A general futurist look at how linear, exponential and discontinuous growth is shaping the future of technology and what may be expected in key areas such as hardware, software, semiconductors, artificial intelligence, nanotechnology, biotechnology, life extension and virtual worlds.
Audio: http://feeds.feedburner.com/BroaderPerspectivePodcast
1. The Future of Technology
Melanie Swan
Futurist
MS Futures Group
Palo Alto, CA
650-681-9482
m@melanieswan.com
http//www.melanieswan.com
Christine Peterson
Vice President and Founder
Foresight Nanotech Institute
Menlo Park, CA
650-289-0860
peterson@foresight.org
http://www.foresight.org
Liana Holmberg and Tess Chu
OS Wrangler & IP Gadfly and Developer
Linden Lab
San Francisco, CA
415-243-9000
liana@lindenlab.com
tess@lindenlab.com
http//www.lindenlab.com
2. The Future of Technology
October 2007
2
Summary
We think about growth and change in linear,
exponential and discontinuous paradigms,
history is a chain of discontinuities
The realm of technology is no longer discrete, technology is
imbuing traditional linear phenomena with exponential and
discontinuous change
Computation (hardware and software) overview: Moore’s Law
improvements will likely continue unabated in hardware; software
however is stuck
Not only will there be linear and exponential growth in the next 50
years but probably also discontinuous change, possibly a change
with greater impact than the Internet in our (current) lifetimes
3. The Future of Technology
October 2007
3
Paradigms of growth and change
Linear
Economic, demographic, biological phenomena
Exponential
Technology: processors, memory, storage,
communications, Internet communities
Discontinuous
Plane, car, radio, wars, radar, nuclear weapons,
satellites, computers, Internet, globalization
Impossible to predict
• Evaluate rapid transition time and doubling capability
• Market mechanisms
Exponential
Discontinuous
Linear
4. The Future of Technology
October 2007
4
The future depends on which coming revolution
occurs first
What will be the next Internet?
Artificial
Intelligence
Molecular
Nanotechnology
Anti-agingVirtual
Reality 2.0
Quantum
Computing
Robotics
Intelligence
Augmentation
Personal
Medicine
Affordable
Space LaunchFab
Labs
5. The Future of Technology
October 2007
5
Evolution of computation
Future of computing
New materials
3d circuits
Quantum computing
Molecular electronics
Optical computing
DNA computing
Electro-
mechanical
Relay Vacuum
tube
Transistor Integrated
circuit
?
Source: Ray Kurzweil, http://www.KurzweilAI.net/pps/ACC2005/
6. The Future of Technology
October 2007
6
Extensibility of Moore’s Law
Source: Ray Kurzweil, http://www.KurzweilAI.net/pps/ACC2005/
Transistors per microprocessor
Penryn
45 nm, 410-800m transistors
Core 2
65 nm, 291m transistors
7. The Future of Technology
October 2007
7
Current semiconductor advancements
Source: http://www.siliconvalleysleuth.com/2007/01/a_look_inside_i.html
Standard Silicon
Transistor
High-k + Metal Gate
Transistor
Historical semiconductors
65nm+
Intel Penryn 45nm chip,
shipping fall 2007
Metal
Gate
High-k
Insulator
Silicon substrate
DrainDrain SourceSource
Silicon substrate
SiO2
Insulator
8. The Future of Technology
October 2007
8
Semiconductor industry roadmap visibility
Source: http://download.intel.com/technology/silicon/Paolo_Semicon_West_071904.pdf
2007
32 nm in 2008, 22 nm in 2010
Molecular manufacturing needed for 10 nm
9. The Future of Technology
October 2007
9
Software remains challenging
Abstract, difficult to measure
Doubling each 6-10 years
Wirth’s law: “Software gets slower faster than
hardware gets faster”
Large complex projects (FAA, CIA) failure
19 m programmers worldwide in 20101
Solutions?
