Two relatively bloggable things happened yesterday so I’ll make some attempt to reconstruct them here in words.

Imperial Physics Interview

I think I’ll do what Farhan did last year in the spirit of open source (kinda) and say something about my interview.

I arrived at 12:30 in time for the tour after just about managing to find the mysterious room 306 (hidden in a sort of conference room). There was someone who had made it all the way from Poland for this and various people who had made arduous journeys from all over the country, so I almost felt guilty about having had such an easy trip – 20 minutes on the No 10 down HSK. We got given a general walk round and free lunch (always a good thing) and were even (jokingly) offered a pint by the tour guide before our interviews!

The 12 of us with interviews that day were split up into three groups of four – I was interviewed with the three others applying for the four-year ‘Physics with Theoretical Physics’ course. We were first all sat together and had the course run past us – it all sounds pretty awesome with ‘complex analysis’ and ‘mathematical analysis’ both being taught in the first year (GL said once the sign of a good maths course is mathematical analysis being taught in the first year). We all went off for a quick (free) tea session in the lunching area (I was hoping to catch some of the ion trapping people from my work experience but they had probably by then left) during which we discussed relativity and space-time diagrams and the concept of ‘now’ which was pretty interesting.

Then we were all sat outside the room and were called in individually for interview. I was the last (a consequence of alphabetical ordering) and the people who went before me seemed to find it OK – one said she had to sketch ’some graph’ and explain ’something physics-ey’ and everyone seemed to have got two questions – so I didn’t think it would be too bad.

So I went in and immediately saw a Newton’s cradle sitting on the desk in front of my interviewer. Her research interest was quantum gravity and was being shadowed by someone who was interested in explosions and generally breaking things Mythbusters style, which is cool. She didn’t mention my personal statement at all and just asked me why I wanted to do physics (as opposed to maths) and why I wanted to go to Imperial. I said I liked being able to see concepts happen in real life, to which she pointed out relativity isn’t exactly the average real life situation. I said something about being able to touch and feel and see stuff in action, and applying maths to stuff and seeing it work, which she seemed satisfied with – ‘I know exactly what you mean’.

She then gestured towards the pad of paper and asked me to differentiate 2^x. Following standard procedure I just rearranged it into e^xln2 and differentiated that, though I didn’t / forgot to turn (ln2)e^xln2 back into (ln2)2^x at the end. She seemed happy and said ‘yup that’s right’ then asked me whether I knew what the thing on the desk was. I successfully identified it as a Newton’s cradle and explained that each collision is elastic and that this results in the inboud ball stopping and the next one going forwards with the same velocity as the inbound one, etc, with some support from a fumbling demonstration.

She then asked me a question about a ping pong ball and a golf ball being dropped such that the former is directly over the latter from 1m, and she asked me how high the ping pong ball would bounce. I invoked the coefficient of restitution and said let the velocity at the bottom be v. Golf ball bounces, goes up with v. Ping pong ball bounces against this, goes up with 3v. Invoking conservation of energy twice the answer came out to be 9m – which was right, apparently. That’s quite high…

She then asked me how long it takes a photon to get from when the universe became transparent to now. I looked confused and for some reason tried to resist the temptation to ask ‘from whose frame of reference?’, though it turned out that’s what the question was asking. I drew a space-time diagram and made a pretty dreadful estimate of the age of the universe [my estimate turned out to be about the age of the earth; note to self: learn some of these numbers sometime...] and asked for clarification on the question. She said it was a trick question and said it’s about frames of reference, at which point I realised it was indeed a relativity question and said ‘zero’ with a slightly botched-up explanation using t = yt’ [note to self: try to remember which side y is on!]. I guess I should have drawn the thing with the axes changing angle on the space-time diagram but nvm…

At the end she said she can’t tell me whether I have an offer and if so it will be with an A* (I *think* I heard that correctly, and it’s possible ‘further maths’ was mentioned in the same sentence – so that was a bit of a surprise). Apparently the school wrote me a good reference, which is good.

EDIT: should probably also say offers/rejections in several weeks

It was all over by 5pm as promised so I had some time to kill before the religion debate at 6:45 (next part of this topically bimodal post).

Intelligence Squared Debate: Religion

More specifically the motion was “The Catholic Church is a force for good in the world”. Matthew, Theo and I, the proud founders of the SPS Sceptic Society, were once again reunited to watch Christopher Hitchens (Writer, broadcaster and polemicist, author of the bestselling book “God is not Great”) and Stephen Fry (Actor, author, comedian and television presenter) debate against Archbishop John Onaiyekan (Roman Catholic Archbishop of Abuja, Nigeria) and Rt Hon Ann Widdecombe (Conservative MP and Catholic convert) (descriptions taken from the I² page). As always I tried to write some notes – here they are in pretty condensed form. {curly brackets} indicate words external to what the speaker said, e.g. comments. I’ve also abbreviated names slightly, and nothing is word-for-word

1st speaker: Archbishop – For

• General stuff about his father and him and all his family being Catholic {Matthew suggested this sounded like the start of a sermon}
• Questioned what sort of ‘force’ the debate was about. He thinks the ‘force’ is a spiritual message, spread around the world, and the force is what this message teaches etc.
• Comments about the sheer size of the ‘force’
• Said that if you ask anyone in Nigeria they’ll tell you the Catholic church is a force for good {according to WJB if you ask 80% of people in Europe they’ll tell you that only GM food has DNA…}
• Quotes statistics about what {I would describe as satisfied customers}
• {Actually I have to say, it did sound like a sermon}

2nd speaker: Hitchens {to much applause!} – Against

• Started with some witty banter
• Said the opposition should have started with a list of apologies {to much applause}
• Started listing crimes against humanity the Catholic Church has committed {a couple were incorrect I think, and also the debate is the present – lots of his examples were from centuries ago. Still valid though, as we will see later}
• Child abuse – the church tried to excuse itself for it instead of apologising
• Said something about antisemitism {lots of audience tuts}
• Religion goes against the method of free thinking and scepticism
• Quote’s Stephen Fry’s situation {Fry turned out to be a really strong speaker because of this later}
• Talked about the ’sale’ of remuneration – paying for people to pray for you
• {I thought Hitchens would be stronger – he was, of course, as always pretty harsh and blunt, but he wasn’t as fired up as he was in some of his previous debates}

3rd speaker: MP – For

• Claims Hitchens misrepresented the Catholic Church {sarcastic applause from Hitchens!}
• Picked up on Hitchens talking about the past, not present
• Picked up on antisemitism thing
• Quotes WWII – helping Jews
• Quotes christians having to renounce faith to join SS {considering Nazism was pretty anti-christian anyways I don’t think this is a particularly valid point}
• Torture – last time’s standards were different so everyone was guilty, not just the church {Fry and the audience tear this apart later}
• Talks about child abuse – church ‘powerless’ to do anything, magistrates etc. also at fault
• Charity – $Bns given to charity {I wonder how much this is in comparison with the church’s wealth…}
• Hope argument – church gives hope to people
• she said ‘I knew condoms would come up’ – tried to make a joke of it {general audience tuts, someone shouted “how dare you laugh at that!”}

4th speaker: Fry – Against

• Started completely differently from Hitchens – said he’s fine with people believing and seeking enlightenment etc. – shows no hostility towards them
• Attacked MP’s point about past vs present – MP basically said ‘history is not important, so let’s forget about it’
• Talked about purgatory, people paying to bypass purgatory / go to heaven; referenced South Park’s version of purgatory (!)
• Quotes ‘outside the church, there is no salvation’ being used to excuse horrific deeds
• Church commands people to be ignorant, prevents them thinking for themselves
• Catholic Church deems itself the only owner of the truth and bullies people into believing
• Current Pope on child abuse: “We do not have the power of a nation” <- yes you do
• Commented on women’s equality
• Apparently the pope wrote a letter / made an announcement to child-abusing priests: [paraphrasing] “do not talk to the police, keep it secret, talk to me instead”. Pope claimed solution is to stop “homosexuals from entering the church”
• {either Hitchens or Fry made this following point} The church ’sentenced’ one child-abusing priest to ‘a lifetime of prayer’ instead of several months / years in prison
• Church doesn’t need to exacerbate existing gay stigma
• Stephen Fry said: “I find it ridiculous that I am being called a perv by such extraordinarily sexually dysfunctional people” {*huge* applause + laugh, proposition looking really pissed off}
• Pope spread false lie that condoms makes AIDS spreading worse – instead of making useful suggestions
• church obsessed with sex. Comparison with food – church equivalent of anorexic and obese {more huge audience support}
• Proposed solution: pope gives back all of Vatican’s wealth to those from whom the church has stolen {even more audience cheer}

Before the debate the audience had pre-voted thus:
FOR: 678
AGAINST: 1102
ABS: 346

Questions

• Catholic Church broke 5 UN conventions on child abuse – should not be allowed to get away with this
• To Archbishop: Q “which Catholic policy are you most ashamed of?” A “I am ashamed of none of them”
• To proposition: “do you need the Vatican’s wealth?”
• To proposition on torture: “even though the standards of the time are xyz, isn’t the truth of the Church doctrine ‘eternal’?”. Church had changed mind on slavery for example. Seems like church in constant state of limbo. MP says “limbo = ’second light’” – {only huge audience groan in entire debate}

Conclusions

Stephen Fry

• MP groaning: “I knew condoms etc would be brought up” – a bit like a burglar groaning in court “I knew my burglary would be brought up” {audience cheer}
• Constantly wasted opportunity for Catholic church to do something by giving away lots of its wealth – until then, not force for good

MP

• Reason for people having children in Nigeria is they need someone to look after them when they are old {relevance?}
• Says no statistical evidence for condoms preventing AIDS {so pope justified in spreading lies?? Theo and I agree she’s crazy}

Hitchens

• Thoughtcrime argument – catholic church essentially enforcing regime of thoughtcrime

Archbishop

• Basically said history doesn’t matter again {even though point previously successfully rebutted by Fry}
• Said he cares about his own relatives and he is happy for them to be Catholic etc. {urgh. There were two parents who fed their baby a litre of salt to punish it. I’m sure they cared about their kid, they just didn’t know giving it salt would be a bad idea. This point isn’t really valid.}

After the debate the results were thus:

FOR: 268
AGAINST: 1862
ABS: 334 {I think – it might have been 34. Can’t check by adding up since audience size was changing throughout debate}

Stephen Fry was a ridiculously strong speaker in this – even stronger than Hitchens, and despite my weak preconception that the Catholic church wasn’t doing much good, after this debate I am now quite convinced that it’s, if anything, a force for evil. Fry shouted twice (or even thrice) in that – he really is passionate about this topic.

