Didya Know
[Another HV twiticism©™®: discrete daily data deliveries in 140K-char chucks; whereas most tweets are rapid & vapid, our twiticisms are slow and substantive. The first was from Bucky Fuller. Next up…]
“Did You Know (Shift Happens)”:
The top 10 in-demand jobs in 2010 didn’t exist in 2004.
We are currently preparing students for jobs that don’t yet exist…
Using technologies that haven’t been invented
Provocative statistic, no. Or more accurately: Provocative statistic? No.
More like a mythstistic. The info above is from the popular “Did you Know?” video. It turns out, tho, the “fact” was wrong when first uttered in 2004, and remains wrong today. The “top 10 in-demand jobs” today all existed in 2004 — half existed in 1904 (nurses, teachers). See the excellent essay “Fact or Fiction? Maintaining Credibility.”
Like many Internet Age “stats,” we want to believe, it matches our worldview, so it must be true. We let perception trump reality, and perpetuate concepts seen on YouTube, heard in some conference, or read on some blog. No research, no fact-checking, just repetition.
With email, txtng, IM, FB, tweets, etc., we are likely reading and writing more lately (not less, as some believe). We’re all becoming de facto published journalists. Perhaps over time we will develop some journalistic instincts, like tracing info to its independently verifiable source, not gulping down data from far downstream — where it’s become tainted and polluted.
“I think we’re in the midst of a literacy revolution the likes of which we haven’t seen since Greek civilization.” –Professor Andrea Lunsford (Stanford Study of Writing), quoted in Wired “The New Literacy“
Many millions of us have enjoyed the “Did You Know” videos (at bottom of page). HV especially liked its lineage: a humble start as a high-school slideshow, later animated, xplaned, 2.0ed, 3.0ed, funny 4.0ed, f’real 4.0ed, and parodied by Barely Political:
The original video is fact-packed. HV figured some may want to access the info w/o having to watch the whole movie. We began this project simply to put the text of the several video versions on a single webpage, accompanied by sources and updates.
In following those sources, tho, we were led less toward “Did You Know?” than “How Do You Know?” — what makes us think we know what we think we know? So, on this page and on our twit feed, HV presents…
“Did You Know?” Statements, Sources, Updates, & Errata
[Clicked source-pages display in frame at bottom (right-click to open instead in a new window/tab).]
0. Click a statement below to show/hide source(s):
1. If you’re one in a million in China, there are 1,328 people just like you. In India, there are 1,166 people just like you. [updated]
CIA World Factbook Population Estimates (July 2009): 1 China 1,338,612,968 2 India 1,166,079,217 3 European Union 491,582,852 4 United States 307,212,123 5 Indonesia 240,271,522
2. The 25% of the population in China with the highest IQ’s exceeds in number the total population of North America.
By 2012, the smartest 25% of children in India will be
200 times smarter than the fattest 10% of children in America.
By 2015, the former will be eaten by the latter.”
—Barely Political parody video: Did You Know?
3. China will soon become the number one English speaking country. [wrong]
Wikipedia: List of Countries by English-speaking Population By Total English (Native and Second Language) Speakers: 1 United States 251,388,301 2 India 90,000,000 3 Nigeria 79,000,000 4 United Kingdom 59,600,000 5 Philippines 48,800,000 …18 China 10,000,000
These numbers are hard to pin down. See the talk page for this article, which includes this China-related explanation:
Wikipedia’s India estimate of 350 million includes two categories — “English Speakers” and “English Users”. The distinction between the Speakers and Users is that Users only know how to read English words while Speakers know how to read English, understand spoken English as well as form their own sentences to converse in English. The distinction becomes clear when you consider the China numbers. China has over 200~350 million users that can read English words but, as anyone can see on the streets of China, only handful of million who are English speakers.
4. If you took every single job in the U.S. today and shipped it to China, China would still have a labor surplus. [no longer true]
[Original source cited by "Did You Know?" was from former Maine Governor Angus King 2005 presentation at SEED (Spreading Educator to Educator Developments podcast | notes)]
5. Today:
11,631 babies will be born in the United States. (7,053 people will die.)
