Post-16 education must be reformed to tackle damaging arts-science divide

Stop dividing children into arts and science specialists at an age when they are not ready to choose. Pete/Flickr, CC BY-SA

Education sits at the heart of our society – and politicians know it. When Tony Blair famously said “education, education, education” it was essentially an election slogan. We are constantly told by our politicians that English A levels are the “gold standard” in education. I say, maybe it’s time for a rethink.

At the heart of the problem is the early specialisation in post-16 education. As a practising scientist I like to think that I can at least have some understanding of any science story presented in the news. But for a large proportion of the population that isn’t the case; our society almost seems to believe that the situation is a virtue. If a politician says “I never could do maths” no one thinks “Philistine”, whereas if they admitted to never having read any Shakespeare or Dickens the reaction would be very different. Why does our society think this is OK?

Science underpins so many decisions; political and personal. In our daily life and jobs, we increasingly need to use quantitative skills: the ability to interpret graphs, utilise spreadsheets and manipulate data. Our national academies recognise this, with a recent report from the British Academy to go with last year’s Vision report from the Royal Society, both calling for all students to continue with some form of maths post-16.

This issue cuts both ways of course. Scientists need to be able to write and communicate better. Whether or not they can quote chunks of poetry, ancient or modern, is not the point. Scientists need to be able to write lucidly and put their work in context. Just about every branch of science is going to touch on the human condition and they need to be able to understand what their research means for the public. Some grasp of history, literature and social science could help them communicate this.

So in my upcoming Presidential Address to the British Science Association, I will be urging politicians to reconsider the structure of our post-16 education. England and Wales are unlike almost all other developed countries in our early specialisation. This leads to a damaging divide between arts and science.

Maybe we could learn a thing to two from Renaissance artists like Leonardo da Vinci? wikimedia

Implicitly, at the point of choosing GCSE topics, a 14 year old will see themselves heading off in one direction or the other. Schools sometimes appear to encourage this, perhaps for the simple reason of easing the timetable. A broader post-16 education would mean moving from the typically narrow choices of A levels to something akin to the European Baccalaureate system (or perhaps the Scottish Higher system), where more subjects are studied for longer.

The teaching shortage

Of course, all this would require an adequate supply of qualified teachers. Currently, however, we neither have enough teachers entering the profession nor staying on for long subsequently. This is a massive problem in many subjects.

In primary school teaching, many schools have no one qualified in science or with a maths degree (the Vision report says only 3% and 5% of primary school teachers have maths and science degrees or specialist teaching qualifications in those subjects respectively). In turn this creates a confidence problem: teachers who haven’t looked at a maths problem since they were 16 are expected to teach numeracy skills they may feel unsure about themselves.

This problem is particularly acute when there is no one else with more relevant experience in the school to whom they can turn for specific advice. This is no criticism of the teachers themselves, but, when teachers have to teach beyond their own areas of confidence and competence, it is harder for them to stimulate the children and to answer their questions.

In the sciences a related problem occurs at secondary school. Teachers may be science teachers, but if their qualification is in biology it is tough for them to teach GCSE physics. Again, this is not meant to apportion blame to the teachers. The Institute of Physics has suggested we need 1000 more physics graduates a year entering the teaching profession if we are to reach a situation where a third of science teachers are qualified in physics – and it would still take 15 years.

To do this would need around a quarter of all physics graduates training as teachers each year. It is hard to imagine that happening, particularly given the level of salaries graduates can otherwise command.

England has this strange habit of splitting our children up into arts and sciences at an age when hormones are surging and peer pressure is liable to be at its most powerful. We should be pressing the government to modify our system so that all children keep studying a broad range of subjects post-16 – and providing adequate funding to do so. In time this would translate into primary school teachers with more confidence to enthuse the next generation in maths and science.

Furthermore, this change would empower everyone to be able to make better-informed decisions about the things that affect them in their everyday life and to make sure that day by day people are able to cope with the numeracy requirements of their jobs with confidence.