Other highlights include us spotting Derren Brown in the audience and a priest in the audience standing up and totally siding with Stephen Fry.

EDIT: whenever I mention the ‘church’ I mean the Catholic church, just to avoid any confusion. As Fry pointed out, he has nothing against Quakers, for example.

Yesterday (16th October) I attended another of the weekly UCL science lectures. This week it was given by Dr Ryan Nichol from the Department of Physics and Astronomy at UCL. It was basically about practical particle physics, discussing experiments such as the LHC and the ANITA experiment (the one with hot air balloons flying over Antarctica). So here are my relatively brief impressions…

The first thing I noticed before the lecture started was the enormous turnout. In fact the lecture was moved from the usual Massey Lecture Theatre to the vastly larger Chemistry auditorium, and even so I ended up sitting on the steps. The audience was largely populated with schools, perhaps explaining the turnout. I also noticed the MIT-style blackboards adorning the front of the lecture theatre.

The talk started with a primer on high-energy physics. This was a pretty standard intro, putting familiar things on a log scale of size, quickly delving into some elementary particle and the ball-exchange explanation of repulsion and attraction in Quantum Field Theory (two people on boats throwing balls to each other), an explanation Feynman would not approve of considering his famous answer to the question ‘what makes two magnets attract’! QFT was elaborated upon slightly in terms of a brief description of the four gauge bosons.

He then moved onto cosmic rays. Apparently it was discovered there were more high-energy particles being detected than there ’should’ have been. In ~1910 Theodor Wulf attempted to detect a difference in muon detection between the top and bottom of the Eiffel Tower, but his instruments weren’t sensitive enough to record any difference. In 1912 Victor Hess did the same experiment, but using hot air balloons instead. As it turned out, as the balloon’s height increased, the radiation detected decreased owing to a decrease in detection of earth-produced radiation due to an increased amount of atmosphere between the balloon and the surface. But after a certain point, the detection rate actually increased, leading the scientists to conclude this excess radiation actually came from space. He also talked about the Pierre Auger Observatory in Argentina. This is a set-up spanning an enormous area consisting of a grid of particle detectors. If a shower occurs, several of these detectors will be hit at the same time. There are also four telescope stations which can, on a clear moonless night, actually ’see’ the fluorescence caused by the shower particles hitting air molecules (Cherenkov radiation?) and produce a pretty picture and model 3-D of the event. Dr Nichol was also at one point involved in ‘CREAM TEA’ which stands for ‘Cosmic Ray Extensive Area Mapping for Terrorism Evasion Application’. Supposedly this uses naturally occurring muons to map an area – detectors can build up a good picture of dense objects (like bombs) within an area by measuring scattered muons.

He then moved on to talking about neutrinos, invented in 1930 by Pauli to conserve energy. I knew they didn’t interact with much, but apparently the average neutrino travels through 53 light years of water before interacting! Luckily the sun produces 100 Bn neutrinos per cm² per second, and the few milligrams of ⁴⁰K in humans emits about 300 M neutrinos per second. I was pretty surprised by the sheer number of neutrinos being constantly produced – beta decay must be a big thing… He then talked about some neutrino experiments. He mentioned the MINOS experiment in which Fermilab fires a massive stream of neutrinos at Soudan (MN) in an attempt to detect a few. He also talked about the Super-Nemo experiment (interesting name…) which was created to discover whether neutrinos are their own antiparticles through the inspection of radioactive decays. Apparently theory has it that neutrinos can turn matter into antimatter (I’m guessing something to do with beta+ decay), so the discovery that neutrinos are their own antiparticles might help explain why the universe has more matter than antimatter.

The talk moved onto the less well-known field of micro-black holes. He didn’t actually say much about this except that most black holes are several hundred/thousand/million solar masses, but that it’s possible to create tiny ones of only several milligrams. Concerning the media scare about the LHC creating a black hole and eating the earth, he explained that only some exotic models allow collisions in the LHC to produce micro-black holes (which can be subsequently studied by measuring the Hawking radiation produced as they [very rapidly] evaporate), and only the most exotic models allow one such black hole to grow large enough to engulf the Earth. Thinking about it, 14 TeV (now reduced to something like 7 or 8 TeV) collisions are unlikely to do much harm considering the Earth is constantly being slammed into by 300 TeV particles from über-novae and other such awesome events. This actually reminded me of an interesting article I read in Scientific American (my preferred science magazine) which discussed formulating QFT on the hyperbolic geometry of general relativity (semiclassical gravity). Apparently as a star begins to collapse into a black hole, a force gets created which opposes this collapse which grows to infinity when the star tends from above to a certain radius. So according to this article, while singularities are theoretically possible, there is no way of going from a non-singularity to a singularity (a bit like travelling faster than light). So Black Holes are actually Black Stars. All very interesting stuff, though I just don’t know enough about the maths to really appreciate it myself. My only experiences of hyperbolic geometry are from a C4 programme about 10 years ago, and from reading GEB by Hofstadter. Not to mention QFT…

Finally, he talked about ANITA, a project he had recently worked on. I think someone came to Halley Soc recently to talk about ANITA, so this sounded pretty familiar. The idea was to fly around Antarctica in hot air balloons in the hope of detecting a Cherenkov cone of RF radiation produced when a neutrino decays in the ice. Apparently the battery box of ANITA had to be painted half black and half white to prevent overheating / freezing (since fans wouldn’t be very effective at low pressures at high altitudes), so they did an art contest for the best 1/2 black and 1/2 white design! Later they analysed their results and discovered no neutrinos, but did find they had detected 6 ultra-high energy cosmic ray air showers – the radiation was so intense that it actually bounced off the ice and was picked up by their detectors.

The Q&A session afterwards didn’t seem as lively as normal, possibly due to the sheer size of the audience, but overall it was an interesting talk and I learnt about several cool experiments I’d never heard of before. I’m always quite interested in how there’s always lots of good theory in practical experiments (cf. my Imperial work experience – week 1, week 2), so it’s all good.

Afterwards we (the three of us from SPS who went) managed to spend 3 solid hours discussing physics and maths in the regular post-UCL-lecture pub ranging from things from the lecture, CERN and QM to Feynman, Galois, Turing and Boltzmann to Primer (my new favourite film), a suitable conclusion to the evening!

EDIT: Urgh that wasn’t brief at all. I should really stop doing this whole mega-post thing…

This is actually my Science Essay Prize submission – here it is for you to read / scoff at…

The title is a reference to a famous article by Hawking and Hartle: ‘Wave Function of the Universe’ (2)

'Coin Toss'. Taken at 21:34 on 3 Aug 2009 at home using an ad-hoc strobe and a 1p coin. Camera: Fujifilm Finepix S5800 (DSLR); 1” exposure; F/3.5 aperture

The Quantum world, the world of the very small, is very different from the world we live in and are familiar with. In this world, the law of conservation of energy can be cheated for a short period of time owing to the time-energy relation to produce short-lived but vital ‘virtual photons’, and even the concept of probability has a quantum twist. This rather bizarre framework has however been enormously successful and has allowed physicists to make very accurate predictions of experiments’ results; it has never been shown to be incorrect or inaccurate. This seems to imply that, however different it may be from classical physics, quantum mechanics is correct. Yet it seems utterly inconceivable that such apparently opposing ideas are in fact both largely correct and thus identical on a macroscopic level, as evidence suggests, and that the correspondence principle, which states that Quantum Mechanics (QM) reduces to classical mechanics at high masses, is true. Classical mechanics provides a full description of a coin toss. Can QM make the same claim while keeping its predictions consistent with those of classical physics?

The Correspondence Principle

The most important relationship between the quantum mechanical description of this coin and its classical counterpart is embodied in the correspondence principle.

The Schrödinger equation is used principally to work with the energy of a system and is displayed here:

Where m corresponds to the mass of the system, E is a number whose possible values are the possible energies the system may take on, and Ψ(x) is a function of the position of the system (the wavefunction) whose square modulus is the probability density of the particle being at position x. In the example of the one-dimensional infinite square well in which a particle is trapped in a one-dimensional universe (such that V(x)≡0) with impassable ‘walls’ placed at 0 and l, it turns out that:

Ψ(x)=sin⁡Ax (In fact Ψ(x)=cos⁡Ax also works)

At both walls (at l and 0):
Ψ(x)=sin⁡Ax=0

So:
A=nπ/l (n=1,2,3…)

When Ψ(x)=sin⁡(nπ/l x) is plugged into the Schrödinger equation, one obtains:

This tells us that the greater the value of n, the more nodes the wavefunction has on the interval [0,l], so the more energy the system has.

