| Wednesday,
August 13, 2003 I put up a lot of
mail last night: scroll up if you have not seen it. And now an exciting
prospect?
subject: a hot diamond cpu
Dear Dr. Pournelle:
Please check http://www.wired.com/wired/archive/11.09/diamond.html
. The article discusses two methods of making diamonds. The Gemesis
process uses heat and pressure to imitate nature and the Apollo process
uses vapor deposition, which I remember from semiconductor manufacture.
Both companies are selling jewel quality diamonds to finance research into
diamond semiconductors.
Sale of manufactured jewel diamonds upsets De Beers.
I know from a previous article they can always spot Gemesis manufactured
diamonds with a 150 power microscope. The flaws look like tiny staples at
that magnification..The Apollo diamonds are harder to spot, but are more
pure than any known mined diamond. De Beers is in trouble, and knows it.
When is the last time Mrs. Pournelle stuffed a 150-power microscope into
her purse when you offered to take her diamond shopping? More important,
if you laid out a one carat ring and a two carat ring side by side,
knowing they were the same price, which would she choose?
But Chaos Manor is about computers. Picture a cpu
made from diamond semiconductor. It will be stable and usable at two
hundred degrees Fahrenheit, Hot enough to destroy any silicon chip. It
will also be faster than present semiconductors, giving Moore's law of
speed doubling a few more years of life. They can make doped diamond
semiconductors now. The process can and will be perfected. Doped diamonds
in your laptop, what a prospect!
regards,
William L. Jones wljones@dallas.net
A long time ago I bought some investment
grade diamonds, in the mistaken belief they were a good investment that
could be worn until liquidated. They were a good investment in the sense
that for a number of years after that I could have sold them for more than
I paid, but I had not reckoned on "You want to sell my
diamonds?"
Man made diamonds threatens the entire
diamond market structure, which is already artificial and came close to
collapse when the Russian collections got into the market.
Diamonds are very useful, and diamond films
may have a lot to do with heat dissipation. Their investment future is
uncertain. And see below.
From Ed Hume:
Unholy Matrimony: Spam and Virus
From http://businessweek.com/technology/content/
aug2003/tc20030812_7863_tc047.htm
Excerpt: Over the past few months, e-mail security
companies have seen mounting evidence that spammers are using
virus-writing techniques to assure that their sales pitches get through.
At the same time, intrepid virus writers have latched onto spammers'
trusty mass-mailing techniques in an effort to wreak widespread digital
mayhem. . . .
Is it time to panic?
Also from Ed Hume:
In a time when manufacturing, IT and other white
collar jobs are leaving the country, there is one place a healthy young
person with a brain can find employment:
PEACETIME OPERATIONS: Recession Recruiting Rocks
From http://www.strategypage.com/fyeo/howtomakewar/
default.asp?target=htpeace.htm
August 12, 2003: What's bad for the economy is good
for the military. With an all volunteer force, the military has to compete
in the job market for new recruits. That effort was taking a beating in
the late 1990s, and probably will again as the economy recovers from the
millennium Internet bubble. Not only are more men and women applying now,
but more first term troops are deciding to stay in. This means that not as
many new recruits are needed each year. Six months ago, the Navy thought
it would need 48,000 new recruits this year. That goal has been scaled
back to 39,971, and will probably go a little lower before the year ends.
Nearly a hundred thousand sailors are returning from the Middle East and
have been told that new deployment plans will keep them at home more often
then in the past. So morale is high, and the civilian job prospects do
look very bright. Thus sailors are staying in uniform. And there's also
the 911 factor. The nation is technically at war, and people in the armed
services are more aware of it than civilians. Everyone is standing a bit
taller, and less eager about getting out of the service.
Overall, the quality of the troops is going up. It
was pretty high to begin with in the late 1990s, but the recruiters can
afford to be more picky now, and they are. Troop commanders know they
won't take a lot of heat from up the line if they toss a troublesome troop
out, and the troops know it too. America has probably never had such a
high quality and combat experienced military before (when there wasn't a
major war going on). Even World War II depended on a lot of guys who
wouldn't be allowed to enlist today. The weapons and equipment are more
capable than ever before, and require brighter troops to make them work.