The fate of the universe: heat death, Big Rip or cosmic consciousness?

Black holes will be all that remains before the universe enters heath death. But the story doesn't end there... NASA/ESA/wikimedia

By piecing together an increasing number of clues, cosmologists are getting closer to understanding what the future and ultimate fate of the universe will be. And I’m afraid the news is not good. Star formation will cease and black holes will take over until they eventually evaporate into nothingness. There could even be a “Big Rip” on the horizon. But for those who don’t mind waiting another 10¹⁰⁵⁰ years or so, things may start to look up as a number of bizarre events could take place.

But before we consider random events in the very far future, let’s start with what we know about the past and the present.

The past

The reason we can investigate the past evolution of the universe is that, in some regards, astronomy is analogous to archaeology. Explicitly: the further we peer away from our home planet, the further back in time we see in to the universe. And when we look far back in time, we observe that galaxies are closer together than they are at present. Although only one strand of evidence among many, this observation – coupled with Einstein’s theory of general relativity – means that the universe started with a Big Bang and has been expanding ever since.

The present

Late last century, one of the most pressing issues in modern cosmology was to measure the deceleration rate of the universe. Given the amount of mass observed in the cosmos it was thought that it might be enough to cause an eventual contraction of the expansion.

Remarkably, two independent teams of scientists found the exact opposite. The universe was not slowing down in its expansion, it was accelerating. This profound discovery lead to the Nobel prize in physics in 2011. However, understanding the implications of it remains challenging.

One way to think about the accelerating universe is that there must be some kind of material (or field) that permeates the universe that exerts a negative pressure (or a repulsive gravity). We call this dark energy.

This may sound a bit far-fetched, but independent experiments have been conducted to corroborate the acceleration of the universe and the existence of dark energy. From 2006, I was involved in the WiggleZ Dark Energy Survey – a scientific experiment to independently confirm the acceleration. Not only did we find that the acceleration is happening, but we provided compelling evidence that the cause of this was dark energy. We observed that dark energy was retarding the growth of massive superclusters of galaxies.

The growth rate of superclusters like Virgo is providing strong evidence for the existence of dark energy. Andrew Z. Colvin/wikimedia, CC BY-SA

We therefore suggested that dark energy is real. If the concept of dark energy and its repulsive gravitation force is too weird, then an alternative to consider is that perhaps our theory of gravitation needs to be modified. This might be achieved in in a similar way that relativity advanced Newtonian gravitation. Either way, we need new physics to explain it.

The future

Before turning to the very distant future, I will mention another relevant survey: GAMA. Using that survey, we found that the universe is slowing “dying”. Put another way: the peak era of star formation is well behind us, and the universe is already fading.

The more “immediate” future can be predicted with some certainty. Five billion years from now, the sun will enter its red giant phase. Depressingly, no more than two more billion years after that, it will consume Earth.

After that, the relative strength of dark energy and how it might vary over time becomes important. The stronger and faster the repulsive force of dark energy is, the more likely it is that the universe will experience a Big Rip. Put bluntly: the Big Rip is what happens when the repulsive force of dark energy is able to overcome gravitation (and everything else). Bodies that are gravitationally bound (such as our local supercluster, our own Milky Way galaxy, our solar system, and eventually ourselves) become ripped apart and all that is left is (probably) lonesome patches of vacuum.

The data from the WiggleZ survey and other experiments do not rule out the Big Rip, but push it in to the exceptionally far future (if at all).

Somewhat more pressing is the heat death of the universe. As the universe carries on expanding, we will no longer be able to observe galaxies outside our local group (100 million years from now). Star formation will then cease in about 1-100 trillion years as the supply of gas needed will be exhausted. While there will be some stars around, these will run out of fuel in some 120 trillion years. All that is left at that point is stellar remnants: black holes, neutron stars, & white dwarfs being the prime examples. One hundred quintillion (10²⁰) years from now, most of these objects will be swallowed up by the supermassive black holes at the heart of galaxies.