Contrasting the shape of the graph of Ψ(x)²=sin²⁡Ax=sin²⁡(nπx/l) (a probability density plot) when n=1 (below, left) with that when n=50 (below, centre), one surmises the one corresponding more closely with what classical physics would predict (i.e. a constant probability density) is the case when n is high; when n is sufficiently high, the graph of Ψ(x)² can be faithfully approximated to a classical constant probability density (below, right) in which the particle can be anywhere along the line with equal probability.

Click to embiggenatrify

Thus as the energy of the system increases, the value of n rises and thus the correspondence between classical and quantum mechanics becomes closer.

Although this is a very specific example, wavefunctions tend to turn out to be trigonometric functions (such as sin⁡x, cos⁡x and e^ix) which are often the only eigenfunctions of the second differential operator in the Schrödinger equation, leading to the occurrence of a positive power of this naturally arising quantum number (n=1,2,3…) which increases with the number of nodes. One can use the periodicity of Ψ(x) to support the argument that this positive integer quantum number n exists for almost every wavefunction Ψ(x) (i.e. for almost every quantum system). So long as the system’s energy is over a certain limit known as the classical limit, it will have a sufficient number of nodes that it will behave effectively identically to a classical system; as its energy tends to infinity, its behaviour will tend to classical behaviour. Extending this concept, since the energy of the coin toss situation is so large relative to atomic energies (the kinetic energy of the entire coin while spinning is many orders of magnitude greater than that of a stream of particles about to collide at 99.9999991% the speed of light in the LHC), n is extremely large so the quantum description of the coin toss approximates extremely well, and in most respects identically, to its classical counterpart.

Ensemble Interpretation

At its heart, the interpretation states that the Quantum Mechanics is a statistical abstraction from reality: No single particle has a wavefunction; rather the wavefunction applies to a group, or ensemble, of classical particles and it merely describes their distribution. Although this interpretation is largely rejected on the grounds of the Young’s double slit experiment run using a light source that emits only one photon at a time in which an interference pattern still appears, it is an apt intuitive argument for the correspondence principle: the time evolution of the graph of position against probability for a single particle with a particular wavefunction is very similar to the that of position against an approximate measure of density of an ensemble of classical particles. Since in the example of a coin toss there are a large number of particles participating with similar wavefunctions, the distinction between a Quantum and a Classical approach to determining the behaviour of the experiment becomes merely academic and a matter of paradigm since both methods will yield very nearly identical results – whether every particle has a wavefunction or the entire system is considered collectively is no longer relevant to predicting the result of a coin toss.

Classical Probability: Coarse Graining

When a decision event occurs, there are several histories associated with it. For example in a simplified universe containing only one photon, a target, and a piece of card with two slits in it, when the photon ‘chooses’ whether to travel through the right or left slit in the Young’s double slit experiment before hitting the target, a history is created for each of the two possibilities: one for when the photon chooses the left slit and one for the right slit. In a more complicated universe containing several particles, every combination of ‘decisions’ the particles ‘make’ constitutes a different history: each possible history (containing complete information about every particle in the universe at every point in time) is a separate history according to quantum mechanics. These histories are known as fine-grained histories.

As I mentioned, the quantum version of probability is very different from its classical counterpart. Instead of assigning each fine-grained history a probability, Quantum Mechanics assigns pairs of fine-grained histories values. This value for a pair of histories A and B can be denoted by a function: D(A,B). The pair of histories can be a pair constituting of just one history, for example D(A,A), which would in fact be a number between zero and one and can be interpreted as the probability of history A occurring. However the function D follows the following rule:

D(A or B,A or B)=D(A,A)+D(B,B)+[D(A,B)+D(B,A)]

The last term, [D(A,B)+D(B,A)], is called the interference term between histories A and B and can have both positive, negative and zero values; if it is not zero, histories A and B interfere with each other, making it difficult, and sometimes impossible, to assign a probability to each separately.

The problem has now been implicitly stated: if interference makes it so difficult to assign probabilities to histories, what makes it possible to say that the probability of tossing ‘heads’ on an unbiased coin is ½; how can such a prohibitive concept of probability be reduced at large scales to classical probabilities?

The solution lies in an idea called coarse graining: histories are organised into sets. For example there might be the set of all histories in which Photon A travels through the left slit – all histories in which this event takes place are organised into a set. The purpose of this classification is to ignore all factors that do not matter to the critical situation by taking the set of all fine-grained histories which agree that the critical event occurs but may disagree on the goings-on in the rest of the universe. This allows the construction of a single coarse-grained history whose probability is:

D(A₁ or A₂ or A₃…,A₁ or A₂ or A₃…)

where A₁, A₂ etc. are all the different histories that make up this set. If all the fine-grained histories can be divided into sets such that each fine-grained history belongs to one and only one set, one obtains a set of mutually-exclusive coarse-grained histories. The result of all this is that, for two coarse-grained histories α and β consisting of fine-grained histories {A_i } and {B_i } respectively (where {A_i } stands for A₁,A₂,A₃,…), D(α,β)+D(β,α)—the interference term between α and β—is the sum of all the interference terms between pairs of fine-grained histories that belong to those two coarse-grained histories: the net interference between α and β is the sum of all the smaller interferences between {A_i } and {B_i }. This summation often leads to cancellation of positive and negative interference terms, leading to a near-zero interference term between α and β: D(α,α) and D(β,β)—the probabilities of α and β respectively—are very well-defined numbers, no longer dependent on interference terms, leading to independent probabilities of mutually-exclusive events. This collapse of D values into classical probabilities via coarse-graining is called decoherence, and statistically, the more fine-grained histories that are summed over, the closer to zero the net interference term becomes, and the more definite and independent the probabilities of coarse-grained histories become. Returning to the coin-toss example, since a huge number of particles interact with the coin’s faces, an enormous number of histories have to be summed over to obtain the two coarse-grained histories of the coin: heads or tails, leading to a near-zero interference term between the heads-up and tails-up histories: it can be said with much confidence that the probability of tossing heads is ½.

Classic Mistakes: Misinterpretations

Common misinterpretations of the implications of the postulates tend to make QM sound more unbelievable and divergent from classical physics than it really is. For example a famous example often quoted when explaining QM is Schrödinger’s Cat. This is a thought experiment invented by Erwin Schrödinger in which a truly random (unpredictable) process produces a Boolean (yes/no) output which determines the fate of a cat contained in a sealed box. If the output is ‘no’, poison is released into the box and the poor animal perishes; if the output is ‘yes’, no poison is released and the cat remains alive. The concept Schrödinger was attempting to communicate by producing this analogy is that of superposition: in the story of the cat, after the cat’s fate has been determined and appropriate actions effected but before the box has been opened and the ‘aliveness’ of the cat observed, the cat is in a superposition of being dead and alive: while classical physics would state that the cat is either dead or alive, quantum physics would assert that the cat is in a curious state of being neither dead nor alive but in some half-way state, and only when the box is opened does the cat decide its ‘aliveness’. My personal opinion is that this is a brilliant analogy and can be a great help in explaining quantum superposition. Unfortunately the analogy is far too often taken literally, leading many to believe that in such a situation the cat really would be neither dead nor alive. As Gell-Mann pointed out in his book The Quark and the Jaguar (1), the cat frequently interacts with air particles inside the box which in turn interact with the box which in turn interacts with the rest of the universe; since a large number of particles are involved the two outcomes (living cat vs dead cat) decohere leading to a more or less classical situation in which the cat is either dead or alive: by the time the cat is observed, the wavefunction has already collapsed and the cat’s ‘aliveness’ is completely classical.

The idea that ‘anything can happen because of Quantum Mechanics’ is another myth. It is to some extent true, but often abominably misinterpreted by unthinking readers, perhaps a situation exacerbated by episodes of The Big Bang Theory and chapters of The Hitchhiker’s Guide to the Galaxy in which references to this idea are made through jokes and invented technology such as the improbability drive without making clear the true nature of the numbers behind these events happening. The probability of a familiar object visible to the naked eye such as a coin spontaneously appearing on Earth within a person’s lifetime is probably smaller than the probability of someone winning the lottery every time for the entirety of his/her lifespan (which itself is in the order of about 10^(-105) ). It is in fact so unlikely that it is, in all practical sense, impossible (an argument often used, and bizarrely often rejected, to support the case against God).

So what about the coin?

In conclusion, the quantum mechanical description of a tossed coin is identical to its classical counterpart. The probability of obtaining heads is the same in both models; the coin spins and is affected by the air in the same way in both models; in neither model is it possible for the coin simply to disappear in mid-air: the coin never has strange quantum-mechanical properties such as superposition or ill-defined probabilities. By obeying the correspondence principle, QM completely describes classical mechanics and in theory, everything previously explainable and predictable by classical mechanics with regards to a coin toss can be fully explained and predicted by quantum mechanics.

Bibliography

1. Gell-Mann, Murray. The Quark and the Jaguar. s.l. : Abacus, 1995.
2. Wave function of the Universe. Hawking, Stephen and James, Hartle. s.l. : Physics Review, 1983.
3. Hawking, Stephen. A Brief History of Time. s.l. : BCA, 1996.
4. Gillespie, Daniel. A Quantum Mechanics Primer. s.l. : International Textbook Company Limited, 1973.
5. Ballentine, Leslie. Quantum mechanics: a modern development. s.l. : World Scientific Publishing Co Pte Ltd, 1998.
6. Gribbin, John. Q is for Quantum. s.l. : Simon & Schuster, 2000.

Bam! School’s in. The inevitable has happened and the most hideously busy few months of my educational life have begun: the university applications term. And to support this packed timetable I reckoned I’d need some pretty decent technological backing. In other words Windows Vista had to go.