Distributed ecologies of software programmers
Open source vs. proprietary systems
Standards, reusable modules
Web-based software
• Aggregating collective intelligence (tagging, RSS,
presence), community platforms as the back end
(FB, LinkedIn, MySpace)
Software that programs software
1
Source: http://www.itfacts.biz/index.php?id=P8481
Lady Ada
Lovelace
10. The Future of Technology
October 2007
10
Arms race for the future of intelligence
Machine Human
Blue Gene/L 360 teraFLOPS (≈.36+ trillion
IPS) and 32 TiB memory1
Unlimited operational/build knowledge
Quick upgrade cycles: performance
capability doubling every 18 months
Linear, Von Neumann architecture
Understands rigid language
Special purpose solving (Deep Blue,
Chinook, ATMs, fraud detection)
Metal chassis, easy to backup
Estimated 2,000 trillion IPS and 1000
TB memory2
Limited operational/build knowledge
Slow upgrade cycles: 10,000 yr
evolutionary adaptations
Massively parallel architecture
Understands flexible, fuzzy language
General purpose problem solving,
works fine in new situations
Nucleotide chassis, no backup possible
1
Source: Fastest Supercomputer, June 2007, http://www.top500.org/system/7747
2
Source: http://paula.univ.gda.pl/~dokgrk/bre01.html
11. The Future of Technology
October 2007
11
Artificial intelligence: current status
Approaches
Symbolic, statistical, learning algorithms,
physical/mechanistic, hybrid
Current initiatives and funding
Narrow AI: DARPA, corporate
Strong AI: startup efforts
Near-term applications
Auditory: speech recognition
Visual: security camera (crowbar/gift)
Physical: buildings and transportation
Format
Robotic (Roomba, mower, vehicles)
Distributed physical presence
Non-corporeal
Kismet
Stanley
12. The Future of Technology
October 2007
12
Molecular nanotechnology
Definition: not work at the nano
scale or with atoms in 2D but
3D molecular placement in
atomically correct structures
Scale
Human hair: 80,000 nm
Limit of human vision: 10,000 nm
Virus: 50 nm, DNA: 2 nm
Microscopy tools
Sources: http://www.imm.org, http://www.foresight.org,
http://www.e-drexler.org, http://www.rfreitas.com
13. The Future of Technology
October 2007
13
Personal fab labs and 3D printing
Community fabs, o/s designs
MIT Fab Labs
Make, TechShop (Menlo Park)
3d printing
Fab@Home, RepRap, Evil
Personal manufacturing
Ponoko (platform)
Fabjectory
http://reprap.org
http://fab.cba.mit.edu/about
MIT Fab Labs
3D printed
plastic avatars
http://www.fabathome.org
Fab@Home
RepRap
Evil Labs
http://www.evilmadscientist.com/
14. The Future of Technology
October 2007
14
Biotechnology, hacking biology
Biology: an information science
Genomics, proteomics,
metabolomics
Faster than Moore’s Law
Sequencing and synthesizing
X Prize for Genomics
• $10M to sequence 100 human
genomes in 10 days
Personalized medicine
RNAi, protein folding
Cure vs. enhancement
Sources: http://www.economist.com/background/displaystory.cfm?story_id=7854314,
http://www.molsci.org/%7Ercarlson/Carlson_Pace_and_Prolif.pdf
15. The Future of Technology
October 2007
15
Anti-aging, life extension and immortality
Aging is a pathology
Immortality is not hubristic and unnatural
Aubrey de Grey
Strategies for Engineered Negligible
Senescence (SENS) and escape velocity
1. Cancer-causing nuclear mutations
2. Mitochondrial mutations
3. Intracellular junk
4. Extracellular junk
5. Cell loss
6. Cell senescence
7. Extracellular crosslinks
Life expectancy test
http://gosset.wharton.upenn.edu/mortality/perl/CalcForm.html
0
10
20
30
40
50
60
70
80
90
1850 1900 1950 2000 2050
U.S. Life Expectancy, 1850 – 2050e
83
77
69
50
39
http://www.methuselahmouse.org/
Research to repair and
reverse the damage of aging
The Methuselah Foundation
Source: http://www.infoplease.com/ipa/A0005140.html
Source: http://earthtrends.wri.org/text/population-health/variable-379.html
16. The Future of Technology
October 2007
16
Human body 2.0, 3.0
Redesign: the digestive system is rebuilt
Auto-nourishment via clothing
Nanobots go in and out of the skin cycling nutrients and waste
Digestive system and blood based nanobots supply precise nutrients
Eating becomes like sex, no biological impact, just for fun
Redesign: the heart is optional
Obsolete organs, heart, lungs, blood; nanobots delivering oxygen to
the cells, don’t require liquid-based medium
Two systems left
Upper esophagus, mouth and brain
Skin, muscle, skeleton and their parts of the nervous system
Sources: Ray Kurzweil, The Singularity is Near, http://lifeboat.com/ex/human.body.version.2.0
17. The Future of Technology
October 2007
17
Virtual worlds, 3D and simulation
Increasing demand for streaming video, data visualization
and 3D data display: learning, work and play
Simulation and augmented reality
Increasingly detailed capture of reality
Geospatialization: Google Earth, Nasa World Wind
Life capture, life logging
Virtual worlds explosion
MMORPG video games and interactive worlds
Participants: enterprise, education, government
Activities: interacting, collaborating, prototyping
Virtual reality 2.0: biofeedback, touch, taste, smell
Wild Divine
18. The Future of Technology
October 2007
18
Affordable space launch
Commercial payload launch
Space elevator
Sub-orbital human flight
Spaceport development
Extra-orbital robotic missions
International participation
Agency partnership
Prizes stimulate development
Peggy Whitson Pam Melroy
19. The Future of Technology
October 2007
19
The future depends on which coming revolution
occurs first
What will be the next Internet?