In this way, the universe will get darker and quieter until there’s not much going on. What happens next will depend on how fast the matter in the universe decays. It is thought that protons, which make up atoms along with neutrons and electrons, spontaneously decay into subatomic particles if you just wait long enough. The time for all ordinary matter to disappear has been calculated to be 10⁴⁰ years from now. Beyond this, only black holes will remain. And even they will evaporate away after some 10¹⁰⁰ years.

At this point, the universe will be nearly a vacuum. Particles that remain, like electrons and light particles (photons), are then very far apart due to the universe’s expansion and rarely – if at all – interact. This is the true death of the universe, dubbed the “heat death”.

The idea comes from the second law of thermodynamics, which states that entropy – a measure of “disorder” or the number of ways a system can be arranged – always increases. Any system, including the universe, will eventually evolve into a state of maximum disorder – just like a sugar cube will always dissolve in a cup of tea but would take an insanely long time to randomly go back to an orderly cube structure. When all the energy the in the cosmos is uniformly spread out, there is no more heat or free energy to fuel processes that consume energy, such as life.

Boltzmann Brains and new Big Bangs

All of the above seem very bleak to say the least. So I will end this article on a highly speculative, probably wrong, completely untestable, but more positive, note.

Fishy? The far future of the universe could rather bizarre. AK Rockefeller/Flickr, CC BY-SA

According to the strange rules of quantum mechanics, random things can pop up from a vacuum. And it is not just a mathematical quirk: the existence of particles suddenly coming into existence and then disappearing again is seen constantly in particle physics experiments. However, there is no reason why so-called “quantum fluctuations” could not give rise to an entire atom.

There has even been speculation that a “brain”, dubbed a Boltzmann brain, could be created in this context. The timescale for such a thing to appear? Well, that has been computed at 10¹⁰⁵⁰ years.

And a new Big Bang? That could be on the way in some 10¹⁰¹⁰⁵⁶ years.

Facebook's digital assistant blends AI with customer service staff – but will it cope without human help?

M – no Bond jokes please. Facebook

With the arrival of its monosyllabic M, Facebook has introduced its own personal digital assistant, following those from Apple (Siri), Microsoft (Cortana), Google (Now) and Amazon (Echo). Technically, M operates partly on the user’s smartphone via the Facebook Messenger app, but it is mostly a cloud-based service. Unlike the others, however, this isn’t just an artificial intelligence but a mix of smart machine learning and human assistance.

What makes M different is that it takes recommendations or answering queries one step further, able to actually make purchases or arrange services for you, and order deliveries. This is the logical conclusion of recommending something, allowing the system to spend your money for you as well. This approach might be risky, or might be brilliant. If it works, suppliers will be clamouring for Facebook’s M to spend users’ money with them, and Facebook will be able to take a percentage in return.

With Facebook’s enormous reach – the site recently claimed one billion concurrent users – even a small percentage of such a large number of users spending even relatively small sums of money would still add up to a great deal of cash for Facebook. Mind you, a few unfortunate misunderstandings of what a user wants to buy might lead to some negative publicity – and one can imagine some Facebook users attempting some very dubious transactions.

Technical and human intelligence

Under the hood, it appears Facebook is not using cutting-edge AI. While its digital assistant’s interface is stored and run from users’ phones, the processing occurs on Facebook’s servers in the cloud where computing power and data can be distributed. It uses technology from wit.ai, which is understood to use conditional random fields, a popular statistical technique dating from the 2000s, and maximum entropy classifiers, based on information theory. These pick up on the structure of the data, and use this to make predictions. These may not be cutting edge, but they are well established and understood. Not only that, but they can use prior knowledge, and one of M’s aims is to improve and to get better through training.

There’s a huge amount of contextual information about the user’s likes and preferences within Facebook’s enormous datasets, and this could help M’s algorithms provide answers. It could also be used to help constrain queries – things to exclude – particularly if both the purchaser and the recipient are Facebook users. But it will take leading edge AI techniques like sentic technologies, which attempt to extract mood, emotion, intention and meaning from text, in order to mine the full value of the text and image datasets generated by Facebook users.