I’d been wanting to rid my laptop of this slick but extremely ungainly and messy monster for some time (ever since I bought the laptop actually) – it crashed at a very embarrassing moment at Young Rewired State (plugged into a projector in front of a big audience with lots of media cameras pointed at me…) which was probably the last straw for me, and it’s had reliably frequent networking issues (wifi and ethernet took it in turns to fail). After testing out Win7 on old machines and liking it, I decided to go for a dual boot with Win7 and Slackware – my opinion is that Linux shouldn’t be all about Ubuntu, Fedora and Debian and I reckoned this would be a good opportunity to give a less mainstream distro a try. There was of course a snag: the main reason I held off uninstalling Vista was that I had a setup of Firefly installed (for my freelance work) which had taken literally half a day to set up so I wasn’t too keen on losing that; clearly virtualisation is the key to the problem here. So I created a new VM in Virtualbox and installed Firefly and supporting software (SQL server, IIS etc.) onto that.

So I backed up my data and installed Win7 and Slackware side-by-side. The last thing Vista did for me was crash while trying to resize a partition so I took a chainsaw (= cfdisk) to my HDD and annihilated Vista. OK I’ll admit the reason I deleted Vista was because I screwed up and hit ctrl+c while cfdisk was running … oops. And I somehow managed to forget to install LILO (boot manager) at first so started off with no boot OS which was just a tad concerning … but in the end everything worked out! I’m using a ~100GB NTFS partition for Win7 + Windows programs. Slackware has ~20GB of ext3 and there’s a 100GB NTFS data drive.

Partition Table for my HDD. In hindsight I should really have used a separate partition for Windows programs and a swap partition. Oh well... Click to embiggen

VirtualBox running in Slackware. It's so much more convenient using virtual machines - they can be transferred from Win7 to Slackware or my desktop or even to someone in China; the portability factor is seriously useful

Linux on laptops, especially Slackware, is all about hackery and cool stuff so I’m hoping to implement sometime soon multitouch gestures (which Mac users have) and customise the OS beyond recognition.

Goods

I was also apprehensive that Windows 7 would fail to realise my laptop has a wifi card thus negating the entire point of the operation, but in the end everything works and wifi was set up as part of the installation process; even standby and hibernate which always crashed the computer in Vista work in Win7! Every time I use the OS I find some small but hugely awesome new little feature that makes me love it just that little bit more. MS have got it right this time in my opinion and I hope it sells well. And I couldn’t help but feel just a little smug when one of my friends came in with a new laptop with Vista on it complaining to me about a wifi failure…

Slackware’s also pretty great – it installed literally in about 15 minutes. I was slightly annoyed at first that it doesn’t have a package manager like apt-get or yum (or at least I can’t seem to find one) but actually now I find svn and make/make install more than adequate substitutes; although installing software is now a lengthier process I get up-to-date packages and have more control over installation. Even better (for me), Slackware starts with a command line and the GUI has to be started manually. So next time someone asks to borrow my laptop to check email be warned: you’ll be using lynx!!

Problems

There are still two things I can’t work out:

1. I’ve got all my Virtualbox information in the shared drive and I’ve managed to boot VMs in both OSes. However if I save the machine state in Win7 and then boot into Slackware it doesn’t seem possible to restore that saved state. If I run that state-saved VM in Slackware and save a new state then return to Win7, Win7 restores the state saved in Win7, not the one in Slackware. A perplexing problem – google time methinks.

2. Linux in general is allergic to Intel Wifi cards. Enough said; though I managed to connect through wifi in Backtrack 3 (not 4 beta though!) so maybe if I do a little driver shuffling it might work eventually.

Overall I’m pretty pleased with this. I didn’t intend to do much advertising in this post but I would certainly recommend Win7. At least give the RCs a try – it is *so* worth it. And having two very good OS’s should give me a huge amount of freedom: Win7 does what I *need* and Slackware does what I *want*. Perfect.

I’ve spent the last two days at the Google HQ in London attending Young Rewired State [hit link for more info about event] (#youngrewiredstate), and it’s been nothing short of epic.

And of course, I’ve taken some photos.

The schedule (shamelessly copied from the site) was as follows:

Saturday 22nd August:
10:00 Start
10:30 Planning session
12:30 Lunch
13:30 Hacking starts
17:30 Dinner
18:30 Home (Hacking overnight allowed!)

Sunday 23rd August:
10:00 Back to hacking
11:30 Brunch
12:30 Back to hacking
16:00 Presentations to Judges and Press
18:30 Prizes announced

On the first day we split into groups and started thinking up ideas. At about 4pm we finally settled on our idea: to make something very similar to RentACoder, but much simpler, targetted at talented coders who need experience in order to get a proper job. Here are a couple of screenshots of the final result (click to embiggen).

We decided on a PHP/MySQL project and as luck would have it, I was the only PHP/MySQL programmer in the group! So it was fairly frantic work (solid coding from 10 till about 3 on the last day) and we ran into all sorts of problems with versioning and people overwriting each others’ work in FTP, especially as the CSS people tended to be working on the same files as I was at the same time!

IRC

As with all hack days, IRC was one of the most important methods of communication. Literally everyone had their laptops out during talks, especially during the presentations at the end and there was a fairly constant stream of chatter on the channel. @samhale123 also put up a bot on the channel to tweet things over IRC – we had several hours of fun attempting to overload the script / twitter / the server!

Immaturity with Twitterfall

Google

Google is an amazing place with by far the best decor I’ve seen in a company building. The floor is laid out like the London underground and the meeting rooms are more or less in the right place for stations (with consistent naming). There are ducks on the ceiling and random awesome other bits of furniture / decor adorning the walls / ceiling / floor.

We were also given a load of Google freebies, including Google yo-yo’s, Google cakes, Google water, Google pens, Google notebooks…

This actually was a telephone box!

Google and Youtube Cakes

People

Of course it was a floor full of geeks, which essentially means a brilliant selection of geek T-shirts (I spotted several from ThinkGeek, at least one from the xkcd store…). The mentors (helping out with coding / guiding the groups) were also working in all sorts of fantastic companies; one of our mentors is working at last.fm, one at moo, one with the BBC etc. And needless to say there was a wide array of OS’ – the large majority seemed to be using Macs, those with PCs were probably split 50/50 between linux (mostly ubuntu, one debian that I know of) and windows.

There was also a brilliant selection of judges, including people from Wired (for some reason looks very familiar; came to school to give a talk maybe?), C4, etc.

Some of the judges

The presentations were good fun – there were something like 40 people from the press / outside making the buzz all the more exciting. And we (@workforpeanuts) won the “Wish I’d thought of that” award!

Anyways, this is the first hack event that I’ve ever been to, and if this is anything to go by, I’m definitely game for another at some point. Heck, maybe DEFCON next year… *MANY* thanks to @hubmum for organising such an amazing event.

And I took other cool photos so go for it and browse!

Today concludes another fantastic week working at Imperial. This post will probably be a lot less massive than the previous one owing to time constraints (I’ve just come back from several exhausting hours of rock climbing at the Westway and Google Calendar tells me I have my driving theory test at some point in the near future).

Firstly, some pics to support stuff from my last post.

The entire table looks like this - I can understand why everything needs to be realigned and tweaked every 15 minutes for the experiment to work! The two blue lasers are I think the main cooling lasers, pumping from the ground to the high energy levels of Ca+.

This is a used Cu O-ring (actually called a gasket) - you can easily see where the knife edge bit into the Cu making a vacuum seal

A top view of the laser setup. You can see the diffraction grating (with an arrow drawn on it) and the connections to the piezo behind it. Click to embiggen.

Equipment

Tantalum Oven

This is the equipment used to produce neutral atoms which are to be ionised.

The oven is suspended between the two electrodes by a Ta wire. The Cu foil is there for a test run of the oven - if it works we should be able to see a spattering of grey Ca on the Cu.

This is the plate on which the entire experiment (ion trap, oven and all) will sit. It will get inserted into one of the holes in the central 6-way cross can.

When I first heard they were going to use an oven, for some reason I imagined some sort of miniature baking oven that somehow emits atoms when turned on! The actual oven is actually a tiny 1cm long tube of tantalum (Ta), sealed at one end by essentially squashing the end, with a tiny hole in the middle of the tube. Ca shavings are stuffed into the open end before the oven is closed, again by squashing. The whole oven is attached by a piece of Ta wire to two electrical contacts across which a potential is applied. The Ta conducts current and heats up, acting as a heating filament. The Ca heats up and the most energetic atoms spit out of the hole (the process is basically evaporating the Ca at very low pressure and high temperature).

I asked why Ta is used – presumably Tungsten (W) has virtually the same properties in that it heats up when current flows through it, and since W is the metal of choice in light bulbs, presumably it’s cheaper? Apparently W can indeed be used; for such applications the criteria for metals are that they are UHV-suitable (don’t trap/adsorb other molecules/atoms on/to their surface which subsequently outgas, ruining Ultra-High Vacuums) and won’t melt at high temperatures. However Ta is normally more suitable than W because it’s more malleable (whereas W is very springy) and can be easily spot-welded (to stick the wire onto the oven). However thoriated W is better than Ta as an electron source since the thorium gives it a much lower workfunction, allowing more electrons to pop out for the same energy input.

Vacuum Pumps

In my previous post about this work experience I omitted some detail on the pumping that I learnt this week. As it turns out, the actual pumping requires three pumps. The first is a roughing pump, to get the pressure down to a very rough vacuum (~10^-2 mbar). Here they were using a rotary vane pump:

This image was nicked from wiki

Essentially as the off-centre internal cylinder turns, the vanes get longer / shorter accordingly such that the pressure at the input is always getting lower and at the output it’s always getting higher, forcing the air out of the output. The principle is essentially PV conservation.