===============
reader, "P" responded to
"Affirmative Action: The Robin Hood Effect" http://www.lagriffedulion.f2s.com/robinhood.htm
Jerry, please pass my comments back to him.
P: Amateur analysis of La Griffe du Lion essay (not a statistician, no
background in demographics or economics)
Griffe: Apologies are unnecessary.
P: Interesting idea, very poor methodology. I suspect that significant
deviations from a meritocracy exist in current income distributions,
however, I believe "racial Robin Hood" effect is less pronounced
than La Griffe calculates and that (female subsidizing male) gender Robin
Hood effect is probably more pronounced despite affirmative action in
opposite direction. There are many, many significant factors in current
personal income distribution besides racial affirmative action and
intelligence. There are significant other factors besides NLSY tested
intelligence which would be paid for in a meritocracy some of which
(whether racial characteristics or not) I contend track strongly with race
in today's population.
Griffe: "Believe," "suspect, and
"contend" are not part of the Griffian lexicon. That said, P
makes some addressable points.
P: Potential problems with La Griffe methodology: 1) Very limited in
adjusting for recognized factors other than IQ and affirmative action in
present actual income distribution; while it raises some pertinent
questions about other factors it dispenses with most of these by asserting
that divvying folk up into quintiles will remove these factors from
significance.
Griffe: Of all variables, IQ is the strongest
correlate of income. There is no longer any doubt that income stratifies
by IQ. (See, for example, TBC or Gottfredson: "Why g Matters",
Intelligence 24(1) 1997.) Dividing the workforce into quintiles reduces
the effect of other factors; it does not remove them. There is bleeding
across quintile boundaries, but keep in mind, P, this is a model, not an
exact representation of economic reality.
P: Factors recognized and dispensed with this way are: Variation in
individual ambition/drive, unmaterialistic personal preference in job
type, opportunity (market imperfection, not every job is available to
every person), Geography (West Virginia/New York pay scale), relevant
personal qualities beyond intelligence (personality, work habits,
ethical/moral values).
This assumes that NLSY tested intelligence is the dominant factor
(excepting affirmative action) in current income distribution and that the
factors above produce only minor deviations in income level. Anecdotally I
would have to say that this is untrue, I think most of us know plenty of
stupid white men who make more than they should in a meritocracy.
Griffe: "Stupid white men"? Could P's
political orientation be peeking out here. The fact that IQ is the single
strongest correlate of income, unpleasant as that may be, cannot be
dismissed. Introduction of other variables does not change this fact.
Geography, is not a human quality. Sure, income stratifies by geography,
but within any region IQ remains the dominant income determining factor.
P: I also know people who score 110% (heh!) on any written test but
wouldn't know what to do with a tool if it bit them and vice versa (good
craftsmen don't necessarily test well or know English, although very good
ones will usually do both in the US).
Griffe: Please don't go there. In a country of
nearly 300 million you can find as many anecdotes as you like to
illustrate any point on any side of any issue. This stuff is OK for The NY
Times, but please keep it out of this discussion.
P: I assert that the geographic deviation is not small (the Median
household income in AK is $55K and in WV is $30K and obviously minority
representation is very different in different locales);
Griffe: If inclined, P might group states according
to per capita income and perform a Griffian calculation for each cluster
of states. The algorithm is laid out in the appendix. You will, however,
find that within each region IQ is the most important variable.
P: ... and that relevant factors other than intelligence are of huge
significance in compensation (criminal background, for instance, which
works to widen LaGriffe's gap, training and education, but also certain
physical traits and cultural traits which are positive and track strongly
to minority groups).
Griffe: Sorry, but criminal background and education
also strongly correlate with IQ. What positive physical and cultural
traits that track to minority groups have you in mind?