M’s natural language processing picks out a message’s intent. But it has a lot to learn. Facebook

David Marcus, vice president of messaging products at Facebook and in charge of M, has said that without explicit consent M won’t embark on such data-mining. In fact there is a limited range of possible services and purchases that the software can perform automatically, while trickier tasks are carried out by the human element behind the scenes - customer service staff working for Facebook. Humans are needed to be able to cover the gaps in the AI’s ability to understand natural language, understanding what users are after, able to sign off purchases to ensure they’re reasonable, and legal.

While the idea is that M learns the right behaviours by associating the user’s intent with the solutions provided by human staff, for this to scale to even a fraction of Facebook Messenger’s 700,000 users, the AI will have to be good enough to relieve the human staff of their role. And that may take a while. Of course, M is being rolled out area by area – currently only San Francisco, of course – so perhaps the firm is just dipping a toe in the water to start with.

So while M may be the personal assistant of the future, at the moment it’s a curious mix of machine learning, automation, and human comprehension. But powered by the tutoring of actual humans and human-created data, in time it could still become more adept than the competition.

Organic 'computers' made of DNA could process data inside our bodies

Will we see DNA in the mainframe? PublicDomainPictures

We invariably imagine electronic devices to be made from silicon chips, with which computers store and process information as binary digits (zeros and ones) represented by tiny electrical charges. But it need not be this way: among the alternatives to silicon are organic mediums such as DNA.

DNA computing was first demonstrated in 1994 by Leonard Adleman who encoded and solved the travelling salesman problem, a maths problem to find the most efficient route for a salesman to take between hypothetical cities, entirely in DNA.

Deoxyribonucleaic acid, DNA, can store vast amounts of information encoded as sequences of the molecules, known as nucleotides, cytosine (C), guanine (G), adenine (A), or thymine (T). The complexity and enormous variance of different species’ genetic codes demonstrates how much information can be stored within DNA encoded using CGAT, and this capacity can be put to use in computing. DNA molecules can be used to process information, using a bonding process between DNA pairs known as hybridisation. This takes single strands of DNA as input and produces subsequent strands of DNA through transformation as output.

Since Adleman’s experiment, many DNA-based “circuits” have been proposed that implement computational methods such as Boolean logic, arithmetical formulas, and neural network computation. Called molecular programming, this approach applies concepts and designs customary to computing to nano-scale approaches appropriate for working with DNA.

It’s circuitry, but not as we know it. Caltech/Lulu Qian, CC BY

In this sense “programming” is really biochemistry. The “programs” created are in fact methods of selecting molecules that interact in a way that achieves a specific result through the process of DNA self-assembly, where disordered collections of molecules will spontaneously interact to form the desired arrangement of strands of DNA.

DNA ‘robots’

DNA can also be used to control motion, allowing for DNA-based nano-mechanical devices. This was first achieved by Bernard Yurke and colleagues in 2000, who created from DNA strands a pair of tweezers that opened and pinched. Later experiments such as by Shelley Wickham and colleagues in 2011 and at Andrew Turberfield’s lab at Oxford demonstrated nano-molecular walking machines made entirely from DNA that could traverse set routes.

One possible application is that such a nano-robot DNA walker could progress along tracks making decisions and signal when reaching the end of the track, indicating computation has finished. Just as electronic circuits are printed onto circuit boards, DNA molecules could be used to print similar tracks arranged into logical decision trees on a DNA tile, with enzymes used to control the decision branching along the tree, causing the walker to take one track or another.

DNA walkers can also carry molecular cargo, and so could be used to deliver drugs inside the body.

Why DNA computing?

DNA molecules’ many appealing features include their size (2nm width), programmability and high storage capacity – much greater than their silicon counterparts. DNA is also versatile, cheap and easy to synthesise, and computing with DNA requires much less energy than electric powered silicon processors.

