#Cognition #Clash in the #IoT #SXSW
Thank you to everyone who attended our (Karl Smith and Thom Heslop) talk at SXSW, it’s the start of a long road into a really complex and contextual problem. But being silent in the crowd as the King walks by with no clothes on is not an option, peoples lives, futures and prosperity is at risk, not to mention the risk of multi-trillion dollar lawsuits that can follow by knowingly distracting people who are engaged in critical tasks.
The IoT – Internet of Things (Ubiquity) is the next great opportunity for commerce to engage with business enterprises and customers. However, there is no unified approach to the mental load between physical interaction, mental interaction and digital interaction. This cognitive landscape is inhabited by associated experiences that gel human behaviour and machine interfaces through, touch, mouse and keyboard. The usage of sight, voice and thought create new complexities and risks which have until recently been the subject of defence technologies (battlefield and strategic), where clear outcomes and prescribed mental models exist.
The diversification of these touch points and multi-point human logic models clash and derail human thinking patterns.
We are looking for people and their knowledge to help create an Ubiquity Open Standard. We are doing this because no one else has noticed this fundamental error in thinking, the hoping that product based companies will work together in creating common standards that are driven by an understanding of human thinking capabilities, cognitive models, relational thinking and machine interactions is unlikely.
While product manufactures continue with supremacy attitude to other ecosystem products and services,
“the human voice and our needs and desires are subjugated to simply another component”
albeit the one that is constantly paying for everything without any input on how it works.
Some Foundations (the rest will go in a technical paper)
Distributed Cognition studies the ways that memories, facts, or knowledge is embedded in the objects, individuals, and tools in our environment. According to Zhang & Norman (1994), the distributed cognition approach has three key components: Embodiment of information that is embedded in representations of interaction Coordination of enaction among embodied agents. Ecological contributions to a cognitive ecosystem.
In Embodied Interaction Dourish -everyday human interaction is embodied; non-rationalising, intersubjective and bodily active. User, not designers, create and communicate meaning and manage coupling. Not just concerned with what people do, but also with what they mean by what they do and how that is meaningful to them. It reflects the sets of meanings that can be ascribed to objects and actions over those objects as part of a larger task or enterprise
Cognition the key to the mind, how people understand what they can do is by comparison a Diagnostic Methodology (goals, adaptations, conventions) with what they already know by accessing the Active Narrative patterns they have created in their own minds according to Smith (2005).
Cognition Groups create a communication method, they carry intention, meaning, risks and benefits.
- Some Cognition patterns are common, shopping basket etc.
- Some Cognition Patterns are social by Family, Sports Team etc.
- Some Cognition Patterns change without notice
Guided Interaction, existing websites offer guided interaction – simplified cognitive pattern encapsulating a plethora of interacting technology and data systems: Shopping Basket – This representation allows for distributed cognition > appropriation > cognitive pattern forming understand– once a user has used a shopping basket they will understand how to use them and generalize: transferable cognitive pattern
Some of the issues with the IoT
- There is no standard of interactivity for humans in the IoT – not a problem if passive background machine-to-machine. A very big problem if actively interacting with humans, who are all different and can create their own meanings for example LOL.
- How does a user form any cognitive patterns from an invisible system?
- IoT combines known patterns as hidden machine-to-machine communications that can create mistrust and security fears
- Detailed component view we have constructed around daily interactions is no longer valid
Some of our initial research
IoT Design Principals
- What is device / service for?
- Where will it be situated?
- When will it be triggered?
- What other devices will it be interacting with?
- Where can it clash?
- Security? – * Lack of security – Shodan
- Design Principal: “Do No Harm”
IoT Design Risks
Context is critical
- Situational interaction problems for consideration
The following barriers reduce our ability to understand the situation
- Perception based on faulty information processing
- Excessive motivation – over motivated to the exclusion of context
- Complacency
- Overload
- Fatigue
- Poor communications
A possible solution
- Avatar (can be visual, sound, texture, smell, taste or a combination) – smart use of Artificial intelligence (AI), where the users cognitive interface is patterned on their unique cognition pattern through a learning algorithm
- This avatar should be directional and instructional like digital signage
- This avatar should respond to the users behavioural interaction and should fall away gracefully as users behaviour becomes more ‘expert* In effect it should be a learning system – learns from the users rather than based on static rules
- For example the AI that George Hotz has built into his self driving car while not the answer points to the kind of thinking required to find the answer, don’t tell the machine to watch and learn from a human and then carry out your task (from 3.33 to 5.04) “the point is to drive naturally like a human, not some engineer’s idea of safety“. For anyone who then thinks this is the final solution, please let us know why you think driving a car is like cooking dinner or navigating the street?
The Full SXSW Talk is on YouTube
Connect to the speakers on LinkedIn here Karl Smith and Thom Heslop