P: 2) Also limited in not even recognizing many other factors in income
distribution. Some of these which spring to mind are:
A)That people making less than $10K a year are not working full time
and that this is often not due to unemployability, (federal minimum wage
in '97 was $4.75*(1500hrs)+ $5.15*(500hrs) = $9700). My 99th percentile IQ
(Nordic) sister works part-time (4-7's) to allow time for volunteer work
(2nd lowest quintile income), my 99th percentile IQ mother works part-time
(~15hrs/wk) for my father's company (a lifestyle choice which probably
relates to her 7-99th percentile IQ children) and pays herself a salary in
the bottom quintile (minimum to qualify for Social Security--since all the
household income including the company profit goes into the same account
anyway). La Griffe does chop off those under 25 and over 64 to eliminate
some of this. Why doesn't he use the readily available statistics for full
time year-round workers of all ages (This puts the top of the bottom
quintile between $17.5K and $20K per person and the bottom of the top
quintile somewhere between $47.5K and $50K and dramatically affects his
calcs)?
Griffe: Sister? Mother? Gimme a break. I chopped off
the over 65 group because many of them don't work and consequently are
neither direct victims nor beneficiaries of AA. I omitted those under 25
to eliminate students who will later enter the work force, but do not yet
fall victim to or benefit from AA. The calculation applies to the slice of
the workforce most affected by AA.
P: Pushing 52% of blacks and 41% of Hispanics into a quintile which is
almost wholly composed of those employed part-time or for less than the
full year is not a logical conclusion from the IQ distribution (this is
also true to a lesser extent of the second quintile from bottom in La
Griffe's divisions).
Griffe: I'm afraid, P, you do not understand the
significance and consequences of a 1 SD gap. It is huge. In a meritocracy,
most blacks and many Hispanics would indeed be pushed into the part-time
and less than full-year employment sectors.
P: Losing/changing jobs happens thruout the IQ distribution (company
fails, down-sizes, moves, family moves because spouse changes jobs, need
to care for parents, health problems, etc.). Certainly someone of marginal
utility is more likely to lose a job or have trouble finding one quickly,
but not to the extent the "meritocracy" distribution would
imply.
Griffe: How can P know the extent of anything
without a measurement? Divine inspiration? More assertions -- easy to
make, hard to back up.
P: B) Assumes income distribution would be unchanged from present both
in Utopia and in meritocracy Not all income is wages and salaries, nor is
all income reported -- I would expect the top and bottom quintiles to
contain large numbers of investors and tax cheats(cash economy). The top
5% also contains (I assert) large sums of over-compensation (capital gains
for largely passive investors) which bear no relation to personal merit or
productivity (emphasis on passive). Which is a completely different issue
of reverse Robin Hood (I make multi-million dollar decisions regularly on
a 5 figure salary which accrue primarily to those whose chief contribution
to society was being born to rich parents or getting lucky once). This is,
of course, like most stereotypes, not true of all top 5% earners.
Griffe: There are bumps like tax cheating and
unreported income that my model does not account for.
P: C)Labor unions are a significant factor in boosting incomes (26%
difference in median for full-time wage and salary workers in 2002) and
blacks are over-represented in unions (32% more likely than whites in
2002). This over-representation is due only in part to affirmative action
in the public sector where jobs were 4 times more likely to be union than
in private industry. Minority populations of cities which are heavily
unionized like Detroit, LA, NY probably plays a bigger role.
Griffe: Unions are indeed a factor, but P's
rhetorical manipulation is worthy of a congressman. Instead of "32%
more likely than whites in 2002," P might have presented the data
from which 32% is calculated, ie, unionized: 16.9% blacks, 12.8% whites,
10.5% Hispanics. Thus viewed, we gauge the effect on income transfer as
small.
D)Income rises with age and minorities are disproportionately young
(this works to widen LaGriffe's gap)
Griffe: Point well taken.
E)Gender differences (31% difference in median for full-time wage and
salary workers) -- % women in workforce and for part-time or full time
differs by race
Griffe: There are part-time/full-time differences by
race, eg, in 2003 81% of white workers were full-time, compared to 85% of
blacks. I cannot separate out the causes, which may or may not be AA
related. As for the gender gap, check out P. Hausman: "I Am Woman,
Hear Me Whine", National review online April 3, 2001, still available
at: http://www.nationalreview.com/comment/
comment-hausmanprint040301.html
P: F) Training, Education, and Experience
Griffe: Trainability is very strongly correlated to
IQ. Ditto education. Experience? No ability --> no experience.