Its drawback is speed: it currently takes several hours to compute the square root of a four digit number, something that a traditional computer could compute in a hundredth of a second. Another drawback is that DNA circuits are single-use, and need to be recreated to run the same computation again.

Perhaps the greatest advantage of DNA over electronic circuits is that it can interact with its biochemical environment. Computing with molecules involves recognising the presence or absence of certain molecules, and so a natural application of DNA computing is to bring such programmability into the realm of environmental biosensing, or delivering medicines and therapies inside living organisms.

DNA programs have already been put to medical uses, such as diagnosing tuberculosis. Another proposed use is a nano-biological “program” by Ehud Shapiro of the Weizmann Institute of Science in Israel, termed the “doctor in the cell” that targets cancer molecules. Other DNA programs for medical applications target lymphocytes (a type of white blood cell), which are defined by the presence or absence of certain cell markers and so can be naturally detected with true/false Boolean logic. However, more effort is required before we can inject smart drugs directly into living organisms.

Future of DNA computing

Taken broadly, DNA computation has enormous future potential. Its huge storage capacity, low energy cost, ease of manufacturing that exploits the power of self-assembly and its easy affinity with the natural world are an entry to nanoscale computing, possibly through designs that incorporate both molecular and electronic components. Since its inception, the technology has progressed at great speed, delivering point-of-care diagnostics and proof-of-concept smart drugs – those that can make diagnostic decisions about the type of therapy to deliver.

There are many challenges, of course, that need to be addressed so that the technology can move forward from the proof-of-concept to real smart drugs: the reliability of the DNA walkers, the robustness of DNA self-assembly, and improving drug delivery. But a century of traditional computer science research is well placed to contribute to developing DNA computing through new programming languages, abstractions, and formal verification techniques – techniques that have already revolutionised silicon circuit design, and can help launch organic computing down the same path.

The theory of parallel universes is not just maths – it is science that can be tested

Scientists are searching for collisions between different 'universe bubbles' in the cosmic microwave bacground. Geralt/pixabay

The existence of parallel universes may seem like something cooked up by science fiction writers, with little relevance to modern theoretical physics. But the idea that we live in a “multiverse” made up of an infinite number of parallel universes has long been considered a scientific possibility – although it is still a matter of vigorous debate among physicists. The race is now on to find a way to test the theory, including searching the sky for signs of collisions with other universes.

It is important to keep in mind that the multiverse view is not actually a theory, it is rather a consequence of our current understanding of theoretical physics. This distinction is crucial. We have not waved our hands and said: “Let there be a multiverse”. Instead the idea that the universe is perhaps one of infinitely many is derived from current theories like quantum mechanics and string theory.

The many-worlds interpretation

You may have heard the thought experiment of Schrödinger’s cat, a spooky animal who lives in a closed box. The act of opening the box allows us to follow one of the possible future histories of our cat, including one in which it is both dead and alive. The reason this seems so impossible is simply because our human intuition is not familiar with it.

But it is entirely possible according to the strange rules of quantum mechanics. The reason that this can happen is that the space of possibilities in quantum mechanics is huge. Mathematically, a quantum mechanical state is a sum (or superposition) of all possible states. In the case of the Schrödinger’s cat, the cat is the superposition of “dead” and “alive” states.

But how do we interpret this to make any practical sense at all? One popular way is to think of all these possibilities as book-keeping devices so that the only “objectively true” cat state is the one we observe. However, one can just as well choose to accept that all these possibilities are true, and that they exist in different universes of a multiverse.

Miaaaaultiverse Robert Couse-Baker/Flickr, CC BY-SA

The string landscape

String theory is one of our most, if not the most promising avenue to be able to unify quantum mechanics and gravity. This is notoriously hard because gravitational force is so difficult to describe on small scales like those of atoms and subatomic particles – which is the science of quantum mechanics. But string theory, which states that all fundamental particles are made up of one-dimensional strings, can describe all known forces of nature at once: gravity, electromagnetism and the nuclear forces.