Artificial
Intelligence
Molecular
Nanotechnology
Anti-agingVirtual
Reality 2.0
Quantum
Computing
Robotics
Intelligence
Augmentation
Personal
Medicine
Affordable
Space LaunchFab
Labs
20. The Future of Technology
October 2007
20
Summary
We think about growth and change in linear,
exponential and discontinuous paradigms,
history is a chain of discontinuities
The realm of technology is no longer discrete, technology is
imbuing traditional linear phenomena with exponential and
discontinuous change
Computation (hardware and software) overview: Moore’s Law
improvements will likely continue unabated in hardware; software
however is stuck
Not only will there be linear and exponential growth in the next 50
years but probably also discontinuous change, possibly a change
with greater impact than the Internet in our (current) lifetimes
21. Thank you
Melanie Swan
Futurist
MS Futures Group
Palo Alto, CA
m@melanieswan.com
http//www.melanieswan.com
Slides: http//www.melanieswan.com/presentations
Provided under an open source Creative Commons 3.0 license
http://creativecommons.org/licenses/by-nc-sa/3.0/
22. The Future of Technology
Melanie Swan
Futurist
MS Futures Group
Palo Alto, CA
650-681-9482
m@melanieswan.com
http//www.melanieswan.com
Christine Peterson
Vice President and Founder
Foresight Nanotech Institute
Menlo Park, CA
650-289-0860
peterson@foresight.org
http://www.foresight.org
Liana Holmberg and Tess Chu
OS Wrangler & IP Gadfly and Developer
Linden Lab
San Francisco, CA
415-243-9000
liana@lindenlab.com
tess@lindenlab.com
http//www.lindenlab.com
Notas do Editor
51, 51, 153
Moore’s Law improvements will likely continue unabated in hardware, even with substrate changes
How the Internet has changed everything example: Efficiency – economics shifting – more resource allocation by market mechanism than by pecking order
Tech & comms = exponential/acc chg – for ex: Internet traffic in a new exponentiation with video traffic – YouTube = 7% Comcast traffic
The next mainstream
Challenge b/c we can see several possible discontinuities but right now it is hard to predict which will occur first…
Also whatever revolution happens first changes a) everything and b) the path to further discontinuities including obviation
Convergence
Perhaps at no time have we had so many possible coming revolutions – Catch 22 tautology – by definition discontinuous change cannot be predicted, so perhaps the next Internet-like revolution will not be one of these things at all, or maybe it will be some part of these areas that we aren’t able to think about now, ex: at some point cars/computers were predicted, but not for mass everyday use – quantum computing now seems like a gov’t encryption preserve – we aren’t thinking of the wristwatch model or human embedded possibilities
VR: phermones, culture and communication occurs at increasing levels of abstraction: desktop/files; 100 IM windows, video gaming, VR worlds, younger generations find less need for true physical presence
Nanomedicine – several diseases involve molecular damage inside the cell, such as from free radicals and radiation. Osteoporosis involves calcium loss from bones and is likely treatable by placing chemical and electrical props at the right places in the body
Nanotech (molecular mfg atomically correct from bottom up)
Nano – pathways biotech, high precision machining, chemical synthesis
Many imply a substantial shift in economics: robotics, AI, nano-mfg
Whatever the next Internet or coming revolutions are going to be, they are going to be computation-dependent
Next, let’s look at an overview of the current status of computation - HW and SW since this is driving realization of all other technology
Information Technologies (of all kinds) double their power (price performance, capacity, bandwidth) each 1-2 years
Do we need the exponentially growing hw power? Yes MRO mars reconnaissance orbiter sends back more than 1 TB / week, hi res photos, more than most planetary missions in whole life
Visualization
Zillionics (kevin kelly) massive info avail; changing science previously hypoth: measurement, now tons of meas.
