of the approach is to apply some of the QuEST tenants to simplified three level model with interaction of bottom-up and top-down signals between adjacent levels. Our approach based on maximal reduction of information at the “gist” level that is represented with binary vector. To overcome data growth with processing problem, we developed multiscale grid processing. We have applied tracking algorithms to MAMI-I dataset, which as an airborne multi-camera EO/IR collection. MAMI-I data provide enough information about vehicles, associated spatial and texture information (appearance, size, etc.), and location and movement history.  The tracks and the corresponding image “features” from MAMI video streams and other data sets have been used to test the algorithm’s ability to autonomously develop hierarchical data representation and potentially predict situation development.

 

We also want to continue our discussion of the Kabrisky lecture – What is QueST? – specifically this week there are parts of the presentation we’ve not made it to and our recent discussion on what are the characteristics of the desired representation make these points important to discuss.   A recent thread of discussion has focused on the missing link for recommender systems – they can’t ‘appreciate’ the information in the data or the context of the human’s environment / thoughts thus they become a ‘feed’ – social media example – but people can’t seem to disconnect – if we design a joint cognitive social media system focused on ‘mindfulness’ and thus a context aware ‘feed’ that provide some value –

The other item on the agenda is Cap has to give several talks coming up and generate some material on historical perspectives in neural science and also computational models associated with machine learning and artificial intelligence so we will have some discussion along those lines.  “Artificial Intelligence and Machine Learning:  Where are we?  How did we get here?  Where do we need to go?  Does that destination require ‘artificial consciousness’?”

 

Specifically – in one recent study cap presented at it was concluded that:

Operationally AI, it can be defined as those areas of R&D practiced by computer scientists who identify with one or more of the following academic sub-disciplines: Computer Vision, Natural Language Processing (NLP), Robotics (including Human-Robot Interactions), Search and Planning, Multi-agent Systems, Social Media Analysis (including Crowdsourcing), and Knowledge Representation and Reasoning (KRR).  In contradistinction to artificial general intelligence:

  • Artificial General Intelligence (AGI) is a research area within AI, small as measured by numbers of researchers or total funding, that seeks to build machines that can successfully perform any task that a human might do. Where AI is oriented around specific tasks, AGI seeks general cognitive abilities. On account of this ambitious goal, AGI has high visibility, disproportionate to its size or present level of success, among futurists, science fiction writers, and the public.

We will want to pull on these threads with respect to the breakthroughs in deep learning and the promise of other approaches to include unsupervised learning, reinforcement learning …

news-summary-39

Advertisements
Categories: Uncategorized

Weekly QuEST Discussion Topics and News, 20 Jan

January 19, 2017 Leave a comment

QuEST 20 Jan 2017

We want to continue our discussion of the Kabrisky lecture – What is QuEST? – and along that line we want to provide more specifics in our use in our Theory of Consciousness of the word ‘situated’ and the word ‘simulated’ and the idea of structural coherence (seems to be the embodied term and related to situated in the psychology literature).

As our use of the term are not consistent with some usage (for example in the Journal of Cognitive Engineering and Decision Making – article in Press by our QuEST colleagues Patterson and Eggleston) – over the last two weeks we have reconciled the terms – at least in the mind of Cap.

What we need to discuss this week is how to engineer systems that have the desired ‘situated’ / ‘simulated’ / structurally coherent nature using current approaches to machine learning and artificial intelligence.

We will initially emphasize applications in making ‘intuitive’ machines that act as interactive multi-sensory content curators.  We define intuition as the quale that humans experience as a result of a ‘sub-conscious’ computation’s outcome being attended to in working memory (thus becoming conscious of the result) without the details of the deliberation.  As intuition is by this definition a quale a machine that replicates that computation is an artificially conscious machine (we define consciousness as the generation of qualia – and when we say above a machine that replicates that intuition computation we are including the constraint that the reproduction includes instantiation of the key defining engineering characteristics of all qualia – situated / simulated / structurally coherent).