The second stage is a turbo pump which is basically an electrical version of the intake fan of a jet engine. It spins extremely quickly (so quickly that it requires a low pressure to operate lest it smash itself to pieces) and the idea is that it spins so quickly that any molecule that hits a spinning blade hits the part of it such that it gets kicked outwards, away from the vacuum. This gets the pressure down to about 10^-5 mbar. The final stage of course the ion pump.

When air is pumped out the can sits in an oven - the idea is to heat up anything that can outgas while pumping, making it outgas more, thus getting rid of 'outgassable' stuff

Wavelength Tuning using Iodine

There are several ways of tuning wavelength (I wrote something about the cavity method – setting up a standing wave – in my previous post), but I found this way of doing it particularly interesting. Like all other elements, iodine has a certain absorption spectrum, a feature used in star spectroscopy to determine elemental composition. But instead of doing what astronomers do (measure wavelengths to identify elements), here we were using a known map of iodine’s spectrum to tune the wavelength: light shining through the iodine has a certain attenuation which is dependent on the wavelength (owing to electron energy levels). By shifting the wavelength around using a piezoelectric it is possible to obtain a local iodine absorption spectrum (wavelength against intensity). By comparing this local spectrum with an ‘atlas’ – iodine’s spectrum for a large range of wavelengths, it is possible to locate the local spectrum within this atlas, thus identifying the wavelength. Apparently a narrow band of local spectrum is sufficient to identify a unique location in the atlas: there are no ‘repeats’. Whether this is non-repeating property is specific to iodine (hence its use) I’m not sure; the isotope used is radioactive so there must be some really good reason to want to use it!

Techniques and Procedures

Saturated Absorption Spectroscopy

All atoms radiate photons. However in a cloud of atoms these photons are affected by the Doppler shift owing to the random movement of the atoms, and a graph of frequency against intensity (basically a spectrum) shows an underlying distribution for this radiation. However the interesting bit of spectroscopy occurs on the surface of this curve, in the form of ‘ripples’ on the underlying curve’s surface. While a human can normally see the ripples roughly by eye, the underlying curve gets in the way of accuracy.

The solution is a method of somehow obtaining the underlying distribution without the ripples using lasers and subtracting this curve from the spectrum, resulting in a graph of just the ripples. I’m still clueless as to precisely how this works / is performed since I didn’t personally bear witness to the process (I heard something about matching lorentz curves to points but that was probably more to do with analysis of the ripples rather than the process of saturated absorption spectroscopy) so it looks like some wiki-ing is called for.

Walking the beam

This isn’t some physicist’s attempt to be a pirate and getting the words muddled; it’s actually a rather clever (though extremely time-consuming) method of aligning a laser beam. Bascially the ideal situation is a laser beam passes precisely through two points. This is very difficult to achieve with just one stand so a setup with mirrors is necessary. Here are several different failed attempts at diagram-ifying the thing:

The dotted lines were added by me to show where the beam will go

At each of the two points the beam needs to go through an adjustable iris is placed (think circular doors in sci-fi films), and mirrors alpha and beta (making up the periscope) can be adjusted so the beam’s height (h) and angle of elevation (e) can be adjusted more or less independently. Then the following two-step process is iterated until the beam is almost exactly where it needs to be.

1. Open B completely, close A so it becomes a tiny hole, and adjust the laser so it goes through A using mirror alpha
2. Open A, close B so it becomes a tiny hole, and adjust the laser until it goes through B using mirror beta.

Illustrations of the steps are as follows:

For some reason it reminded me of numerical analysis / Newton Raphson type things – constantly optimising and getting closer and closer to perfection yet never reaching it. GL’s cobweb illustration of numerical analysis seemed particularly similar to this situation. Anyways while I quite like how it works, walking the beam does start to lose its novelty after doing it for a couple of hours…

Scanning Tunnelling Microscopy

Danny also explained some awesome stuff on this and how it works. Basically the idea of STM is to use quantum tunnelling calculations to make a map of a surface. A probe is held (say at +5V) very near a surface (grounded), and owing to quantum tunnelling, a certain current flows between the probe and the surface. This current is proportional to exp(-l) (or something like that) so it is possible to measure l to a very high degree of accuracy. As the probe is scanned across the surface, a matrix of measurements of l against (x,y) can be created, thus mapping the topology of the surface. This mapping can in fact be so accurate that it can pinpoint individual atoms sticking out from / adsorbed to an otherwise flat surface.

Conclusion

The last two weeks have been nothing short of awesome. I’ve learnt (and sometimes noted down) many new things on every one of the last ten working days and I’ve recounted here and in last week’s post merely a handful of the more interesting bits and pieces. I even solved an apparently insurmountable practical problem thus moving the entire scientific community forwards! I mean … I came up with a (pretty good) solution to unscrewing a stuck nut… Many thanks to Danny Segal for giving me such a wonderful opportunity.

There is one thing I still can’t work out though:

5W Green Laser used for pumping Titanium-Sapphire Laser

I’ve been looking forward to this for quite some time – two weeks of work experience in a lab at Imperial College working with some PhD students with an ultimate goal of making some progress towards the construction of a quantum computer that doesn’t take tens of man-hours to perform each calculation. After a week I feel I’ve learnt a lot about formal lab work and large-scale experiments (*slightly* different from the 20-minute assessed practicals from AS!) and about the general physics and concepts behind some of the experiments and equipment – I’ve been (not so) conscientiously filling pages of my notebook with messy notes and cryptic diagrams so hopefully some of the stuff I write here will make some vague sense and not quite directly contradict truth.

There are also several interesting bits and pieces lying around the place. There’s an enormous Newton’s Cradle in which each ball looks like it could be heavy enough to be a ship’s anchor. There are also enormous capacitors lying around everywhere for the people working on high-density fluxes.

Capacitor banks lying around

The Building

The Physics / Maths / Computer Science part of Imperial is somewhat bizzarre – from what I gather it consists basically of two adjacent buildings which were built at different times and were haphazardly connected together by knocking down bits of walls. Unfortunately the actual floor levels are out of alignment and the floor heights are also different, which means there’s a crazy staircase joining the two buildings together and floors 6 and 7 in one of the buildings had to be rechristened 6 and 6M for the sake of keeping the numbering consistent with the lifts. There are also only connections on certain floors of each building so it’s possible to leave the Blackett Lab on the bottom floor, go up one level and be confronted with a solid-looking wall where the connection should be. As if things aren’t crazy enough, there’s a set of lifts placed almost exactly at the junction (so to speak) between the two buildings, making it really confusing to navigate the whole 3D maze. It’s pretty good fun actually!

The Lab and Equipment

Optics

I was quite surprised when I first saw the lab – I was expecting a Leonard Hofstadter style lab (as seen on The Big Bang Theory) but actually quantum computing with ions (ions therefore being the main focus of most of the projects which I’ll come to later) involves a fairly large amount of optics work, so each of two adjacent, connected labs I was working in has an optical table as its centrepiece littered with lasers and ridiculously complicated setups of mirrors and lenses which have been tuned very accurately to direct laser beams into tiny optical fibres and whatnot. Speaking of accuracy, the setups are so sensitive to small shifts that they need to be tuned almost constantly. The PhD students told me they detect a lot more drift during the daytime when other experiments are going on in other labs which release radio waves and traffic is rumbling overhead (despite being two floors below ground level and over a block away from a small road) than at night when there is less activity.

The equipment is also sensitive to tiny temperature fluctuations. Most of the lasers are basically diode lasers:

It’s a fairly standard laser setup in which electrons and holes come together in the depleted region between n and p type semiconductors and then either wait for a nanosecond or so before annihilating and releasing a photon (spontaneous emission) or get hit by a photon, resulting in stimulated emission. What was experimentally interesting was that the cavity length in fact determines the wavelength of the laser owing to the fact that a standing wave needs to be created which ‘fits’ exactly in the cavity (a whole number of half-wavelengths need to fit in the cavity) and this is sensitive to temperature. So each laser box has four BNC sockets: one for providing the laser with electricity, one for a thermistor which is hooked up to a feedback loop system which regulates temperature using a Peltier junction heat pump (which occupies another socket on the laser), and one for a piece of Piezo (placed on the diffraction grating) which can change width depending on the voltage across it (or maybe current through it, or something) thus allowing the cavity length to be adjusted, though my suggestion to manipulate the piezo in the feedback loop to compensate for temperature changes would fail since the temperature-dependent expansion of the cavity is several orders of magnitude greater than anything the piezo can correct. When I heard that I was pretty astonished the laser cavity had to be adjusted to such an exact length – several orders of magnitude more exact than the expansion of a bit of metal when raised by a few degrees. The entire laser is covered by a thick black piece of foam to protect it from temperature fluctuations in the room.

Electromagnets

It was pretty cool to find out that I was to be working in the same room as a 2.5 Tesla electromagnet! Ion trapping, as I will also come to later, involves not only charge and potential fields but also magnetic fields, so the Penning Trap the researchers there were using was sitting inside an enormous superconducting electromagnet.

The superconducting electromagnet uses liquid He to keep cool – as a sidenote I asked why they (and CERN) use cool superconductors (more expensive liquid He) instead of the more recently discovered crazy warm ones (cheap liquid N₂); the reason is because above a certain current, superconductors end up failing and develop some resistance causing heat to be produced resulting in a quench (the He boils off, expands to something like 15x its volume and the whole can explodes in a fit of freezing fury), and the cool superconductors can carry a much higher current before this happens, allowing more powerful electromagnets. Of course this comes at a very high cost. As can be seen from the diagram, the He (at ~4K) is shielded from room temperature by a layer of liquid N₂ (at a balmy ~77K). The He needs to be replaced about once every couple of months, while the N₂ is replaced about twice a week. The superconducting coil, power supply and cables are eventually going to have 80A coursing through them – a truly formidable current!