ST: persistence of ICs with new materials, molecular mfg and 3d circuits
LT: quantum computers, Analog Quantum computers introduction in 2007 from Dwave Systems. In ten years, possibly quantum computers with multiple thousands to millions of qubits.
QC: lots of govt and corporate sponsored research, conferences, papers, activity, early commercialization – ROADMAP
Managing Moore’s Law limitations: Penryn definition changes, 3D molecular ICs;
9/18/07 – Moore says only 10 years left…implicit: in current paradigm
Otellini: http://www.internetnews.com/ent-news/article.php/3700336
Gate closed is 0, current doesn’t flow through
Intel’s move to 45nm process technology for its Core microarchitecture Penryn, made possible with the use of high-k and metal transistors
"the biggest change in transistor technology since the introduction of polysilicon gate MOS transistors in the late 1960s."
The new breakthrough involves the use of hafnium, an unstable silvery metallic element chemically related to zirconium. It is currently used mostly for control rods in nuclear reactors. Hafnium is far denser than silicon and can be etched to even greater degrees of precision, meaning chips can be made still more dense and more powerful.
Trend: graphics built onto the chip, not separately, enables the gaming and 3d simulation and virtual worlds
Nehalem, which will succeed Intel's Core microprocessor architecture, will have up to eight cores and each core will be able to process two threads simultaneously, giving each processor the capacity to process up to 16 threads at the same time.
http://www.internetnews.com/ent-news/article.php/3700336
Looking even farther down the road, Otellini said Intel will begin production using 32-nanometer process technology in 2009. He proudly showed attendees what he called the world's first 32-nanometer product, a 291-megabit array die with more than 1.9 billion transistors.
"This gives us the confidence to build a mainstream microprocessor in two years on this technology," he said, adding that 32-nanometer products will be designed using the second generation of its high-k metal gate transistor technology which replaces silicon dioxide with hafnium as an insulator to produce faster and more energy-efficient chips.
For a seemingly advanced society, we have not been doing a good job on SW
Diversity of claims regarding improvement
Large project failure
No reusable modules (like building and construction industry)
Open source ability to maintain and extend dominance over proprietary standards
Ada Lovelace is popularly credited as history's first programmer. She was the first to express an algorithm intended for implementation on a computer, Charles Babbage’s analytical engine, in Oct 1842
A SW PROB: Need qqch re: the internal mechanism building new concepts, new machinery and new skills to create new understanding
AI/human will be equivalent; transfer our software from meat-based processor to more capacious hardware
Some assumptions in the calcs; also since not 100% human brain focused on any task, only need 1/1000 the capacity to simulate which we do have
Important point is that the HW is not there yet; SW lagging even farther behind but with another few cycles could have adequate simulation power
Most think ~2015; ccortex
The 360-teraFLOPS machine handles many challenging scientific simulations, including ab initio molecular dynamics; three-dimensional (3D) dislocation dynamics; and turbulence, shock, and instability phenomena in hydrodynamics. It is also a computational science research machine for evaluating advanced computer architectures.
Will we have intelligent machines that can solve any problem or will we hack the brain to improve human intell; prob combo both
intelligent agent is a system that perceives its environment and takes actions which maximizes its chances of success
Applications: pattern recognition, robotics
Problems: the lack of raw computer power, the intractable combinatorial explosion of algorithms, the difficulty of representing commonsense knowledge and doing commonsense reasoning, the incredible difficulty of perception and motion and the failings of logic.