However, for string theory to work mathematically, it requires at least ten physical dimensions. Since we can only observe four dimensions: height, width, depth (all spatial) and time (temporal), the extra dimensions of string theory must therefore be hidden somehow if it is to be correct. To be able to use the theory to explain the physical phenomena we see, these extra dimensions have to be “compactified” by being curled up in such a way that they are too small to be seen. Perhaps for each point in our large four dimensions, there exists six extra indistinguishable directions?

A problem, or some would say, a feature, of string theory is that there are many ways of doing this compactification –10⁵⁰⁰ possibilities is one number usually touted about. Each of these compactifications will result in a universe with different physical laws – such as different masses of electrons and different constants of gravity. However there are also vigorous objections to the methodology of compactification, so the issue is not quite settled.

But given this, the obvious question is: which of these landscape of possibilities do we live in? String theory itself does not provide a mechanism to predict that, which makes it useless as we can’t test it. But fortunately, an idea from our study of early universe cosmology has turned this bug into a feature.

The early universe

During the very early universe, before the Big Bang, the universe underwent a period of accelerated expansion called inflation. Inflation was invoked originally to explain why the current observational universe is almost uniform in temperature. However, the theory also predicted a spectrum of temperature fluctuations around this equilibrium which was later confirmed by several spacecraft such as Cosmic Background Explorer, Wilkinson Microwave Anisotropy Probe and the PLANCK spacecraft.

While the exact details of the theory are still being hotly debated, inflation is widely accepted by physicists. However, a consequence of this theory is that there must be other parts of the universe that are still accelerating. However, due to the quantum fluctuations of space-time, some parts of the universe never actually reach the end state of inflation. This means that the universe is, at least according to our current understanding, eternally inflating. Some parts can therefore end up becoming other universes, which could become other universes etc. This mechanism generates a infinite number of universes.

By combining this scenario with string theory, there is a possibility that each of these universes possesses a different compactification of the extra dimensions and hence has different physical laws.

The cosmic microwave background. Scoured for gravitational waves and signs of collisions with other universes. NASA / WMAP Science Team/wikimedia

Testing the theory

The universes predicted by string theory and inflation live in the same physical space (unlike the many universes of quantum mechanics which live in a mathematical space), they can overlap or collide. Indeed, they inevitably must collide, leaving possible signatures in the cosmic sky which we can try to search for.

The exact details of the signatures depends intimately on the models – ranging from cold or hot spots in the cosmic microwave background to anomalous voids in the distribution of galaxies. Nevertheless, since collisions with other universes must occur in a particular direction, a general expectation is that any signatures will break the uniformity of our observable universe.

These signatures are actively being pursued by scientists. Some are looking for it directly through imprints in the cosmic microwave background, the afterglow of the Big Bang. However, no such signatures are yet to be seen. Others are looking for indirect support such as gravitational waves, which are ripples in space-time as massive objects pass through. Such waves could directly prove the existence of inflation, which ultimately strengthens the support for the multiverse theory.

Whether we will ever be able to prove their existence is hard to predict. But given the massive implications of such a finding it should definitely be worth the search.

Shift from electronics to spintronics opens up possibilities of faster data

levoodoo, CC BY-NC

Electronics is based on measuring the tiny electrical charge of electrons passing through electronic circuits. An alternative approach under development is spintronics, which instead relies not on electrons’ charge, but on another of their fundamental quantum-mechanical properties: spin.

Spin can be visualised as the Earth turning on its own axis while rotating around the sun. In the same way, an electron spins on its own axis while rotating around an atom’s nucleus. Spin is either “up” or “down”. In the same way traditional electronics uses charge to represent information as zeros and ones, the two spin states can be used to represent the same binary data in spintronics.

Spin can be measured because it generates tiny magnetic fields. Ferrous metals such as iron become magnetic, for example, when enough particles have their spin set in the same direction, generating a magnetic field of the same polarity as the spin.