Refilling with N2

N2 is very cold!

Apparently the way they get the electromagnet to start conducting current is to simply arrange the coil in a loop – they can’t expose the 4K superconductor to air, so it is necessary to induce the current in the superconducting coil. Once this is done, the power supply can be switched off and the current in the superconductor just keeps going round (owing to the lack of resistance), allowing a very strong noise-less magnetic flux to be produced (the flux’s precision is something like 10^-6%)

Vacuums

This was particularly new to me. I’d never worked with 2.5 Tesla or 5W lasers before, but while I’ve come across magnets and lasers in experiments, I’ve never really observed experiments involving vacuums before (apart from the bell-ringing-in-a-jar/gerbil-squeaking-in-a-jar one to show sound doesn’t travel well through a near-vacuum). There’s a lot of novel (to me) and interesting experimental stuff that goes on here.

Basically the idea of creating a seal when joining two flanges together is to use a copper O-ring. Each flange has a ‘knife edge’ (90° very sharp edge) and when they’re pressed together with a Cu O-ring in between, the knife edges cut into the soft Cu; thus the Cu itself becomes the seal.

While flicking through Inward Bound by Abraham Pais (recommended by CAPS) I read about various attempts at making a good vacuum pump. Modern technology has come a long way since the mercury-filled jar, and now creating a very good vacuum is a multi-stage process. First all the equipment is cleaned thoroughly – for some reason fats and oils from people’s hands (for example) are disastrous for a vacuum so everything needs to be wiped squeaky clean with something like acetone or isopropanol. Then everything is sat on the optical table which has a source of clean dust-free air on the ceiling which constantly blows on the equipment, keeping dust off and constantly cleaning it of bits of dust that have settled. Then everything is put together using gloves, nuts, bolts, Cu O-rings and *a lot* of effort (believe me, putting flanges on sideways while stopping the Cu O-ring from slipping out is infinitely more difficult than measuring SHM of a cork in a tub of water – reference to AS practical; one of the researchers also described putting He into the cannister having first cooled it sufficiently to stop everything boiling off immediately as a dark art rather than a science). The air is pumped out using a conventional pump until the pressure inside is something like 10^-6 millibars, at which point an ion pump is turned on to essentially evacuate the remaining air molecule by molecule. The pump essentially ionises the gases inside the chamber and use charged plates to attract them out. The final result is a very good vacuum.

The Projects

Photon Ionisation

There was a MSC researcher from Germany sharing the lab with the PhD students from Imperial, and he was working on a different method of ionisation. The supervising prof, Dr Danny Segal (a reader in Quantum Optics), explained that the previous approach to getting ions was to use a ’splat gun’ approach – basically a stream of neutral atoms from an oven hits a stream of electrons from an electron gun, and those electrons will tend to knock out some electrons from the stream of particles, resulting in a few ions. This has a few problems: lots of atoms never get ionised so end up getting deposited on the side of the chamber, screwing up the shape of the potential well in the ion trap; lots of electrons end up floating around in the chamber and get deposited on insulators, again causing irregularities in charge distribution.

The German MSC researcher was working on using photons to create these ions – a much more tenuous stream of neutral particles is projected into a beam of photons which, via the photoelectric effect, knock out electrons creating ions. This should have a higher rate of ionisation leaving fewer ‘waste’ atoms sticking to the inside, and the number of photoelectrons knocking around the chamber should be much lower than the number of electrons being shot from the electron gun. I suggested the photons might knock electrons off other bits of the apparatus, again screwing up the flux; apparently this should happen infrequently enough to allow a reasonably controllable flux, though some researchers using a Paul trap (involving an oscillating EM field) apparently detected ionisation using photons of the wrong frequency for direct ionisation leading them to believe electrons were being knocked from the apparatus and these, accelerated by the oscillating field, slammed into atoms causing ionisation.

Anyways the setup was more or less thus (a picture is worth a thousand words):

The way to ionise these Ca atoms is to first use a laser to excite the atoms – push some electrons up to a higher energy level. The UV LED then does the actual ionisation from that energy level. The picture above is of a half-finished setup (optics haven’t been sorted out yet and there are two unsealed flanges).

Laser cooling

GSM/KPZ gave us an article in class last year about laser cooling (‘Cool things to do with lasers’, Ifan G Hughes et al 2007), and it turns out it’s useful for Quantum Computing – a jittery ion is presumably pretty bad for physicists who want a stable wavefunction. Well, here’s the setup.

Click to embiggen

It’s in fact mostly about using lasers to manipulate electron energy levels in a Ca⁺ ion:

The Ca⁺ ion has an energy level electrons can fall down to (RHS of diagram) where they would stay for quite long before falling back down which is undesirable considering the cooling involves shuttling electrons between the leftmost levels (in the diagram). So four red lasers are required to pump those back up to the top energy level.

QED

I’m not really sure what’s going on in this experiment but basically, since QED is only significant at high charges (something like that), the researchers go to GSI to conduct this research. The idea at GSI is to slam super-high energy ions through a gold foil which apparently strips them of all electrons. Different ions are separated via a very similar system to how a mass spec works.

Some Other Physics-ey Stuff

Ion Trapping

There are of course lots of different methods of trapping ions – I mentioned one in my post about the UCL antimatter lecture. Apparently it’s provable from Maxwell’s equations that it is impossible to create a static 3-dimensional potential well to trap ions, so there are currently two main methods: using a purely electromagnetic system (using either some feedback system to wobble the ions towards the centre of the trap or a constantly oscillating field like in a Paul trap), or to use magnets:

A cation is sitting at the bottom of a potential well in the z direction. It is surrounded in the xy plane by oppositely charged plates. As it is attracted to the plates, the z-directional magnetic field causes it to move in a circular motion (as seen in cloud / bubble chambers to determine momenta of ejected charged particles), represented on the diagram by ‘micro OOO’. The charge on the plates are then somehow tuned to make the large-scale motion of the particle resemble a circle and so it eventually loops back on itself, so its path shape looks like what is labelled in the diagram as ‘tuned, get O’

The Ca⁺ Ion

The actual quantum computation to be done with the Ca⁺ ion (not a typo: just one +; this isn’t chemistry!) involves electron energy levels. An electron can be in one of two energy levels, and that is the qubit. In Ca⁺ there are two more or less independent distinct situations in which an electron can be in one of two energy levels, allowing two qubits to be encoded into one ion.

The use of this isn’t only to cram more qubits into fewer ions (I read a research group somewhere is making base-5 ‘qudits’ using microwaves and superconducting things) but also to allow easier entanglement – since both qubits are in the same ion it’s supposedly easier to make them interfere in a predictable manner, which allows a quantum NOT gate to be set up which is critical to quantum computing; supposedly only two research groups in the world have managed to get this quantum NOT gate to work.

Ion hopping

The biggest limitation apart from the sheer fiddly-ness and slowness of everything in the quantum computing world is the fact that it’s impossible to put more than about 8 ions in one trap before they start screwing up each others’ wavefunctions. The PhD researchers had previously been working on a solution to this problem that the theorists came up with – getting the ions to hop around in the trap, thus manipulating each ion more or less individually. This has already been done using Paul traps (I think) but the researchers here were trying to use Penning traps and show they are in fact better for quantum computing (or at least can do the same things as Paul traps).

Overall

There’s a lot more I would say if I had the time but as with all blog posts, you’ve got to stop somewhere. But overall I’ve never done modern practical physics before (at UCL we looked at some particle traces on the computers which is the closest I’ve really got so far) so this is a pretty damn amazing experience for me, hence the mega-post.

Introducing: the world's smallest allen key!

Last week I was in Cranfield participating in the Aerospace Challenge Finals. The challenge this year was to come up with a design for a device to drop humanitarian aid accurately (within 20 metres of a target) from 3000 metres up. Our idea managed to make it to the finals which turned out to be a week of lectures on general aerospace engineering, activities and flying! Photos are here.

Flying

Each person got two flying experiences, both of which included some time piloting the aircraft: about 10-20 minutes in a helicopter and about half an hour in a fixed-wing plane.

My first flying experience was with a small Robinson helicopter, which can only really be described as terrifyingly, exhilaratingly awesome. The pilot managed the take-off which was one of the most breathtaking experiences I’ve ever had – in a helicopter you’re literally sitting in a big transparent flying bubble with the engine behind you, so the view and experience is truly amazing as the land falls away beneath you… I later took over and found control extremely difficult – even a tiny movement of the stick causes the vehicle to tilt violently in that direction making a beginner like me very prone to overcorrection leading to a serious case of increasing-amplitude SHM! The actual stick is situated between the pilot and the copilot and a rotating handle is stuck on the end allowing dual control, so my rather flailing and uncontrolled flight was abruptly and expertly rectified when the pilot took control (though not before I turned and prepared to land by erratically lurching towards a patch of grass). The pilot then demonstrated some cool things one can do with a helicopter including skid landing and take-off, going backwards and sideways while spinning etc.

