2. SEAMLESS AND INTELLIGENT DEVICE ENSEMBLES
WEAVING A WEB OF APPLICATIONS
PEOPLE AS PERIPHERALS
COMBINATORIAL FORECASTS
INTUITIVE INTERFACES
TOTAL RECORDING AND TRANSCRIPTION
NEW OS FOR PERSONAL DATA
CYBERSECURITY: INOCULATING THE INTERNET
E3: ENERGY EFFICIENCY ENGINEERED
THE DIGITAL CONNECTOME
ALGORITHMIC ARMS RACE
3. CRITICAL TENSIONS
Glass Houses vs. Black Boxes
The Human Internet vs. The Machine Internet
The Cosmopolitan Internet vs. Internet Fragmentation
Augmentation vs. Dependency
4. f
lo w
S
E
N
S
E
E
M
E
R
G
E
C
O
N
T
R
O
L
massivedatastorageandsearch|summarizationandstorification
mobileandembeddedsensornets|powerandbatterydesign
dynamicnetworkarchitectures|object&userecognition
context-awareapps&services
object&
userecognition
| open resources
context-aware
apps
&
services
automated process optimization | wearable computers
machine learning | context-aware apps and services
high-speeddataminingandmatching
predictiveanalytics|machinelearning
multi-sensory interfaces | augmented reality | smart surfaces
cognitivecom
puting
power and battery design
cloud supercomputing
| dy
nam
ic
networkarchitectures
cloud supercomputing | dynamic network architectures
high-speed data mining and matching | data surveillance
context-awareappsandservices|openresources
multi-factor biometric authentication
multi-factorbiometricauthentication|datasurveillance
SeamlessandIntelligentDeviceensembles
W
eaving
a
w
eb
of
applications
People as Peripherals
Intuitive Interfaces
TotalRecordingandTranscription
e3: energy efficiency engineered
the
digita
l
connectome
AlgorithmicArmsRace
NewOSforPersonalData
Cybersecurity: Inoculating the Internet
ocuspec.com
technologyreview
.com
lifelapse.com
wired.com
sec.gov
sec.gov
ibm
.com
scientificamerican.com
greenlight.calit2.net
futurity.org
pittpatt.com
pittpatt.com
databanker.com
bis.gov.uk
apple.com
technologyreview.com
w
ebintents.org
ifttt.com
projects.csail.mit.edu
taskrabbit.com
TotalRecordingofChildhoodCognitivescientistDeb
Royinstalled11camerasand14microphonesinhishomebefore
thebirthofhisson.Herecordedatotalofmorethan90,000hours
ofvideoover3years,morethan200terabytesofdata,completely
documentingtheprocessofhisson’slanguagelearning.
By
2022,people
and
organizationswillhavetheabilitytokeepdigitalscrapbooks
ofalloftheirinteractionsandtobeabletominerecordsofallpastmeetingsfor
patterns
and
insights.
Voice, haptics, and gesture will dominate the next era of interface, but will soon
be
joined by responsive and anticipatory systems that will leverage im
plicit intention
markers such as eye tracking, emotional signals, and other non-verbal cues.
Software initially developed for crowdsourcing and
social enterprise applications is now able to perform
task assignment and coordination roles traditionally
occupied by human managers and supervisors.
M
eta-services
and
progr
am
m
atic
constructs thattransform
disparate
appl
ication
functionalities into
real-tim
e
solutions
w
illb
e
the
connective
tissueoftheInternet,m
uch
as hyperlinks
w
ere
to
theearlyweb.
Anintegratedecosystemofnetworkeddeviceswill
allowforseamlesscontent,contextualawareness,
andintelligentassistancebasedonuserneeds.
wallets,repositories,orclouds—allservedbyintelligentdigitalagents.
searchablepersonaldatacontainers—vaults,banks,lockers,eventnetw
orks,
By2022,muchofouronlineinteractionwillbeexperiencedthrough
secure,
balance
openness and security.
code-sniffing
technologies, bio-authentication methods, and new ISP rules to
Pernicious
code is increasingly infecting the Internet, encouraging use of
technology (ICT) energy efficiency.
needs and to improve information and communication
developed and scaled to meet growing device power
cells, and biofuel for mobile devices are being rapidly
charging surfaces, solar and kinetic chargers, fuel
Universal power supplies, wireless charging,
connection
pathw
ays
ove
rothers.
networks w
ith
form
ation
s
designed
to
privilegecertainnetwork
phenom
ena, su
ch
as
w
iring
together
engineered
to
take
adva
ntage
ofneural-likefrontierNetworks architecture
w
illbe
deliberately
againstotherpredictivealgorithms.
astheychangetheirowncodesinordertocompete
butwillalsocreateemergentandvolatileproperties
meaningfulpatternsofhumanandmachinebehavior,
Algorithmswillharnessmassivedatatogenerate
Lifelapse
LifelapseisaniPhoneapplication.Oncetheuserputs
the
phone
in
a
specialpocket,itautomaticallytakespicturesevery30
seconds.These
arethencompiledastime-lapsevideostobeshared.
