Letters: Garden feud
Greg Wilson’s otherwise discerning review of Pamela McCorduck’s Aaron’s
Code (Review, 27 April) is marred by a backhander on artificial intelligence’s
‘failures during the past 10 years’.
If this is an assessment against the criteria set by McCorduck’s overheated
expectations of artificial intelligence, then well and good. By this measure
almost any realistic advance would seem disappointing. But if Wilson is
offering a serious remark then he stands in the footsteps of a long line
of neighbourly leaners over the wall of artificial intelligence’s garden.
When in a bad mood about one’s own plants, audible disparagement of others
revives the spirits.
What counts in the end is yield rather than discontents of onlookers.
In a special issue of The Knowledge Engineering Review, artificial intelligence
is debated in the context of similarly unsubstantiated criticisms made in
1972-73. However, the subject continues to thrive and moreover does so on
a fraction of the support which is today devoted to some more speculative
areas, such as-to pluck an example from the air-parallel computing.
Donald Michie The Turing Institute Edinburgh
Letters: Scratched backs
Grant Lewison may have missed an important factor in determining how
often research papers are cited (Forum, 4 May). In my attendance at conferences
both in Europe and in the US, I have noticed that in each coherent field
of research there is a core group of active workers who publish regularly
and dominate conferences in that field, which occur every year or two. Because
of this and the fact that they send each other copies of their papers, they
tend to cite each other’s work more frequently than papers of those outside
this peer group.
In addition, in my fields of interest the European Commission and the
European Materials Research Society are very proficient at holding conferences
at which these peer groups can meet. Because the Commission financially
supports cross-frontier research, it is not surprising that multi-European
papers are cited more frequently.
G. B. Brook High Wycombe, Bucks
Letters: Happy with scum
Andy Coghlan asks: ‘It may be green but is it clean?’ (Technology, 4
May). I have used nothing but Ecover for five years and my clothes are holding
out well. The naked eye cannot detect bacteria and fungi and the deposits
of scum do not weigh me down. I am happy to use a product which has less
impact on the microorganisms in my clothes if it has less impact on the
microorganisms in our rivers, lakes and reservoirs.
Coghlan failed to mention that eutrophication is not the only environmental
problem associated with phosphate. I also prefer to use a phosphate-free
product to reduce the quantity of heavy-metal waste produced during the
manufacture of phosphate additives.
It is not surprising that Albright & Wilson do not draw attention
to this by-product of the detergent industry in their report on non-phosphate
powders. They have been widely condemned for the serious heavy-metal pollution
of the Irish Sea caused by their chemical plant in Cumbria, ordered to clean
up by the National Rivers Authority, and face court action from Greenpeace.
I feel strongly that they should clean up their own act before attacking
competitors whose priorities differ from their own.
Michael Hirsh London
Letters: Whose safety?
Mick Hamer’s article on the use of inflatable air bags as protection
for motorcyclists (Technology, 27 April) left me with a number of nagging
doubts. I am uncertain as to how closely the motor-cycles used in Bryan
Chinn’s experiments resemble real road-going machines. For instance, in
his experiments on leg protectors, Chinn removed the cylinder heads, barrels,
carburettors and exhaust pipes from the BMW R80 which he used. How were
the motorcycles used in this more recent set of tests modified?
The article also left me uncertain as to how well the dummies used in
these experiments model a real motorcyclist. Does he have these dummies
sitting in an upright posture with their hands on the petrol tank, as was
the case in the previous studies? I am also mystified by Chinn’s reluctance
to reveal the results of the experiments done using a moving car.
My final point is one of emphasis. It seems perverse to be discussing
methods of reducing the severity of injuries caused in road accidents in
the absence of any mention of the prevention of those accidents by improved
driving and riding standards. I would even question the wisdom of introducing
devices which make the driver feel safer. It can be argued that the motorist
will modify his style of driving to return to the perceived level of risk
that existed before the introduction of the safety measure. A natural consequence
is that as the motorist feels safer, the burden of risk falls more upon
the cyclist or pedestrian.
This is consistent with the observation that on the introduction of
seat belt legislation in this country, the number of deaths on the road
was not significantly changed, but a larger proportion of those deaths were
of pedestrians. This line of argument might be extended to the point of
repealing seat belt and crash helmet legislation for the benefit of more
vulnerable road users.
Duncan McKenzie Imperial College of Science, Technology and Medicine
London
Letters: Daresbury queries
The MPs Chris Butler and Doug Hoyle correctly drew attention to the
breakdown of traditional peer review procedures over the proposed closure
of the Nuclear Structure Facility at Daresbury (Letters, 6 April). In particular
they referred to the formation of a review (Science and Engineering Research
Council press release, 7 February) to ‘assess the importance of the science
in the context of Council work as a whole’. We now discover this review
will not consider the nuclear structure science that may be performed at
the NSF, the only UK-based facility for this research.
In his reply, Sir Mark Richmond, the chairman of the SERC, rebuts the
MPs’ letter, accusing the MPs of giving the impression that the review be
limited to the NSF (Letters, 27 April).
I can inform Sir Mark that not only did the MPs assume that the SERC
press release meant that future NSF based work would be included in the
review, so did all the physicists I talked to at the time.
P. J. Woods Department of Physics University of Edinburgh
Letters: Dino-mania
Nick Hodgetts calls on palae-ontologists, and more specifically dinosaurologists,
to account for themselves in the light of the serious plight of existing
species like the giant panda and the habitants of the rainforest (Forum,
16 March).
Though the study of dinosaurs has not directly affected the survivorship
of any particular species it has taught all of us not to be so narrow-minded
in our perception of the biosphere and how individual species interact with
each other.