P: G) Unemployment and Prison population What effect do higher
unemployment and higher prison population have on minority earnings--one
could argue that this winnows the field of workers and increases
competitiveness of remaining minorities since stupid people are more
likely to get caught or be unemployable.
Griffe: Sorry P, but criminality is also strongly
linked to IQ. (See TBC.) Based on incarceration rates I estimate the male
gap in "criminality" as 0.81 SD for African Americans, and 0.43
SD for Hispanics, both with respect to non-Hispanic male whites. http://www.lagriffedulion.f2s.com/fuzzy.htm
Unemployed are covererd by my model. They have income; it's just zero.
P: 3)Zero Sum game -- The assumption that there is a certain amount of
income to go around and that giving one person more takes away from
others. This is not a 1-1 trade-off. The size of the economy depends on
production. Providing resources to people which allow increased production
increase the size of the pie. Affirmative action in employment may be a
zero sum game or even a negative sum game ... but affirmative action in
education, training, capital access, etc. may well be a positive sum game
by raising the productivity of those with lower potential more than it
impairs the productivity of those with higher potential. There is
definitely not a fixed amount of education to go around. I do not assert a
positive effect, I merely suggest an effect should have been accounted
for.
Griffe: AA in education, capital access etc. may
influence income distribution, but the robin hood effect, being an
"integral" rather than a "differential" approach to
income transfer, considers the _sum_ of AA influences rather than
individual factors.
P: If you get thru all of this I'd be amazed but I wanted to get this
off my chest.
Griffe: Thanks for the careful reading.
P: Best, Ben A. Pedersen, P.E.
===== La Griffe du Lion http://www.lagriffedulion.f2s.com
=For the moment let's you and him fight
Subject: A lawyer advises
wait-and-see on SCO
This is a nice, clear article written by an
attorney. If you know someone who is actually worried about SCO, you
should show this article to them.
http://techupdate.zdnet.com/techupdate/stories
/main/0,14179,2914464,00.html
-- Steve R. Hastings "Vita est" steve@hastings.org
http://www.blarg.net/~steveha
Lawyers always advise you to hire an
attorney, pay him, and then wait and see; but this is in fact a good
article and I can second the advice to read it.
==============
Harry Erwin on
Creative Destruction
On Creative Destruction
I call myself a liberal democrat (with the emphasis
on the 'liberal') although it is no longer clear to me how I really differ
from my mother's 'Chase' brand of conservative republicanism. I think I
have become more pragmatic than anything else. Having lived in the USA for
more than fifty years and then in the north of England for almost three, I
have relevant experience from both countries.
It is clear that capitalism is thriving at the local
level in both. Central planning simply does not serve the needs of the
local market, and that simple fact is accepted by most Americans and
English. It's at the next level up that we begin to see problems.
It is characteristic of ruling elites in both
countries that their reach exceeds their grasp. By this I mean that their
solutions to problems are almost always simple, coherent, and wrong.
Policy making in both countries tends to be captured by entrenched
interests, even when a large majority of the population is ill-served. An
example that stays out of politics--the actively church-going population
in England (1.7%) seems to have shrunk to be smaller than the actively gay
population (~2%). For most people here, the established church is tainted
by its historic links to the UK government, and alternative churches are
not mainstream enough to attract much membership. Yet, church governance
remains centralized and unresponsive to the needs of most people. The
church I go to has about 80 mostly elderly parishioners at each service,
and five priests on staff.
So I agree with the need for creative destruction,
although many English do not. I think Peter Drucker once pointed out that
closing the company should always remain an option for a board of
directors, and I think that should be the case for any institution that is
no longer useful. On the other hand, 'useful' is not just a short-term
economic term. The UK government seems to be laboring under the illusion
that the only people who benefit from publicly financed university
education and research are the people who take the degrees, but one
consequence of this perspective is that almost all engineering programs
are closing in England--they're expensive to teach and unpopular with
students. What kind of modern country can import all its engineers?