Emotion-Detecting
Smartphone
Sam
sung
is prototyping
new software that can infer a smartphone users’ m
ental state
from
behaviors such as typing speed, use of backspaces and
deletes,
and shaking.
Ocuspec This powerful, low-cost gesture recognition technology
works across platforms allowing it to be integrated into everything
from a TV to a laptop.
TaskRabbit Taskrabbit allows users to
offer payment to a crowd of screened helpers
for help with chores.
Soylent
Soylent, developed by Michael
Bernstein at MIT, embeds crowdsourced editing
and proofreading functions directly into MS Word,
allowing users to receive input from hundreds,
in real-time, while drafting a document.
W
eb
Intents
Thisfram
ework
lets
users
enabletheirInternetapplicationsto
communicatewithand
issue
com
m
ands
toeachother,likeTwitterorDelicious.
AirplayMirroringAirplayMirroring
letsyouwirelesslystreamfromoneormore
iPadsoriPhonestoyourHDTVthroughan
AppleTV.MidataMidataisavoluntaryprogramthattheUKGovernment
MidataMidataisavoluntaryprogramthattheUKgovernmentis
undertakingwithindustry,givingconsumersincreasingaccess
totheirpersonaldatainaportable,electronicformat.
Databanker
TheFacebookPrivacyInformerApp
calculates
r
isk
metricsbasedonhow
anindividualhasconfigured
Facebook
privacy
settingsandgeneratesafinalnumberthatratespersonalprivacy
risk.
PittPatt PittPatt provides a facial recognition software develop-
ment kit (SDK) that allows anyone to integrate this capability more
easily into their products and services.
Scavenging Power From The Air
Georgia Tech researchers are developing
printable, flexible devices that can scavenge
ambient electromagnetic energy and convert
and store the power.
Greenlight Calit2 researchers developed
a system of networking architectures,
sensors nets, quantitative software analysis,
server arrays, and visualization tools for greater
efficiency and reduced carbon footprints.
Uncovering
Pro
perties
o
f
Cognitive
Netw
o
rks
Recent research
suggests that the brain’s neurons
align
to
form
a 3D
grid.
C
o
g
n
itive
C
o
m
puting
IBM
is
currently
m
odeling
the
brainsystemsof
com
plex
anim
alsasa
precursortowhatitis
calling
“cognitive
com
puting,”orcomputation
properties
leveraged
bythebrain.
Polyworld
Thisisoneofseveralplatforms
designedtocreateanenvironmentforAIto
evolvenewapproachesandcapabilities.
PredictiveSoftwaredisrupts
financialMarketsRegulatorshave
foundautomatedtradingsoftwaretobe
akeyfactorinsomeextrememarket
fluctuations,includinga“flashcrash”in2010.
D
ig
ital
Ants: Stigm
ergic
CyberSecurity
Programs
w
ander through
com
puters searching out and swarming viruses.They
leave
“digital pherom
ones” when they sense a problematic behavior.
If
Th
is, Th
en
Th
at
This service allows users
to
create
”triggers” across platforms, “I write a
tw
eet, post it on
Facebook and
then share the
“triggers” so
friends can
use them
too.”
TheSamsungSmartWindow
The
SamsungSmartWindowisatransparentLCD
whichturnsaregularwindowinyourhouse
intoaon-demandtouch-screeninterface.
Legend
This map highlights the technological
changes occuring with the Internet and
how they will transform the human Internet
experience in the coming decade. It contains
seven foundational technology forecasts
and ten combinatorial forecasts. The
foundational technology forecasts are divided
into two clusters: platform and workload,
located along either side of the figure.
The combinatorial forecasts are arrayed
around the human figure and divided into
four functional potentials: FLOW, SENSE,
ONTROL and EMERGE.
Fou nd ationa l
T ec h nol ogy For ec as ts
Fundamental changes are
occurring in the way the Internet
functions. Forecasts for the core
computational and network
architectures for mobile and
embedded systems are shown
in the platform clusters. New
and powerful functionalities for
networked computing systems
are forecasted in the workload
clusters.