Spintronics has several advantages over conventional electronics. Electronics require specialised semiconductor materials in order to control the flow of charge through the transistors. But spin can be measured very simply in common metals such as copper or aluminium. Less energy is needed to change spin than to generate a current to maintain electron charges in a device, so spintronics devices use less power.

Spin states can be set quickly, which makes transferring data quicker. And because electron spin is not energy-dependent, spin is non-volatile – information sent using spin remains fixed even after loss of power.

Upgrading hard disks using spin

The first application of spintronics to computers saw Professors Albert Fert and Peter Grünberg awarded the 2007 Nobel Prize in Physics for their discovery of giant magnetoresistance (GMR). They realised it was possible to use electron spin to increase the rate at which information could be read from a hard disk drive and developed ground-breaking technology to harness this feature.

A hard drive, showing circular platters and read/write head mounted at the tip of the arm. drive by mike mols/shutterstock.com

A hard disk drive stores data as ones and zeros encoded magnetically on rotating disk platters within the drive. The magnetic field is generated when electrons flow through wire coils mounted in the drive write heads which move across the face of the platters, changing the alignment of the magneto-sensitive particles on the platter surface. Reversing the electron flow reverses the field; the two directions represent one and zero. To read from the disk the process works in reverse.

A hard disk drive read/write head. amagill, CC BY

A GMR drive head consists of two ferromagnetic layers, one with a fixed magnetic field direction and the other free to align with the magnetic field encoded on the disk, with a non-magnetic layer sandwiched in between.

When an electron passes through a magnetic field its spin state may change, known as scattering. Where electrons have random, scattered spin states this creates greater resistance to electric current. By aligning electrons’ spin state to that of the magnetic field in the layers of the drive head, GMR technology dramatically reduces resistance, speeding up data transfer. First introduced by IBM in 1997, GMR technology has led to faster and higher-density drives than was previously possible.

Putting a fresh spin on memory

Spintronics researchers have since been working on introducing the same technology to computer memory, aiming to replace electric current-based dynamic random access memory (DRAM) with magnetic RAM (MRAM). The first commercial product by Everspin has been used in Airbus aircraft and BMW motorbikes due to its reliability under heat stress or cosmic-ray exposure – something that affects aircraft cruising at high altitudes.

MRAM exploits the same spin-based magnetic field approach, but uses a magnetoresistance cell to store data rather than a spinning disk platter as in a hard drive. While it is not as fast as DRAM, magnetic cells are able to maintain their stored spin orientations, and so the data they represent, without power. MRAM is likely to replace commonly used flash memory such as SD cards and compact flash first, as it is faster and doesn’t suffer from flash memory’s limited lifespan.

Other manufacturers such as Intel, Qualcomm, Toshiba and Samsung are developing MRAM to use as processor cache memory, where by virtue of their smaller size MRAM chips of greater capacity can be incorporated into smaller packages that will be faster, and use up to 80% less power than current cache memory.

As electronics approaches the limits of silicon, spintronic components will play an important role in ensuring we enjoy steady performance gains, and faster, higher-capacity storage at lower power and cost.

Get used to it: quantum computing will bring immense processing possibilities

D-Wave, CC BY

The one thing everyone knows about quantum mechanics is its legendary weirdness, in which the basic tenets of the world it describes seem alien to the world we live in. Superposition, where things can be in two states simultaneously, a switch both on and off, a cat both dead and alive. Or entanglement, what Einstein called “spooky action-at-distance” in which objects are invisibly linked, even when separated by huge distances.

But weird or not, quantum theory is approaching a century old and has found many applications in daily life. As John von Neumann once said: “You don’t understand quantum mechanics, you just get used to it.” Much of electronics is based on quantum physics, and the application of quantum theory to computing could open up huge possibilities for the complex calculations and data processing we see today.