51,344 babies will be born in China. (25,892 people will die.)
69,517 babies will be born in India. (19,903 people will die.)
368,314 babies will be born on Earth. (4,718,562 people will die.)
CIA World Factbook Birth Rate (2009 est.) per 1,000 population (Rank Country Births/1,000 Population) 153 United States 13.82 150 China 14.00 87 India 21.76
CIA World Factbook Death Rate (2009 est.) (Rank Country Deaths/1,000 Population) 102 United States 8.38 130 China 7.06 155 India 6.23
US Census Bureau World Vital Events Per Time Unit 2009 (Figures may not add to totals due to rounding) ------------------------------------------------- Natural Time unit Births Deaths increase ------------------------------------------------- Year 134,434,533 56,622,740 77,811,793 Month 11,202,878 4,718,562 6,484,316 Day 368,314 155,131 213,183 Hour 15,346 6,464 8,883 Minute 256 108 148 Second 4.3 1.8 2.5 -------------------------------------------------
6. The average young U.S. baby-boom worker held 10.8 jobs between the ages of 18 and 42. (Two-thirds of these jobs were held from ages 18 to 27.) [updated]
US Dept of Labor: Bureau of Labor Statistics:
National Longitudinal Surveys ("Number of Jobs Held in a Lifetime')
Average number of jobs for persons ages 18 to 42
in 1978-2006, by Ages:
Total 18-22 23-27 28-32 33-37 38-42
10.8 4.4 3.3 2.6 2.2 1.9
The above DOL: BLS: NLS reports: “The average person born in the later years of the baby boom held 10.8 jobs from age 18 to age 42.” Not surprising since: “nearly two-thirds of these jobs were held from ages 18 to 27.”
And contrary to common Internet Age expectations, this average number of jobs seems not to have shifted much the past two decades:
DOL: BLS "Median years of tenure with current employer for employed wage and salary workers by age and sex, selected years, 1996-2008" 1996 1998 2000 2002 2004 2006 2008 3.8 3.6 3.5 3.7 4.0 4.0 4.1
Also see DOL: “Employee Tenure Summary.”
7. Length of time with the same employer- Ages 18 to 22: 72% of their jobs ended in less than a year; 94 percent ended in fewer than 5 years. Ages 38 to 42, 31 percent ended in less than a year and 65 percent ended in fewer than 5 years.
US Dept of Labor: Bureau of Labor Statistics: National Longitudinal Surveys: "Number of Jobs Held in a Lifetime" Cumulative percent distribution of duration of completed employment relationships Ages <1year <2yrs <5yrs <10yrs <15yrs 18-22 72.3 85.2 94.1 97.1 98.0 23-27 59.2 75.9 88.8 94.0 95.7 28-32 52.5 69.7 85.5 91.6 93.6 6.2 33-37 42.8 60.7 80.6 88.2 88.9 11.1 38-42 30.5 46.6 65.1
Although job duration tends to be longer the older a worker is when starting the job, these baby boomers continued to have large numbers of short-duration jobs even as they approached middle age. Among jobs started by workers when they were ages 38 to 42, 31 percent ended in less than a year, and 65 percent ended in fewer than 5 years.
—”Number of Jobs Held, Labor Market Activity, and Earnings Growth among the Youngest Baby Boomers: Results from a Longitudinal Survey” (pdf) June 2008
["DYK?" vid quotes Ian Jukes in "From Gutenberg to Gates to Google (and beyond); Education for the Online World" (pdf). Jukes asserts the erroneous: "The US Department of Labor now estimates that today’s learners can expect to have 10 to 14 career – not 10 to 14 jobs, but careers."
That'd be news to the DOL:
The Bureau of Labor Statistics (BLS) never has attempted to estimate the number of times people change careers in the course of their working lives. The reason we have not produced such estimates is that no consensus has emerged on what constitutes a career change.