The classic example is the terminal Cretaceous extinctions. Not only
did the dinosaurs perish, but the entire biosphere was altered by extinction
and origination. We can learn something from that.
The emergence of dino-mania has several factors in its favour. First
and foremost is the education value, something missed by Hodgetts’s essay.
No other science has an impact on young children like dinosaurs. It is usually
their first introduction to science.
If we are serious about saving the rainforest we do not have to knock
palaeontology.
Tim T. Tokaryk Saskatchewan Museum of Natural History Regina, Canada
Letters: Graphology-schmology
Describe an old wives’ tale as something-ology and almost anyone will
believe it. Doesn’t Alison Brooks (Forum, 27 April) realise that when applicants
for a job are asked for a sample of their handwriting it is not for analysis
by some eminent scientist. Nothing could be further from the truth. The
samples are sent to the typists, who pick out the ones they can read.
Richard James Cambridge
Letters: Licensed to encrypt
The letter from Jennifer Seberry of the University of New South Wales
(20 April) raises the question of whether encryption programs claiming to
use DES do in fact do so and how long they take when they do.
My own DES implementation is an assembler program based on a Pascal
algorithm which I understand was originally distributed by the National
Physical Laboratory. This program can encrypt or decrypt a half megabyte
file in less than 10 minutes on a 25-megahertz 386 PC. That’s 25 per cent
more text than contained in an average James Bond novel, and while it is
hard to guess the size of ‘the battle plans for the recapture of Kuwait’,
it certainly looked less complicated than Live and Let Die when I watched
it enacted on the telly.
The Pascal version takes nine times as long, so perhaps the reason Seberry
knows of no DES algorithm which would not need hours for this task is that
she has only seen high-level language implementations.
Iolo Davidson Tetbury, Gloucestershire
Letters: Auxiliary devices
I write regarding Chris Bower’s letter concerning the lack of information
about tests on environmentally desirable auxiliary devices for car engines
(13 April).
I worked at the New South Wales government vehicle emission laboratory
for nine years before moving to my present position two years ago and in
that time was plagued by such auxiliary devices, all of which were going
to save the world.
Several dozen devices were evaluated, most being rejected on scientific
or engineering grounds without testing. Several showing theoretical promise
were subjected to emission and fuel consumption testing, and several which
were rejected on theoretical grounds were tested under political pressure.
The testing of one device was paid for by a television station as part
of its programme of debunking misleading advertisers. The device was a plastic
sphere in two halves, each containing a magnet, which was installed around
the fuel line. It was advertised as a ‘polarizer’.
The ‘theory’ behind the device was that the magnets caused the fuel
molecules to line up, thereby increasing combustion efficiency. The fitting
instructions included advancing the static ignition timing, which, of course,
gave a small fuel consumption reduction (as one would expect), but also
increased emissions to illegal levels, and caused marginal pinking (again
as one would expect).
With the widespread adoption of engine management systems and fuel injection
by vehicle manufacturers in Australia since 1986, following the adoption
of unleaded petrol and US 1975 emission standards, there has been a drastic
reduction in the number of such devices being offered. We are now more inclined
to see fuel additives being advertised for ‘upper cylinder lubrication deficiency
caused by unleaded petrol’, which has about the same value as the polarizer’s
claims.
My advice to Chris Bowers is to view claims without some theoretical
basis with extreme caution and to insist on independent test results by
an authoritative body as part of any evaluation.
Jack Haley National Roads & Motorists’ Association Sydney, Australia
Letters: Gold fusion
Michael Kenward’s interesting review of Frank Close’s book Too Hot to
Handle tells of the furore which followed the announcement by the chemists
Martin Fleischmann (Fellow of the Royal Society) and Stanley Pons that they
had produced fusion by chemical means (Review, 16 March). Kenward explains
that ‘it had all happened before’, referring to the recent breakthrough
on high-temperature superconductivity – with the significant difference
that many experimenters quickly confirmed hot superconductivity.
Indeed, ‘it had all happened before’ that a Fellow of the Royal Society
had announced the achievement of transmutation by chemical means. Dr James
Price MD was elected FRS in 1782, with his certificate of recommendation
describing him as ‘well versed in various branches of natural science, and
particularly in Chymistry’. In 1782 some distinguished men visited Price’s
chemical laboratory at his country house at Stoke in Surrey to witness his
transmutation of mercury into gold. He quickly published a pamphlet telling
of his success, which created a major sensation.
In the early years of the Royal Society alchemy had been taken very
seriously by Boyle, and especially by Newton; but in 1782 Joseph Black and
other chemists rejected scornfully Price’s claims for transmutation. Sir
Joseph Banks, the President of the Royal Society, reminded Price ‘that the
honour of the Society as well as his own was at stake, and that the experiments
should be repeated before officially appointed fellows as witnesses . .
.
‘Early in August the three delegates of the society arrived at Stoke
and Price conducted them to his laboratory. Leaving them for a moment, he
drank a concoction containing hydrocyanic acid (prussic acid), returned
to the laboratory, and collapsed and died before their eyes.’ (E. J. Holmyard,
Alchemy, Penguin, Harmondsworth, 1957, p 261).
Could it have been the recollection of that regrettable episode which
has restrained the Royal Society from inviting Fleischmann to demonstrate
his experiment before witnesses?
Garry J. Tee University of Auckland New Zealand