2. Exocortex is part of you
• Exocortex is extension of your neocortex via external
technological devices and systems
– Especially ones more deeply and seamlessly interfaced
• But this is a goal of many computational/communication devices
• Natural drivers in this direction
• All computational and communication devices that extend
your effective intelligence are part of it
• Some of your exocortex is shared with other people
– Beginning of group mind
• The devices that are part of your mind and effective
intelligence are as important and should be as secured /
safeguarded / not be trespassed on as your physical brain.
3. What kind of tools do you dream of?
• Voice interfaces
• Software assistant
– Takes notes (voice?), does research, notes what I need and fetches it,
notifies me of relevant things
– Is my appointment secretary
– Is my budget manager..
– Remembers everything
– Finds anything relevant and useful to what I am doing
– muse
• Internet of Things
– Not just physical things hooked to internet with all their information
– Internet of conceptual things as well
• No dead unconnected data
4. Human Computer Interface
• What kinds of interfaces are best for humans?
• Take advantage of our strengths, augment our weaknesses
– Strengths
• Chunking
• Pattern recognition (especially visual)
• Induction and abduction
• Complex (mostly social) inter-relationships
– Weakness
• Memory
• Computation
• Analysis with too many non-chunkable free variables
• Large scale sifting of information
• Finding patterns in data sets to large for us
5. Dumb data, duplication
• Information about an object X of type Y should always be
associated with that instance
• Instances should persist and have identity for long lived “things”
• Today the connection, the “what is this about” is largely not
automated but in users heads or particular software
• Can’t have internet of things like this or much reasoning and pulling
together relevant information.
• Many copies of somewhat differing data about same things in many
formats
• Relationships between things seldom 1st class entities in programs
and systems
• Persistent, distributed cloud of software objects / agents / actors
6. Software Agents
• More or less autonomous software programs
– Long running
– Task focused (specialized or task assignable)
– May cooperate with other agents in agent ecology
• Some agents are mobile
– Many are designed as loadable modules on any suitable
platform
– May propagate copies much like a worm or virus
• Search agents are a simple example
• Shopping and auction agents
• Some types of bots would qualify
7. Software Agents
• One being’s worm is another being’s research
tool
– Traveling software entities phone home
• The invisible hand in the world of software bots
– Bidding on tasks
– Contracts
– Subcontract
• Describing what is needed
• Representing knowledge
8. Mobile Computing
• There’s an app for that
– Not yet for many things and not that smart but..
• Location based computing is huge
– context awareness
– Augmented reality
– Proximity notices (people, events, tasks..)
• Sensor processing
• Explosion of cheap ubiquitous partial solutions
– Input an issue
• Voice input much needed
– Pulling out a device and navigating menus to an app and
navigating its menus is too slow
• Causes much less use than would be beneficial
– Flash mobs
• People/machines (cyborgs) flocking on information feeds
9. Digital Research Assistant
• System that scans a corpus of articles and other
sources
• Does data mining and other techniques to extract
concepts
• May be tuned to look for particular keywords
– System to find bi-metals tried and suggest new ones is
famous example, several patentable finds
• Scan of blogosphere and google results looking
for information on given topics
– Some automation composing results with references
10. Everything is intelligent
• Any object with state can have sensors and give off
information
• Augmented objects can change state on external
request / message
• Sets of such objects can exhibit unified complex
behaviors
• Adaptive systems reconfigure within programmed
limits
– Based on past history
– Current data
– Situation models
– State machines
11. Wearable computing
• Land of cyborgs
• Feeds into senses
– Eye patch or other visual input
– Ear buds
– Haptic
• Input is still a problem – needs voice
• Getting better at external brain “mind reading”
• Power is big challenge
• Not looking like a complete cyborg is important
• Need wearable systems to record everything
– Don’t record and you can’t learn more from what you experience
– Don’t record and there is less of you in the metaverse
12. Context Aware Computing
• Who I am
• What does the system know about me?
– History
– Other sources
– Derived rolling model
• Task model
• Interest model
• Personality model..
– Feedback from user
• Where am I?
• When is this interaction taking place?
• Likely current goals
• interruptability
13. Ubiquitous computing
• Computers everywhere
• Nearby computer recognizes you
• Your working sets are in the cloud
– With proper encryption and other security
• Nearby systems load the parts of your working
set that are needed for your current activities
• Devices and systems work for you in current
location
– Depending on availability, permissions..
14. Social Computing
• Cognitive Surplus
• Approximately 1% of hours usually spent on TV gave us wikipedia.
• People are more connected/aware of more of each other’s
lives than ever before
• Group mind effects of information filtering, aggregation
and spread
• Collaborative filtering, auctions, prediction markets,
reputation systems, tagging, sorting, verification
• Methods and tools to attract hundreds to millions of minds
to useful computational tasks
– By accident and on purpose
• Collective Intelligence