—US Dept of Labor: BLS: FAQHow many times does the average worker change careers? Statistically speaking, no one knows. One idea that is commonly -- but incorrectly -- attributed to the U.S. Department of Labor is that people change careers about seven times in a lifetime. But the Labor Department does not gather that kind of data.
—David Terkanian, DOL: BLS economist, "Lifetime 'career' changes," DOL's Occupational Outlook Quarterly, Summer 2006
8. According to former Secretary of Education Richard Riley, the top 10 in-demand jobs in 2010 didn't exist in 2004. [wrong]
US Dept of Labor & Dept of Educaton: CareerVoyages.gov: "Top 50 In-Demand Occupations" Projected Need for Employees Projected 2006 Media Top In-Demand Occupations (2006 - 2016) Growth Hourly Wage --------------------------------------------------------------------------- 1 Registered Nurses 1,001,000 18-26% $29 2 General and Operations Managers 441,000 0-8% $43 3 Physicians and Surgeons 204,000 9-17% >70 4 Elementary School Teachers 545,000 9-17% Annual: $47K 5 Accountants and Auditors 450,000 18-26% $27 6 Computer Software Engineers 300,000 27+% $40 7 Lawyers 228,000 9-17% >70 8 Sales Representatives 476,000 0-8% $24 9 Computer Systems Analysts 280,000 0-8% $35 10 Management Analysts 264,000 18-26% $34
The source of this oft-repeated miscalculation is a 2004 book, now out-of-print, The Jobs Revolution: Changing How America Works. It quotes fmr Edu Sec Riley: “None of the top 10 jobs that will exist in 2010 exist today, and these jobs will employ technology that hasn’t been invented to solve problems we haven’t yet imagined.”
Turns out Riley guessed wrong in 2004. So why do we continue to perpetuate the mistake in 2009?
Read how one guy took on the task of tracking down some facts: “Fact or Fiction? Maintaining Credibility,” Retail Industry Leaders- Newsletter 2009-04-06:
So, I went in search for the original comments from former Secretary of Education Richard Riley. It turns out that he did say this and was quoted in a book titled: The Jobs Revolution: Changing How America Works. But, here is the important fact that solves the mystery — the book was published in 2004. If you go back and look at the quote, it now seems clear what has happened.
In 2004, Riley was making a prediction that he likely intended to make a point about how the pace of change would continue in the upcoming years. That is far different than reporting his prediction as a fact in 2009 when it has clearly not come true.
9. We are currently preparing students for jobs that don’t yet exist, using technologies that haven’t been invented, to solve problems we don’t even know are problems yet. [NOT]
10. Name this country: Richest in the World, Largest Military, Center of world business and finance, Strongest education system, World center of innovation and invention, Currency the world standard of value, Highest standard of living. The country is: England… In 1900. [maybe]
First of all, England is a country on an island — Great Britain — which is part of the United Kingdom. So when DYK? sez England, let’s assume they mean the UK.
The UK’s military, the British Armed Forces was likely the strongest in 1900. But the “largest military” may have been China’s. The Qing Dynasty threw 600K at the First Sino-Japanese War, 1894-5. The closest I could find for the Brits was 250K in the 1853-6 Crimean War.
All this info is from Wikipedia, btw, so second-hard at best. We’ve gone google-eyed trying to find better figures, and are now giving up: maybe someone else will have more success — the best search term we tried was “troop strength.”
FYI, the British Empire was the largest ever, at its 1920 peak reigning over 25% of the world’s population and 25% of the Earth’s landmass — just squeaking by the Mongols, and a good 10M km2 larger than the various Russia/USSR monoliths.
As to what country was “Richest in the World” in 1900: the metric is GDP, or GDI, once called GNP, soon to be PPP, for which the IMF and CIA have differing data, but neither had nada for 1900. ‘Course for “Highest standard of living” take all above and add “per capita.” Couldn’t come up w/ bupkis on that either. Case closed on this one.