The second characteristic of intuition that is emphasized by our colleagues is the use of ‘recombinations’ of prior experiences / memories that can be used in the ‘sub-conscious’ computation.  In the QuEST models we use the word ‘simulation’ to capture this similar idea that the computation of all qualia is the generation of a representation that much of the content of the representation is inferred versus a recall of previously experienced episodes.

In summary:

Common ground –thanx to our 711th colleagues we’ve converged on a path that is comfortable with respect to using terms and our focus on engineering computational characteristics that currently are not emphasized in AI/ML solutions but appear to be key constraints in consciousness

In our QuEST world we define ‘intution’ – as the quale that is evoked in consciousness to provide an actionable conclusion to a computation that is being accomplished without conscious awareness of the details of the deliberation – ‘I think walking into this environment is not a good idea’ – ‘I’m not going to enjoy this class’ – ‘that boy is not the right match for my daughter’, ‘that truck rumbling down the hill out of control towards a gas station is a bad thing’ …

As all qualia – the representation is the key not only for the experience but for the computation of the quale – and in the literature (thanx to Robert, Bob, Anne — they’ve shown us how our QuEST developed tenets can be reconciled with their view of intuitive cognition) – it is clear that it has to instantiate being situated, simulated and be structurally coherent consistent with the tenets we’ve developed in QuEST over the last decade

So as a clear first step in our endeavor to make a fully conscious computer we seek making a computer with intuition – and I’m still convinced the best first step for this is in content curation – the computer that ‘feels’ you might be interested in this part of the multi-modality (audio, video, text, …) world versus that part of the sensory streams and based on your response to what it provides you (both estimating your conscious and subconscious representational states ~ the system is emotionally intelligent) changes the interaction (what it provides you next) – by the interaction it increases your emotional intelligence also

In this world that is going towards ubiquitous computing, virtual and augmented reality – this intuitive computer will always be on and learn from natural sources be multi-sensory and will reason (manipulate its own representation and do so to facilitate accomplishing tasks thus understanding) – and part of that manipulation will form new qualia via imagining unique combinations of existing qualia ~ chunking to facilitate gisting and a key means for abstraction – our use of the word simulation

 

The other item on the agenda is Cap has to give several talks coming up and generate some material on historical perspectives in neural science and also computational models associated with machine learning and artificial intelligence so we will have some discussion along those lines.

 

Specifically – in one recent study cap presented at it was concluded that:

Operationally AI, it can be defined as those areas of R&D practiced by computer scientists who identify with one or more of the following academic sub-disciplines: Computer Vision, Natural Language Processing (NLP), Robotics (including Human-Robot Interactions), Search and Planning, Multi-agent Systems, Social Media Analysis (including Crowdsourcing), and Knowledge Representation and Reasoning (KRR).  In contradistinction to artificial general intelligence:

  • Artificial General Intelligence (AGI) is a research area within AI, small as measured by numbers of researchers or total funding, that seeks to build machines that can successfully perform any task that a human might do. Where AI is oriented around specific tasks, AGI seeks general cognitive abilities. On account of this ambitious goal, AGI has high visibility, disproportionate to its size or present level of success, among futurists, science fiction writers, and the public.

We will want to pull on these threads with respect to the breakthroughs in deep learning and the promise of other approaches to include unsupervised learning, reinforcement learning …

news-summary-38

Categories: Uncategorized

Weekly QuEST Discussion Topics and News, 13 Jan

January 12, 2017 Leave a comment

QuEST 13 Jan 2017

This week we want to several things – we want to continue our discussion of the Kabrisky lecture – What is QueST – and along that line we want to clean up some concern on our use in our Theory of Consciousness of the word ‘situated’.  As our use of the term is not consistent with some usage (for example in the Journal of Cognitive Engineering and Decision Making – article in Press by our QuEST colleagues Patterson and Eggleston) – we want to reconcile the terms.  The other item on the agenda is Cap has to give several talks coming up and generate some material on historical perspectives in neural science and also computational models associated with machine learning and artificial intelligence so we will have some discussion along those lines.