Here you can see how control over steering is shared between pilot (me) and real pilot (instructor)

The next day I got in a PA28 – my first fixed-wing experience. The pilot had to go through an enormous list of things to check before taking off and explained a little about what she was doing (mostly checking the engine could rev at certain RPMs and wouldn’t give out in certain situations, flicking on and off various lights and calibrating [and pointing at] instruments). The runway was also ridiculously long so she didn’t even bother with flaps for takeoff. This was much easier to fly than the helicopter and the dials and instruments in the cockpit didn’t obscure the view as much I had inferred they would from MS Flight Sim’s portrayal. I did a few rather ginger turns and pitch adjustments before relinquishing control back to the pilot who then demonstrated some steep banks, a stall (which sounded dangerous and seemed to imply the engine cutting out) and a dive (which was extremely cool). Later that week Matthew and I were inspired enough to ask about possible places to get flying instruction – flying has always been one of those things I’ve wanted to learn but I’ve always ended up not having enough time or money to start…

Here the instructor is doing a steep bank. She even did a pretty steep dive totally relaxed and with that pen in her hand!

Me flying the PA28!

Activities

The week started with some group leadership exercises which consisted of attempting to place 30 cards in the correct pattern (easy) and work out the shape and colour of two missing shapes while blindfolded (hard) – both were much more enjoyable than I had expected from that genre of exercises.

The first engineering challenge we were given was an egg-drop challenge – the idea was to construct a package which will protect an egg from a drop of 4 metres. We were given limited materials and each material had a price; the idea was to make the cheapest package that doesn’t crack the egg. Our attempt turned out to be the most epic non-fail in history – literally seconds before the end of the construction phase we managed to pop two balloons which made us completely change our plan and in the last few seconds and in great haste we crammed stuff into a crumple zone and added a parachute … and it somehow worked and turned out to be the cheapest package (if wastage is deducted)! I guess that really proves the KISS principle: Keep It Simple Stupid.

The second engineering challenge was along similar lines – dropping aid – though it was from a more macro perspective. The game was called ‘airlift’ and sold by Elite – the idea was to plan an air route through several African villages which uses the least fuel, while dropping packages of aid which we had to construct out of wooden blocks, paper and tape while making sure everything fits in the cargo hold. The first thing I pointed out when time started was that both problems were NP-complete: the packing problem was almost exactly the same as the knapsack problem and the route planning was basically the Travelling Salesman problem with fuel added in as a factor. In other words we had to be either very good at intuitive problem solving or somehow get lucky. As it turned out, as perhaps a combination of the two, we somehow managed to come up with both the the optimum packing configuration as well as the best route, and finished literally as the final buzzer went – not bad!

The rest of the week was dotted with things like paper plane competitions (which included an awesome flying paper ring which seems impossible when you first see it fly), a game of (actual) CTF and some sports.

Lectures

Over the week there were daily lectures. Much as I would love to discuss them all here in depth I haven’t got that much time / space and besides most people aren’t as interested as I am in the effect of negative angles of attack… But I’ll go a little into some of the most interesting lectures.

Fly by Wire (FBW)

The problem for a long time had been that when going sufficiently quickly, adjusting the controls from the cockpit was really quite hard work – the air going past has so much momentum and the mass flow rate is so high that to change its direction by (for example) adjusting the ailerons requires a lot of force. To make things worse, at supersonic speeds a shock cone is developed (some awesome videos of this are on Youtube) – if this touches the aileron the stick can be wrenched out of the pilot’s hand. Some of these controls were partially solved by making the stick adjust small tabs in the wing instead of the entire aileron, reducing the force required to steer, and by making controls non-reversible (force on the aileron doesn’t affect the flying stick). There are of course some problems with these such as lack of ‘feel’ of the controls. So recently manual stick-aileron transmission was replaced with an electronic motor which receives instructions from the cockpit and adjusts the ailerons itself. Not only does this take all the strain off the pilot, but it also allows a computer to neutralise bad judgements on the part of the pilot such as initiating a sharp dive at 50 feet, implemented by a feedback mechanism from the aircraft to the computer. It also simplifies the cockpit – instead of filling the area with controls, dials an instruments, a computer screen with a joystick and throttle suffices to fly a FBW plane. I asked whether, since FBW significantly reduces the pilot’s direct control over the aircraft, FBW might actually make complicated manouevres more unsafe or indeed completely impossible. John Farley, who was giving the talk, said that, from his vast experience, pilots, however experienced, cannot really be trusted to fly planes safely all the time, and in fact he would feel safer trusting a computer’s judgement and letting a computer do such manouevres than a pilot. That talk also proves that a Boeing 747 probably has non-reversible controls so that scene in Snakes on a Plane (I think it was that film) in which the pilot asked the co-pilot to help pull back on the stick very hard was probably a load of rubbish. Not that you needed to be told that.

Basic Aerodynamics

One of the interesting things from this talk was the reasoning for why helicopters don’t go fast. There is always one part of the rotor going forwards, and if the helicopter moves forwards sufficiently quickly that part of the rotor travels at supersonic speeds generating a shockwave that could rip apart the rotor. In addition, even at lower speeds, there is an imbalance between the airspeed of the fowards-going part and backwards-going part of the rotor meaning a gimbal has to change the angle of attack of the blade depending on which way it’s going: the angle of attack of the rearwards-going blade has to increase to increase lift on that side otherwise the helicopter would just roll over. Of course, there is a maximum angle of attack this blade can be set to before it stalls which is about 20°. This limits the helicopter’s speed at subsonic speeds.

An RAF Hawk landed at the airstrip for us - here is the pilot demonstrating how the entire tailplane rotates

Automation and the future

This was probably the most interesting talk of the week; unfortunately it was cut short for us owing to a jetstream flight. Apparently currently pilots of Euro Fighters get sensor fused info presented to them in the form of advice as to what to do and they simply act upon that, which means half the time the plane is telling the pilot what to do: it is telling the pilot how to control it: semi-automation. Even in commercial aircraft a system called TCAS (Traffic Collision Avoidance System) senses other aircraft and advises the pilot on how to manoeuvre. There is clearly room for improvement: unmanned aerial vehicles are coming. This of course led to the whole humans v computers discussion but for every example of a pilot doing something heroic and saving the plane, there are several examples in which pilots screwed up and computers would have saved lives – Chris Roberts, the speaker, asked whether it *really* is desirable to have a pilot flying the plane, and whether the problem of pilots becoming de-skilled from letting the autopilot take over really is such a problem after all. I also found it very interesting and surprising that currently many landings of commercial aircraft are performed by the autopilot in low-visibility situations.

Anyways overall it was a fantastic week. Whatever the results of the competition turn out to be, I for one got a lot out of six days in Cranfield. I learnt a lot, made some friends, made some good contacts in the industry, and had some great fun relaxing in the English countryside!

Walking in the English countryside

This week has gone pretty quickly and I’ve mostly been working on the text analyser / summary program. I even managed to take some photos! The week started with @dumbledad (= Tim) showing me some of the visualisation stuff he and an intern had been working on to visualise a book, some of which will appear shortly on a site somewhere… It’s all in the spirit of new and interesting data presentation in the spirit of Information Aesthetics and he sent me a link to some stuff he did on ManyEyes – word clouds (or ‘wordles’) comparing frequencies of words in narrative and speech. Some of the other ones are more difficult to describe but I’ll be sure to tweet link to them when they get published.

The idea of the summary program was that it split the book into sections then compared a histogram of word frequency densities in each section with another histogram for the entire book, then picked out the words which were most likely to be important to the section by choosing the most unusually frequently used ones. The problem with that was the program wasn’t picking out main characters because they were being mentioned all throughout the book. So I was to implement a system to split words into three categories: local to the section, local to the book (main characters) and common to the English language. The existing framework for a two-way local to section vs local to book had already been written so I was to implement the three-way split.

Factor graph showing the model

By Wednesday I’d finished the actual implementation so I started trying to invent a visualisation. My original idea was to have a ’story line’ (no pun was actually intended) along which various threads would undulate, and the further out from the story line they are, the more important they are; think of it as a radial graph – I think I was probably inspired by the RealPlayer (yuk, I know) ‘cosmic string’ visualisation. I built a really flickery version as a mockup which was approved, and since I was by then starting to shy away from WPF I ended up learning DirectX overnight to implement a final 3D non-flickery version of it. After spending a whole day stressing over the edges of the scene getting cut off and finally realising I’d set the camera’s maximum viewing distance ridiculously low, I finally got it to work, and after writing some homebrew bezier curve code it looked pretty good (if I may say so myself); Tim tells me he’ll probably add a screen video of it to the online display of visualisations so … watch this space.

Another excitement of the week was a talk from TrueKnowledge (= TK), an internet answer engine. It’s similar to the famous Wolfram Alpha (= Walfa); however in my opinion it actually has more potential. Walfa throws manpower at writing new code to scrape information from various different sources on the fly which essentially means the more information you want, the more you’re going to need to work. TK on the other hand stores information in an enormous database which has a structure suitable for storing any type of information, and although work is done to ‘crawl’ Wikipedia and other sources for knowledge, it also sources the community for information which means it can gather lots of important knowledge very quickly with minimal effort. It also has awesome features of natural language parsing (ask it ‘what colour are red cars’ for example) and it can also give you a step-by-step explanation of the logical process that leads to its final answer.

The bottom half of the screenshot shows TK's stages of logical inference

It of course differs from Walfa in that it hasn’t got a tonne of Mathematica code behind it – its strengths are in factual and inferred knowledge as opposed to evaluating integrals. It’s currently in Beta and has an API (yay!) so I strongly encourage anyone who has used Walfa to give TK a go.