11. The U.S. is 20th in the world in broadband Internet penetration, behind Belgium and Estonia. [update]
Rank Country HH-pen 1 South Korea 95% 2 Singapore 88% 3 Netherlands 85% 4 Denmark 82% 5 Tailwan 81% ...20 USA 60%
12. In 2003 Nintendo invested 140M $USD in research and development. [true] The U.S. Federal Government spent less than half as much on Research and Innovation in Education. [false; US Dept of Edu 2003 budget- Research: $229M]
Budget of the United States Government: Fiscal Year 2007
Department of Education (pdf): Spending
2007 Estimate- Research and Statistics….. $554M USDNintendo Annual Report: Consolidated Statements of Income (pdf)
2007 Actual- Research and development expenses….. $370M USD
In 2003 those research-dollars totaled $140M for Nintendo and $229M by US.Dept.Edu. Starting 2008, the US dropped an Edu line-item for Research and Statistics. The word “research” does appear in the description of a new program:
“The American Competitiveness Initiative includes an array of research and education programs that focus on improving the Nation’s long-term economic competitiveness.”
Estimated 2007 spending on ACI programs: $412M.
13. One out of every eight couples married in the U.S. last year met online. [doubtful]
Nineteen per cent of the couples met online compared with 17 per cent who met at work and 17 per cent who met through friends.… In the current survey, 31 per cent of married couples aged 45 to 54 met online, against 18 per cent of 20 to 44-year-olds.
—New Scientist “Go online to find your future spouse” August 2008.
Trouble is, no info is given on whether this survey was a scientific polling of a random sample or a “self-selected opinion poll” — known as SLOP in survey circles, for good reason: with SLOP the sample isn’t random, so the results aren’t representative of the general population.
Because the sponsor is a company with a stake in the results, the survey is worthy of heightened scrutiny. Neither the newsroom of eH or HI mentions the survey.
I’m no Nate Silver, but I do know that without info on methodology, margin of sampling error, or even a list of the survey questions, it’s impossible to gauge the statistical value of the results — if any.
And, let’s face it, asking an online-only audience about whether they met online kinda skews the results. What about the nearly 20% of US that aren’t online? Far as I can see, they aren’t even considered. One more thing: how many married folk do you know that met online? Is your figure anywhere close to the survey’s “31% of married couples aged 45-54 met online”?
When you see the words “online survey” chances are that it isn’t a probability-based sample, although it could be. You will want to find out more about how the respondents were selected, particularly if the survey claims to be representative of all adults in the U.S.
—Understanding and Interpreting Polls News University (online course- reg required)
See: American Association for Public Opinion Research “Questions to Ask When Writing About Polls” and National Council on Public Polls “Principles of Disclosure” and “20 Questions A Journalist Should Ask About Poll Results“.
14. More than 300 million people use Facebook. MySpace has 123 million users — the average MySpace page is visited 30 times a day. [update]
“Facebook now serves 300 million people across the world.… We’re also succeeding at building Facebook in a sustainable way. Earlier this year, we said we expected to be cash flow positive sometime in 2010, and I’m pleased to share that we achieved this milestone last quarter.”
—”300 Million and On” Facebook blog, Mark Zuckerberg, September 15 2009
“In May, Facebook had more than 307 million users, compared to MySpace’s 123 million.”
—”More Americans Go To Facebook Than MySpace” PC Magazine [quoting comScore stats]
As to the figure– 30/dy-visits to avg Space page: That comes from these these notes to an iBreakfast talk by Shawn Gold, MySpace Senior VP of Marketing & Content, titled “Marketing in a Networked Culture.”
Interesting related info: TechCrunch “Modeling The True Value Of Social Networks: 2009 Edition”
To put this social net hype in perspective, however, do remember that folk’s current financial fanaticism for Facebook was just as fervently flung a mere two years ago at MySpace — which has since laid off half its staff (”MySpace Fires Another 300” Business Insider).