Specifically – in one recent study cap presented at it was concluded that:

Operationally AI, it can be defined as those areas of R&D practiced by computer scientists who identify with one or more of the following academic sub-disciplines: Computer Vision, Natural Language Processing (NLP), Robotics (including Human-Robot Interactions), Search and Planning, Multi-agent Systems, Social Media Analysis (including Crowdsourcing), and Knowledge Representation and Reasoning (KRR).  In contradistinction to artificial general intelligence:

  • Artificial General Intelligence (AGI) is a research area within AI, small as measured by numbers of researchers or total funding, that seeks to build machines that can successfully perform any task that a human might do. Where AI is oriented around specific tasks, AGI seeks general cognitive abilities. On account of this ambitious goal, AGI has high visibility, disproportionate to its size or present level of success, among futurists, science fiction writers, and the public.

We will want to pull on these threads with respect to the breakthroughs in deep learning and the promise of other approaches to include unsupervised learning, reinforcement learning …

news-summary-37

Categories: Uncategorized

Quest Kabrisky Lecture, 6 Jan

January 6, 2017 Leave a comment

QuEST 6 Jan 2017:

The Kabrisky Memorial Lecture for 2017: ‘What is QuEST?’  The first QuEST meeting of each calendar year we give the Kabrisky Memorial Lecture (in honor of our late colleague Prof Matthew Kabrisky) that brings together our best ‘What is QuEST’ information.  As all QuEST meetings this will be an interactive discussion of the material so anyone who has never been exposed to our effort can catch up and those who have been involved can refine their personal views on what we seek.

QuEST – Qualia Exploitation of Sensing Technology – a Cognitive exoskeleton

PURPOSE

 

– QuEST is an innovative analytical and software development approach to improve human-machine team decision quality over a wide range of stimuli (handling unexpected queries) by providing computer-based decision aids that are engineered to provide both intuitive reasoning and “conscious” deliberative thinking.

 

– QuEST provides a mathematical framework to understand what can be known by a group of people and their computer-based decision aids about situations to facilitate prediction of when more people (different training) or computer aids are necessary to make a particular decision.

 

 

DISCUSSION

 

– QuEST defines a new set of processes that will be implemented in computer agents.

 

– Decision quality is dominated by the appropriate level of situation awareness.  Situation awareness is the perception of environmental elements with respect to time/space, logical connection, comprehension of their meaning, and the projection of their future status.

 

– QuEST is an approach to situation assessment (processes that are used to achieve situation awareness) and situation understanding (comprehension of the meaning of the information) integrated with each other and the decision maker’s goals.

 

– QuEST solutions help humans understand the “so what” of the data {sensemaking ~ “a motivated, continuous effort to understand connections (among people, places and events) in order to anticipate their trajectories and act effectively” for decision quality performance}.1

 

– QuEST agents implement blended dual process cognitive models (have both artificial conscious and artificial subconscious/intuition processes) for situation assessment.

 

— Artificial conscious processes implement in working memory the QuEST Theory of Consciousness (structural coherence, situation based, simulation/cognitively decoupled).

 

— Subconscious/intuition processes do not use working memory and are thus considered autonomous (do not require consciousness to act) – current approaches to data driven, artificial intelligence provide a wide range of options for implementing instantiations of capturing experiential knowledge used by these processes.

 

– QuEST is developing a ‘Theory of Knowledge’ to provide the theoretical foundations to understand what an agent or group of agents can know, which fundamentally changes human-computer decision making from an empirical effort to a scientific effort.

 

1 Klein, G., Moon, B. and Hoffman, R.R., “Making Sense of Sensemaking I: Alternative Perspectives,” IEEE Intelligent Systems, 21(4), Jul/Aug 2006, pp. 70-73.

Categories: Uncategorized