Imagine a computer processor able to harness super-position, to calculate the result of an arbitrarily large number of permutations of a complex problem simultaneously. Imagine how entanglement could be used to allow systems on different sides of the world to be linked and their efforts combined, despite their physical separation. Quantum computing has immense potential, making light work of some of the most difficult tasks, such as simulating the body’s response to drugs, predicting weather patterns, or analysing big datasets.

Replica of the first ever transistor, manufactured at Bell Labs in 1947. Lucent Technologies

Such processing possibilities are needed. The first transistors could only just be held in the hand, while today they measure just 14 nm – 500 times smaller than a red blood cell. This relentless shrinking, predicted by Intel founder Gordon Moore as Moore’s law, has held true for 50 years, but cannot hold indefinitely. Silicon can only be shrunk so far, and if we are to continue benefiting from the performance gains we have become used to, we need a different approach.

Quantum fabrication

Advances in semiconductor fabrication have made it possible to mass-produce quantum-scale semiconductors – electronic circuits that exhibit quantum effects such as super-position and entanglement.

Quantum circuitry. Paul Koenraad/TU Eindhoven, Author provided

The image, captured at the atomic scale, shows a cross-section through one potential candidate for the building blocks of a quantum computer, a semiconductor nano-ring. Electrons trapped in these rings exhibit the strange properties of quantum mechanics, and semiconductor fabrication processes are poised to integrate these elements required to build a quantum computer. While we may be able to construct a quantum computer using structures like these, there are still major challenges involved.

In a classical computer processor a huge number of transistors interact conditionally and predictably with one another. But quantum behaviour is highly fragile; for example, under quantum physics even measuring the state of the system such as checking whether the switch is on or off, actually changes what is being observed. Conducting an orchestra of quantum systems to produce useful output that couldn’t easily by handled by a classical computer is extremely difficult.

But there have been huge investments: the UK government announced £270m funding for quantum technologies in 2014 for example, and the likes of Google, NASA and Lockheed Martin are also working in the field. It’s difficult to predict the pace of progress, but a useful quantum computer could be ten years away.

Building quantum computers. Michael Thompson, Lancaster Quantum Technology Centre, Author provided

The basic element of quantum computing is known as a qubit, the quantum equivalent to the bits used in traditional computers. To date, scientists have harnessed quantum systems to represent qubits in many different ways, ranging from defects in diamonds, to semiconductor nano-structures or tiny superconducting circuits. Each of these has is own advantages and disadvantages, but none yet has met all the requirements for a quantum computer, known as the DiVincenzo Criteria.

The most impressive progress has come from D-Wave Systems, a firm that has managed to pack hundreds of qubits on to a small chip similar in appearance to a traditional processor.

Quantum secrets

The benefits of harnessing quantum technologies aren’t limited to computing, however. Whether or not quantum computing will extend or augment digital computing, the same quantum effects can be harnessed for other means. The most mature example is quantum communications.

Quantum physics has been proposed as a means to prevent forgery of valuable objects, such as a banknote or diamond, as illustrated in the image below. Here, the unusual negative rules embedded within quantum physics prove useful; perfect copies of unknown states cannot be made and measurements change the systems they are measuring. These two limitations are combined in this quantum anti-counterfeiting scheme, making it impossible to copy the identity of the object they are stored in.

Adding a quantum secret to a standard barcode prevents tampering or forgery of valuable goods. Robert Young, Author provided

The concept of quantum money is, unfortunately, highly impractical, but the same idea has been successfully extended to communications. The idea is straightforward: the act of measuring quantum super-position states alters what you try to measure, so it’s possible to detect the presence of an eavesdropper making such measurements. With the correct protocol, such as BB84, it is possible to communicate privately, with that privacy guaranteed by fundamental laws of physics.

Quantum communication systems are commercially available today from firms such as Toshiba and ID Quantique. While the implementation is clunky and expensive now it will become more streamlined and miniaturised, just as transistors have miniaturised over the last 60 years.

Improvements to nanoscale fabrication techniques will greatly accelerate the development of quantum-based technologies. And while useful quantum computing still appears to be some way off, it’s future is very exciting indeed.