The Internet Big Picture: Internet Usage Statistics:
World Internet Users and Population Stats
Population: Internet Users: Penetration: Growth: (2009 est.) Dec 2000 Aug 2009 %-Population 2000-2009 6,767,805,208 360,985,492 1,668,870,408 24.7% 362.3%
15. If MySpace were a country, its inhabitants would live inside computer chips.
16. There are 7 billion searches on Google each month. To whom were these questions addressed B.G.? (Before Google)
Rank: Provider: Searches (000): Share of Searches" - All Search 10,812,734 100.0% 1 Google 6,986,580 64.6% 2 Yahoo! 1,726,060 16.0% 3 MSN/Windows Live/Bing 1,156,415 10.7% 4 AOL 333,231 3.1% 5 Ask.com 186,270 1.7%
When DYK? video came out in 2006 there were: 2.7B Google searches/month.
17. We humans send 2,000,000,000,000 (that’s 2 trillion) text messages… daily.
18. There are about 615,000 words in the English language. About 5 times as many as during Shakespeare’s time. [update]
Number of entries: 291,500
Number of word forms defined and/or illustrated: 615,100.
There’s a range for the numbers of English words. OED, above, lists 292K words in 615K word-forms. Global Languiage Monitor sez 1,001,543 (Sept 2009). Here’s other estimates: Number of Words in the English Language.
This last source estimates 50-60K words in Shakespeare’s time, aka, in the Early Modern English language. Can’t confirm.
19. More than 3,300 new books are published… Daily.
Compare to Worldmapper: “In 1999 there were a million new book titles worldwide.” Worldmapper 1999 data, Books Published: 1 million; Books Borrowed: 6156 million.
20. In 2008 humans created, captured, and replicated 487 exabytes of digital information (that’s almost 5 x 1020; 1 EB = 1 billion gigabytes).
Amount of Digital Information- created, captured, and replicated In Exabytes (1EB = 1 billion gigabytes = 1,000,000,000,000,000,000 bytes) 1983 1996 2000 2003 2006 2007 2008 -------------------------------------------------------- 000.09 000.17 2.8 5.6 161 281 487
Sources: Digital Universe white papers by IDC- sponsored by EMC:
May 2009 “As the Economy Contracts, the Digital Universe Expands” (pdf | kinda kewl multimedia white-paper); March 2007 “The Expanding Digital Universe: A Forecast of Worldwide Information Growth Through 2010″ (pdf)
[See RayOnStorage's skeptical analysis of IDC's "Exabyte-a-day" stat.]
Information Management and Systems, University of California at Berkeley:
October 2003 “How Much Information? 2003”
Latanya Sweeney, Computer Science, Carnegie Mellon University:
2001 “Information Explosion” (pdf)
Perspective (1 zettabyte = 1,000 exabytes):
000.5 zettabytes = the amount of human-manufactured info in 2008.
300 zettabytes = “the amount of visual information conveyed from the eyes to the brain of the entire human race in a single year.” —Cisco Visual Networking Index – Forecast Q&A: June 2009.
Related: Minnesota Internet Traffic Studies (MINTS)
Year-end 2008 monthly Internet traffic estimate (exabytes) U.S. 1.2 - 1.8 World 5 - 8 Current (year-end 2008) annual Internet traffic growth rates U.S. 50-60% World 50-60%
And: the Language Log’s “More on the 5 Exabyte Mistake.”
21. The amount of digital information created by humans every recent year exceeds the total for all previous years combined.
22. Bell Labs optically transmitted 100 trillion bits per second over 7,000 kilometers (like sending 400 DVDs per second over transoceanic cable).