On Tuesday the weekly Mexican food van appeared – until then I’d never realised quite how amazingly good burritos can be! While we were eating we started discussing presentation of text. The problem is that a conventional layout presents the reader with a formidable block of text interspersed with some images which is difficult to follow and annoying to read since one always has to alternate between studying the image and reading the text. However attempts at producing non-linear presentations of information such as embedding text into the image as tooltips or expandable areas of the image etc. have always resulted in people simply not reading very much of the text and consequently missing out important stuff. The best solution we came up with is using an old method of collapsible clauses, just like collapsible code. For example, if a relative clause which in this case is italicised and relatively long yet somehow doesn’t contribute much to the sentence thus merely adds length and unnecessary information to the text making the ultimate meaning more difficult to discern is considered superfluous to the meaning of the sentence, it could be replaced by a small button that only shows the clause if clicked – such ideas are particularly relevant to German sentences which tend to have huge diversions into clauses before the verb is revealed right at the end. This way readers can quickly get the gist of what’s going on so they may study the image in an enlightened way, then go back and expand the text to get the full meaning.

There are also a few things I noticed about MSR in general. There is a strong sense of company loyalty – all employees seem to use Bing, and everyone I’ve seen even goes as far as using IE instead of Firefox! Using only Microsoft products to perform tasks however did make me aware of the wide range of programs they do produce – they even have Virtual Machine software and an internal proprietary alternative to SVN. I guess it does help the developers of these applications a lot if they have an enormous internal test group: all the employees and interns. There’s also pretty close integration with Redmond (Outlook + Office Communicator + global WAN shares) so feedback could be quite efficiently delivered. The entire place also operates in the spirit of trust – all users have admin rights (necessary for developers anyway) – which is so much better than what is implemented at school: a highly restrictive policy which, despite recent changes for the better, still filters out most protocols (FTP included) and in fact, instead of preventing people from doing things simply makes everything so much more difficult to do. Now I have to connect through encrypted VPN to use FTP…

Anyways overall it was a great two weeks. I enjoyed it hugely, I didn’t need to touch Excel, I didn’t make anyone coffee and I didn’t do any filing (who needs paper anyway? It’s a software company!) – instead I worked on real (and rather cool) projects, learnt some useful things, and made new acquaintances.

In other news, I’m off tomorrow to Cranfield for the Aerospace Challenge Finals – I’ll get to fly (actual!) planes, take lots of photos and it should be another great experience. They’d just better have wifi, though I’m bringing my Alfa Awus (ridiculously powerful) along in case of weak signal!

I’m currently doing Work Experience at Microsoft Research in Cambridge (MSR Cambridge) and despite my gripes about Microsoft software, the research centre and some of the stuff people are doing there is pretty cool. Here’s what I’ve done so far and what I think of the place / stuff.

I arrived in Cambridge to a 4-storey house all to myself (!) complete with dishwasher, washing machine, decent cooking stuff etc. I saw the wifi router but it wasn’t broadcasting beacons – it was a hidden SSID. So I spent about 3 hours capping packets to no avail on a nearby WEP wifi and finally realised that (a) there was an ethernet cable sticking out of the wifi router which I could plug into, and (b) the router had its hidden SSID written on it.

After waking up the next day to three alarms (alarm clock, laptop, watch; 5 minutes apart from each other; I wanted to make sure I actually got up on time … for once) I left my laptop at home with three layers of Lifehacker-inspired protection (physical lock, Yawcam for motion detection webcam and LaptopAlarm – a thief would have to have his/her photo uploaded to an FTP server and walk around Cambridge with a chest of drawers attached to a laptop that starts screaming the second it’s unplugged in order to steal my laptop) and walked the 1.6 mile commute to MSR in the West Cambridge Site NW of the city. I eventually arrived where I was issued a temporary pass, a Microsoft Research rucksack, some freebies and a ‘wottle’ – MS in an attempt to do the environment some good decided to issue refillable recycled plastic/rubber (?) bottles. The person who came up with the idea evidently had the same lack of aptitude as I do when it comes to names; it’s got ‘wottled by you’ written on the side…

I also met my supervisor who explained a little about what the project I was to be working on (Infer.NET in the machine learning area) is all about – Bayesian inference. It’s basically a system of statistical modelling which takes in information about distributions, observed values and interaction between random variables and infers new distributions by applying Bayes’ Theorem (which SJRH gave us a talk in the dying weeks of term). In terms of the theory all he said was when there are several distributions it gets pretty complicated and uber-badass integration (not his words) is necessary which, until an approximate method was found a couple of years ago, wasn’t really possible computationally. He also introduced me to a couple of the team members, one of whom has a firm belief that random variables should be a part of programming and are as valid as, if not more than, ‘normal’ everyday data types like integers.

Eventually we settled on my project. I’d mentioned the Monty Hall problem when discussing Bayes’ Theorem and he proposed that I work on a display implementing that using Infer.NET as a simple project, and also to give me an idea of how something like that would be implemented using this framework. At that point I was more or less left to my own devices to familiarise myself with the workstation and Visual Studio and to complete a stack of paperwork I’d been given. I think it was at about this point that I discovered I was working on an 8-core Dell Precision workstation. Serious overkill – as Guy pointed out, probably employees spend most of their time gaming and consequently don’t get much done!

My supervisor (I think the legal paperwork of which there was much permits me to give his name – John Guiver) took me out for lunch according to some sort of tradition involving new people joining the team at the Cavendish Laboratory Canteen (the Cavendish Lab was near the MS building). Actually the canteen there was pretty standard (I was expecting wonders from the great Cavendish) though the food is good and cheap – that’s all that really matters. One of the other team members, John Winn (random variables in programming dude) had been to MIT. Apparently they have whiteboards in the loos and people leave each other problems on them, sangaku style!

There was some sort of filming going on in the main lecture theatre – bright white theatre floodlights adorned the roof and several cameras were propped up on tripods all over the place. After lunch I ended up agreeing to help out which mostly involved milling around displays of Infer.NET related programs for a time lapse vid. This turned out to be really awesome because of the way the displays were implemented – massive multi-touch touch screen coffee tables! They could also be operated by circular counters placed on the tabletop, used like dials (twist to change settings), similar to a touch screen I’d seen on youtube demonstrating a physics simulator. There was also a(n incomplete) machine vision display which proudly labeled a pair of scissors as a book…

Afterwards I grabbed myself the Infer.NET dll’s and started trying to implement the Monty Hall problem. There was a problem with the model and I ended up (wilfully) working almost an hour overtime on my first day! I got home and tried for the first time (and failed) to play Dawn of War Soulstorm with Guy over Hamachi-induced WAN.

Thursday was a particularly interesting day. The first thing that happened was, while infuriated with WPF’s apparent inability to do frame by frame animations and in a *very* roundabout method of doing things, I worked out how to do multithreading in C#. It’s not actually that hard at all (1 line of code captures the essence of it) but was just one of those really useful things I’d always wanted to learn but hadn’t owing to preconceptions about ploughing through hundreds of lines of unclear example code from an obscure online tutorial. Traditionally Thursday lunch is an Infer.NET group lunch so I also got to chat more with the rest of the team about what they’re doing. One was trying to help a mathematician with something programmatic and was stressing about why on earth he kept insisting on doing things in unprogrammatic and ultimately unsuccessful ways, which ended up as a big discussion about mathematical v computing methods, and how the two disciplines have drastically different approaches and concepts of ideal solutions despite their fundamental similarities. It also allowed me to use my rather scanty knowledge of linear algebra to justify the mathematician’s approach to writing an algorithm to determine if a graph has loops: create the graph’s incidence matrix and find its determinant. If the det is zero the graph has a loop. Of course finding the determinant is an O(n³) affair using Gauss Jordan (exponential if you use the FP1-style cofactor method) so computationally expensive . Thursday afternoon saw my discovery of how to make dlls in C# using Visual Studio, the second simple yet important discovery of the day.

But most interesting / fun was the evening. A go karting event had been arranged for all the interns (and some employees) as part of a summer series of events and as a work experience employee I’d been invited. I discovered I wasn’t the only first timer which made me feel a little less worried about failing epically; so long as the damn thing doesn’t have a clutch I’ll be OK (just had my second driving lesson). Anyways it was all a lot of fun – I only crashed once because I first overestimated then underestimated the tyres’ grip (it was a classic understeer-oversteer-skid affair) and didn’t lean the right way (apparently you should lean towards the outside of a bend to get more grip which seems to me as a cyclist extremely counterintuitive). I also discovered it’s true that racing makes you ridiculously thirsty; luckily I had my wottle to hand! Afterwards while we were waiting for the coach back I noticed one of the employees was wearing a T-shirt with ‘i bing’ on the front and ‘u bing?’ on the back – apparently people who had worked on the Bing project were all issued them at the launch.

By Friday I’d more or less finished the Monty Hall implementation, so John and I discussed possibilities for next week. One was some intelligent text ‘understanding’ which uses a neat little trick involving Bayesian inference on a massive scale to partition a text into phrases (without using punctuation as a guide). Another alternative was to use statistical analysis on the relative locality of certain words to generate a summary of a very long text. I suggested the possibility of using Infer.NET for auto correcting which seemed viable. I was also led over to the uber-cool gadgets area of the building (‘computers in the home’) which was densely populated with enormous screens, multi-touch touch screen coffee tables, wireless [just about everything] and more or less all the stuff I’d want to have sitting in my room!

John also explained a little about modelling statistics and factor graphs: graphs in which nodes are either variables or functions, and edges are directed. Messages in the form of entire distributions are passed along these edges between the variables and functions and you end up getting complicated-looking graphs with arrows and squares and rectangles. These are used for describing models and the maths is apparently beyond 1st year undergrad… The storm that day was also quite epic – the electricity in the MS building seemed affected as the lights kept flickering out – I assume the PCs were all UPS’d.

Anyways those are my thoughts at half time – I still have another week to go. I haven’t actually taken [m]any photos at all since I’ve been spending most of my hours either working or cooking/eating or sleeping (I’m back in London for the weekend); hopefully I’ll be able to take some next week.