C omb inator ia l For ec as ts
In a messy and complex world
where accelerating change
creates technological disruptions,
dynamic social reconfigurations,
and rapid organizational
innovation, linear extrapolation
of technological trends is
insufficient. Instead, technological
change will be driven by the
combination and recombination
of foundational elements.
Combinatorial forecasting,
then, involves the systematic
intermingling of distinct
technology trends with the goal of
perceiving alternative outcomes
that could occur when these
elements are mixed together.
T ec h nol ogy E na b l e rs
Each combinatorial forecast
contains a list of several core
technologies that are changing
conditions of possibility.
These technological trends or
advancements enable the shifts
described in the forecasts.
S igna ls
Signals support each forecast
with concrete examples from
the world today. These include
new technologies, research,
businesses, or other endeavors
that demonstrate the direction
of the forecast as a whole.
fo u r po t entials
FLOW
Functional assets are being
fragmented and recombined
into new arrays to provide
more relevant, efficient, and
personalized service.
SENSE
Computational interfaces are
becoming more ubiquitous,
human-centric, and invisible.
ONTROL
Boundaries are difficult to
maintain with the Internet, but
new forms of control will aid in the
struggle for secure transactions
and private interactions.
EMERGE
Machine learning will advance
quickly, unleashed from human
control and allowed to develop
in unexpected ways.
TotalRecordingofChildhoodCognitivescientistDeb
Royinstalled11camerasand14microphonesinhishomebefore
thebirthofhisson.Herecordedatotalofmorethan90,000hoursof
videoover3years,morethan200terabytesofdata,completely
documentingtheprocessoflanguagelearning.
NewOSforPersonalDatarepositories,orclouds,allservedbyintelligentdigitalag
en
ts.
chablepersonaldatacontainers—vaults,banks,lock
ers,
eve
ntnetw
orks,w
allets
,
By2022,muchofouronlineinteraction
willbe
expe
rienc
ed
th
ro
ugh
secure
,sear
PLATFORM
networks
Re-Engineering The Network:
Quality Performance
Wireless Networks: Dynamic Configurations
computational cloud
Cloud-Served Supercomputing:
New Operating Systems
Open Resources: Data, Apps, and Services
Exotic Hardware: Biological and Quantum
embedded and local systems
Smart Surfaces: Scalable, Ubiquitous,
and Invisible
Smart Spaces: Sensing, Recording,
and Learning
Coordinated Devices: Heterogeneous,
Connected, and Adaptable
mobile hardware
Mobile Sensor Platforms:
Pervasive and Powerful
Mobile Computing: Handhelds
and Wearable Computing
Personal Robots:
Drones, Swarms, and
Anthropomorphic Assistants
workload
rendering and simulation
Utilizing Big Data: Simulations, Models,
and Visualizations
Video Web: Synthetic, Interactive,
and Hyperlinked
Dimensionality: 3D, Immersive,
and Holographic
mining and reasoning
Autonomous Processes: Machine-to-Machine
Contextual Assistance: Analytics,
Optimization, and Validation
Understanding Humans: Machine Learning
sensing and recognition
High-Resolution Sensing: Motion,
Volume, and Activity
Real-World Interaction: Recognition,
Discovery, and Filtering
5. networks
A cascade of fundamental changes in net-
work software and hardware will greatly improve
connections.
computational cloud
-
tational cloud of super intelligent services, including
-
cations with machine intelligence and learning.
embedded and local systems
displays of every size, sensors, actuators, and
mobile hardware
A vast and complex array of connected
pervasive in human experience and environments
in unprecedented ways.
FOUNDATIONAL TECHNOLOGY FORECASTS
rendering and simulation
applications outsourced to the cloud, devices’
rich user interfaces and graphics. Advances in
rendering, simulations, and modeling will make
displays a major driver of CPU chip performance,
with photorealistic simulations dominating network
device usage.
mining and reasoning
Machine intelligence and learning will
applications, automating a vast amount of what
was once human work.
sensing and recognition
Devices equipped with sensor software
anticipating user needs.
PLATFORM CLUSTER WORKLOAD CLUSTER
6. HOW TO USE THIS MAP:
Thinking Systematically about the Future
F LOW, SENSE, ONTROL EMERGE.
Working with this Map
EXPLORE F LOW, SENSE,
ONTROL EMERGE
FOCUS
DRILL DOWN
CONSIDER
RE-COMBINE
create
Institute for the Future
Technology Horizons Program
Acknowledgements