Paris, September 28, 2009 – Alcatel-Lucent (Euronext Paris and NYSE:ALU) today announced that scientists in Bell Labs, the company’s research arm, have set a new optical transmission record of more than 100 Petabits per second.kilometer (equivalent to 100 million Gigabits per second.kilometer). This transmission experiment involved sending the equivalent of 400 DVDs per second over 7,000 kilometers, roughly the distance between Paris and Chicago. This is the highest capacity ever achieved over a transoceanic distance and represents an increase that exceeds that of today’s most advanced commercial undersea cables by a factor of ten. To achieve these record-breaking results the Bell Labs researchers made innovative use of new detection techniques and harnessed a diverse array of technologies in modulation, transmission, and signal processing.
Old NTT records: “14 Tbps over a Single Optical Fiber: Successful Demonstration of World’s Largest Capacity — 140 digital high-definition movies transmitted in one second” —Nippon Telegraph and Telephone Corporation Science and Core Technology Laboratory Group September 29 2006 News Release (see also September 25 2008.
Apparently, NTT has also achieved teleportation. Beam me up.
23. 300 million new personal computers shipped worldwide in 2008.
The latest forecast meant that global shipments of PCs, which include desktops, laptops and low-cost netbooks, will likely hit 301.5 million units this year, up from 299.4 million units in 2008, based on iSuppli figures.
EETimes.com –
Notebooks to outsell desktops in ‘09, says iSuppli
2009-07:
Global PC shipments are expected to decline to 287.3 million units in 2009, down 4 percent from 2008, the first year-to-year contraction for PC shipments since the Dot Com bubble burst, according to a revised forecast by iSuppli.…
ISuppli predicts that shipments of desktop PCs will decline to 124.4 million in 2009, down 18 percent from 151.9 million units in 2008. Entry-level servers — which iSuppli includes in its definition of PCs — also will suffer a decline, with shipments falling to 6.9 million units, down 9.5 percent from 7.7 million in 2008, the firm said.
Notebook PC shipments in 2009 will rise by 11.7 percent to reach 155.97 million units, up from 139.6 million in 2008, iSuppli said.
iSuppli Notebook PC Shipments Exceed Desktops for First Time in Q3
24. The $100 laptop project was expected to ship 50-100 million laptops per year to children in underdeveloped countries. [not yet]
“At least 750,000 kids — a number that will double by June, Chairman Nicholas Negroponte said — are tapping away on One Laptop per Child-produced XO laptop keyboards.” —CNN Laptops bring lessons, maybe even peace” 2009-03:
“Only about a million OLPCs have shipped so far.” —UN Dispatch “One Laptop Per Child – The Dream is Over” 2009-09.
Also see: OLPC News.
25. By 2023 the computation capability of a supercomputer will exceed that of the Human Brain. [maybe]
| Name | FLOPS |
|---|---|
| petaFLOPS | 1015 |
| teraFLOPS | 1012 |
| gigaFLOPS | 109 |
| megaFLOPS | 106 |
Neither quantifying the brain’s computational powers nor predicting a future supercomputers specs is easy, or possibly accomplishable.
The original DYK? prediction mistakenly used the year 2013. But their source, Ray Kurzweil’s 2001 “The Law of Accelerating Returns” (full excerpt below), sez: “We achieve one Human Brain capability (2 * 10^16 cps) for $1,000 around the year 2023.”
Kurzweil defines his “cps” as “calculations per second,” and conflates it with FLOPS, one of two common computional metrics:
FLOPS = FLoating point Operations Per Second.
MIPS = Millions of Instructions Per Second.
Kurzweil’s 2001 essay also forecast “IBM’s ‘BlueGene’ supercomputer, now being built and scheduled to be completed by 2005, is projected to provide 1 million billion calculations per second” (i.e., 1 petaFLOPS).
BlueGene never hit petaflops-itude. But its IBM successor, Roadrunner, did in 2008:
The new No. 1 system, built by IBM for the U.S. Department of Energy’s Los Alamos National Laboratory and called “Roadrunner,” achieved performance of 1.026 petaflop/s—becoming the first supercomputer ever to reach this milestone. At the same time, Roadrunner is also one of the most energy efficient systems on the TOP500.