Similarly, the UK PhD program has been gutted to the extent that US
universities no longer consider the degree equivalent to a US PhD. What
kind of modern country can import all its real PhDs?
That, I think, is the real issue. Creative
destruction is necessary, but the criteria should look beyond short-term
economics. -- "I would like to be able to love my country, and
justice too." (Albert Camus)
Harry Erwin
Apparently not everyone understands what I
have been trying to say:
Unrestricted Free Trade Capitalism (19th
Century Liberalism) is probably the best engine for economic growth and
the creation of wealth.
Unfortunately, people are not equal. Fully
half of them are below average. The distribution of
wealth does not coincide with the distribution of votes. The problem of
political democracy has been, since ancient times, the problem of wealth
and debt relief and property. You can't have capitalism without
capitalists and the capitalists have to be allowed to accumulate capital.
But democracies don't vote for such measures
except in certain circumstances and those change.
The trick is to provide enough political and
economic freedom to allow growth to continue, while overcoming the
tendency of politicians to accumulate votes through divisionary politics,
and through the politics of envy. In the present situation it's worse
because the public service unions have become what amount to a first
charge against the tax base, never decreasing in numbers or pay or
pensions, and always organized to pay politicians to keep it that way.
The Libertarian Party may be devoted to
smaller government, but it insists on purity of candidates. Smart
political operators know better. They try to outbid the opposition on
narrow grounds. They know the politician won't stay bought. The result is
that outfits that care a lot on narrow issues generally get their way.
My arguments on tariff boil down to this:
there are a lot of displaced people in this country and they were formerly
part of the middle class and voted for middle class measures, which by and
large are pro-capitalist and pro-free trade, and as we said in the Old
South when I was a youth, "Tariff for revenue only" (i.e. not
for protection).
But the balance of power is always delicate,
as Schumpeter saw. I never believed he was right when he thought socialism
was inevitable if you kept democracy; but the tendency is toward more and
more government benefits (which also implies more and more government
employees) and the creation of more and more entitlements which will be
defended by their recipients, who will organize politically to defend
those entitlements. See Teachers Unions, Prison Guard Unions, Public
Service Unions, Social Security as the Third Rail, Medicare, etc. etc. for
examples.
Preserving both liberty -- economic freedom
is one of the essences of liberty -- and democracy is very difficult, and
there are no hard and fast rules and theories that work through all times
and places.
The people must have confidence that the
government really is trying to do its best for them; and that restraint in
what they grab is a good idea. California has gone about as far as
you can go in the other direction. No one has any confidence in the state
government now, but no one has the foggiest notion of what to do about the
entrenched interests which have their iron triangles of organized fund
raising to sell legislation to the highest bidder.
It will not be all that long before it
happens on a national scale if we continue to export manufacturing jobs.
Half the people are below average in
intellect. You cannot force them to compete with the above average in
other countries if you don't give them satisfying non-intellectual work
they can do.
I am waiting for someone to show me why that
analysis is wrong.
On the new Frankenworms:
As expected, the "Blaster" (aka "MSBlast"
and "LoveSAN") worm has new variants that are a bit more
efficient, and have different names for the executable program that is
installed on the computer. These new variants are not damaging to files on
your computer, but the potential is there. It would be quite easy for a
Blaster clone to contain a simple command to format your hard drive or
delete files, etc.
So it is quite important for your readers to ensure
that they have protection against this worm. They need to apply the
Microsoft patches to be protected. Note that the worm doesn't arrive in an
email message, so anti-virus protection, or not opening attachments,
doesn't protect your computer.
A firewall of some sort will help (there are reports
that the XP Firewall can help), as long as the user doesn't inadvertently
allow access to the worm.
The 'mantra' will help: "update patches, update
virus definitions, don't open attachments"; repeat three times....
Rick Hellewell Information Security Dweeb
digitalchoke@digitalchoke.com
=Thanks. No one has sent me an account of
infection through any means other than mail, so I can't describe how that
happens.
TOP
CURRENT
VIEW
|
CLICK HERE for
a Special Request. 