Definition: not conducting work at nanoscale or even 2d atom placing, 3d molecular/atomic specific placement from the bottom up; manipulating atoms as we do bits
Water bridge: http://www.physorg.com/news110191847.html
Manipulating atoms as we do bits: Discrete entities handled rapidly with digital control to reliably form new patterns
Molecule, Atom? Atomic precision about one angstrom, 1/10 nm
Molecular mill: www.e-drexler.com/p/04/04/0512molMills.html
Lots of plastic 3-d printing at the moment
Public health improves or degrades?
Consumerism increases/decreases; matter as entertainment; period decorations and events; SL b/c RL
Regulation re: production of weapons, disease, genetic material, etc.
Personalized medicine is the use of detailed information about a patient's genotype or level of gene expression and a patient's clinical data in order to select a medication, therapy or preventative measure that is particularly suited to that patient
Hall: Patching up/fixing genetic deficiency vs. augmentation, the same thing depending on where you are on the scale of human variation; Intelligence, ‘beauty’/symmetry, height; Susceptibility to sunburn; Oxygen in cells: anemic vs. super-athletic capability. Enhancement rationale matters? (e.g.; help Mt. Everest rescue workers (respirocytes) vs. beauty (vanity)
Digital People: from Bionic humans to androids by Sidney Perkowitz 2004 - 10%? Americans are augmented w. non-biological machinery embedded – terminology: cyborg/robosapien (scary) or pacemaker, diabetic pump, hip replacement, corneal implant, hearing aid, cochlear implant, prosthetic limbs, etc.
Augmenting intelligence-it’s a wonder what paper & pencil will do for multiplying 10 digit numbers
Human life is arbitrarily limited at present
bioMEMS blood-borne devices that deliver hormones such as insulin have been demonstrated in animals
Dr. Ron Kahn, Joslin Diabetes Center, id the fat insulin receptor (FIR) gene – controls accumulation of fat by the fat cells
Increasing demand for data display visually and in 3D, already seen in ubiquity of streaming video
3rd annual American Cancer Society Relay For Life - $90,000 raised – 40 teams
Video games (9m World of Warcraft players), Second Life (9 million residents 40,000 concurrency, $1m daily economy)
Science 2.0 – simulation replaces experiment, hypothesis post-facto
BCI – brain computer interface with SL; headpiece with electrodes sensing motor cortex movement; think about walking
Geospatialization of data, Google earth KML, nasa world view, VIS demand; aug reality
Virtual reality 2.0: more immersive, incorporating biofeedback, touch, taste, smell
Spaceport America (near Las Cruces NM, 2008-2010 construction), UAE
Earth is just an 8000 mi wide space ship with 10,000 feet of life support; get out of diapers and colonize other parts of spaces
X Prize Foundation announcing automotive prize in early 2008; clean fuel car that can reach 100 mpg
Elevator climbing (power beaming) & tether strength competition– used to have the cranes for tether climbing at the air show, but 400 foot crane not a good idea…
Virgin Galactic has sold 100 of its $200,000 sub-orbital flight slots
NASA says they spend $6,000 per pound but other estimates suggest real cost is $20,000-$35,000/pound, needs to be about $500/pound
NASA Centennial Challenges
The next mainstream
Challenge b/c we can see several possible discontinuities but right now it is hard to predict which will occur first…
Also whatever revolution happens first changes a) everything and b) the path to further discontinuities including obviation
Convergence
Perhaps at no time have we had so many possible coming revolutions – Catch 22 tautology – by definition discontinuous change cannot be predicted, so perhaps the next Internet-like revolution will not be one of these things at all, or maybe it will be some part of these areas that we aren’t able to think about now, ex: at some point cars/computers were predicted, but not for mass everyday use – quantum computing now seems like a gov’t encryption preserve – we aren’t thinking of the wristwatch model or human embedded possibilities
VR: phermpnes, culture and communication occurs at increasing levels of abstraction: desktop/files; 100 IM windows, video gaming, VR worlds, younger generations find less need for true physical presence
Nanomedicine – several diseases involve molecular damage inside the cell, such as from free radicals and radiation. Osteoporosis involves calcium loss from bones and is likely treatable by placing chemical and electrical props at the right places in the body
Nanotech (molecular mfg atomically correct from bottom up)
Nano – pathways biotech, high precision machining, chemical synthesis
Many imply a substantial shift in economics: robotics, AI, nano-mfg
Moore’s Law improvements will likely continue unabated in hardware, even with substrate changes