—TOP500 Supercomputing Sites | June 2008
Here’s the top five supercomputing speed demons so far this year, and the winner for 2005-2008:
Now here’s Kurzweil’s 2001 cortical calculations and suputer projections:
Here’s My estimate of brain capacity is 100 billion neurons times an average 1,000 connections per neuron (with the calculations taking place primarily in the connections) times 200 calculations per second. Although these estimates are conservatively high, one can find higher and lower estimates. However, even much higher (or lower) estimates by orders of magnitude only shift the prediction by a relatively small number of years.
Some prominent dates from this analysis include the following:
- We achieve one Human Brain capability (2 * 10^16 cps) for $1,000 around the year 2023.
- We achieve one Human Brain capability (2 * 10^16 cps) for one cent around the year 2037.
- We achieve one Human Race capability (2 * 10^26 cps) for $1,000 around the year 2049.
- We achieve one Human Race capability (2 * 10^26 cps) for one cent around the year 2059.
Others have expressed gray matter in MIPS:
It takes robot vision programs about 100 computer instructions to derive single edge or motion detections from comparable video images. 100 million instructions are needed to do a million detections, and 1,000 MIPS to repeat them ten times per second to match the retina.
The 1,500 cubic centimeter human brain is about 100,000 times as large as the retina, suggesting that matching overall human behavior will take about 100 million MIPS of computer power.
—Hans Moravec “When will computer hardware match the human brain? 1997
The above Moravec likely ended up in this oft-repeated mis-quote:
A human brain’s probable processing power is around 100 teraflops, roughly 100 trillion calculations per second, according to Hans Morvec, principal research scientist at the Robotics Institute.
—Michelle Delio This Is Your Computer on Brains Wired 2002-11-19
She seems to have transposed his “100 million MIPS” to her “100 teraflops”? MIPS and FLOPS measure different things so don’t convert.
But the mis-quote stuck — there’s 100s of Google hits for the exact phrase: “A human brain’s probable processing power is around 100 teraflops.”
She also repeatedly misspelled “Moravec” as “Morvec” and Wired posted this warning atop above article:
Reader’s advisory: Wired News has been unable to confirm some sources for a number of stories written by this author.
Once again this page’s info-quest becomes a hunt for how information originates, morphs, and moves. So, moving on…
Notice the suputer specs above are in FLOPS, so we’ll need a FLOPS estimate of cerebral power for comparison. (MIPS are now known as Meaningless Indicator of Processor Speed, aka, not the preferred metric.) Oxford University’s Future of Humanity Institute put out a Brain Emulation Roadmap (Nick Bostrom and Anders Sandberg 2008) with “Estimates of the computational capacity/demands of the human brain” (Appendix A, page 84 pdf | scribd). Some figures are in FLOPS. Of them, Kurzweil’s above 2*1016 FLOPS (20 petaFLOPS) is by far the lowest. Others range from 1018 to 1028 FLOPS.
So when computers will catch up? Dunno. But IBM Roadrunner’s recent record is a mere 1015 FLOPS, i.e., many orders of magnitude to go.
26. By 2049 a $1,000 computer will exceed the computational capabilities of the human species. [who knows?]
Already, IBM’s “Blue Gene” supercomputer, now being built and scheduled to be completed by 2005, is projected to provide 1 million billion calculations per second (i.e., one billion megaflops). This is already one twentieth of the capacity of the human brain, which I estimate at a conservatively high 20 million billion calculations per second (100 billion neurons times 1,000 connections per neuron times 200 calculations per second per connection). In line with my earlier predictions, supercomputers will achieve one human brain capacity by 2010, and personal computers will do so by around 2020. By 2030, it will take a village of human brains (around a thousand) to match $1000 of computing. By 2050, $1000 of computing will equal the processing power of all human brains on Earth.
Of course, this only includes those brains still using carbon-based neurons. While human neurons are wondrous creations in a way, we wouldn’t (and don’t) design computing circuits the same way. Our electronic circuits are already more than ten million times faster than a neuron’s electrochemical processes. Most of the complexity of a human neuron is devoted to maintaining its life support functions, not its information processing capabilities. Ultimately, we will need to port our mental processes to a more suitable computational substrate. Then our minds won’t have to stay so small, being constrained as they are today to a mere hundred trillion neural connections each operating at a ponderous 200 digitally controlled analog calculations per second.
But as pointed out in #25 above, his back-of-napkin brain power estimates were several orders of magnitude lower than any other.
Well, loyal readers: That’s it for our DYK? updates, sources and errata. Here’s some miscellanea…
For help getting started on the last couple source-checks, much thanks to Doctor Recommended for “FLOPS, MIPS, Watts and the Human Brain.”
Keep in mind, tho, some like neither MIPS nor FLOPS as computational metrics: “Of all the misleading performance metrics out there, MIPS and GFLOPS ratings have got to be two of the most widespread.” —Ars Technica “Behind the benchmarks” 1999.
In Feb 2002 Smart Computing attempted to aggregate computer/brain comparisons in “The Great Race: CPU Speeds Close In On The Human Brain” (pdf).
The 2003 Oracle- Think Quest project’s Computer vs The Brain had this version of Moravec’s 100 MMIPS estimate in “Man Vs Machine:”
Overall, the retina seems to process about ten one-million-point images per second.
Because the 1,500 cubic centimeter human brain is about 100,000 times as large as the retina, by simple calculation, we can estimate the processing power of a average brain to be about 100 million MIPS (Million computer Instructions Per Second ). In case you’re wondering how much speed that is, let us give you an idea.
1999’s fastest PC processor chip on the market was a 700 MHz pentium that did 4200 MIPS. By simple calculation, we can see that we would need at least 24,000 of these processors in a system to match up to the total speed of the brain !! (Which means the brain is like a 168,0000 MHz Pentium computer). But even so, other factors like memory and the complexity of the system needed to handle so many processors will not be a simple task. Because of these factors, the figures we so childishly calculated will most probably be a very serious underestimate.
The current Blue Brain Project is “the first comprehensive attempt to reverse-engineer the mammalian brain, in order to understand brain function and dysfunction through detailed simulations” (Ecole Polytechnique Federale de Lausanne in Lausanne, Switzerland; Henry Markram, project Director). Here’s one of their FAQ:
Is the brain like a computer?
In some ways yes, but in most ways it is not at all like a computer. The brain performs many analog operations which cannot be performed by computers and in many cases it achieves hybrid digital-analog computing. The most important feature of the brain that makes it different from computers is that it is constantly changing. If the resistors and capacitors in a computer started changing, then it would immediately malfunction, whereas in the brain such equivalent properties change constantly on the time scales of milliseconds to years. The brain is more like a dynamically morphing computer. We are still far from understanding the rules that govern the brain’s genetically and environmentally driven self-organization in response to external stimulus.
As for state-of-art silicon, this from Wikipedia IBM Roadrunner:
The US$133-million Roadrunner is designed for a peak performance of 1.7 petaflops, achieving 1.026 on May 25, 2008… Roadrunner occupies approximately 560 square metres (6,000 sq ft) and became operational in 2008. The DOE plans to use the computer for simulating how nuclear materials age in order to predict whether the USA’s aging arsenal of nuclear weapons are safe and reliable. Other uses for the Roadrunner include the sciences, financial, automotive and aerospace industries.
So Santa won’t be bringing you that this year, but you can “Tap Into the 12-Million-Teraflop Handheld Megacomputer | Inforporn” — Kevin Kelly, Wired 2008.
That’s all, folks. Thanks for reading.
[Click a numbered statement above to show/hide source(s). Clicked source-pages display in frame below (right-click to open instead in a new window/tab. This link brings back the DYK? video playlist.]
[...] our new Didya Know page, we’re breaking down the stats & sources from the popular fact-filled “Did You [...]
Leave a comment: