In the past century, computers have  gone from room-sized calculators, to smartphones that let me watch a video of Angry Bird speedruns of all  time while I’m at the airport.

But as much as computers have  changed, they’ve also stayed the same. At least in terms of how they do their computing.

But in the last few years, scientists  have been racing to build a totally new kind of computer, called a quantum  computer, that may spark a revolution. And if you’ve seen any news about this research, you may have also seen that quantum  computers have a distinct look. They look like a complicated stack of gold  discs and wires hanging from a ceiling.

Sort of like a steampunk chandelier. And you may have wondered: Why do  quantum computers look like that? And we can answer that for you.

But the thing is, while they really do  look like that, they also … kinda … don’t. [♪ INTRO] Any kind of regular computer,  be it the one in your phone or the one in your smart toaster, runs  on exactly the same principles. At some level, it’s processing  information in the form of bits: abstract, mathematical objects that  are always in one of two states. For convenience, those states are  usually labelled “zero” and “one”.

Meanwhile, every computer contains  physical objects that can also be in one of two states, usually in the form of electric  current either flowing or not flowing. In other words, these objects  physically represent those bits. To do the actual computing, the  rest of the computer processes bits by switching them between the zero  state and the one state as needed.

These days, the most advanced  microchips can have hundreds of billions of these switches on them, meaning you can fit incredible computing power into a tiny space. Since the 1940s, we’ve built countless different kinds of computers with this approach. But in recent years,  scientists have been trying to build computers using something totally different.

As their name suggests, a quantum  computer runs on the principles of quantum mechanics: the weird,  counter-intuitive theory that describes the universe  on the smallest scales. And the main difference with these computers is that instead of using bits, they use qubits. A qubit is like a regular bit, except that  instead of being in either the zero or the one state, it can somehow be in some  combination of the two states at once.

This is the strange quantum  phenomenon known as superposition, and there’s nothing like it in the everyday world. It’s as if a coin could  somehow be in both the ‘heads’ and ‘tails’ states at once. It’s just kinda odd.

We’ve known about the kinds of quantum systems that produce qubits for a century now. And in recent decades, people  have learned that qubits can process information very  differently from regular bits. This is due to that superposition,  as well as other quantum phenomena like entanglement and interference  … which are super complicated in their own right, so we’ll  set them aside for this episode.

What it all comes down to is this: the extra  possible states that a qubit can be in … those different zero-and-one states …  can interact in totally different ways with each other than your  regular zero and one states. Researchers can exploit these  effects to do clever tricks. And sometimes, it means a  quantum computer winds up performing a specific computing task  much faster than a classical computer.

So recently, a sort of arms race  has begun between universities, big tech companies, small startups,  and even militaries to build working quantum computers, and  discover new uses for them. It’s a big industry with a lot of  hype: In June 2025, one quantum company bought out a rival for just  over a billion dollars. And that’s downright wild, given that,  basically, not a single quantum computer has made it out of the prototype stage.

And none are actually being used right now to do computations that a classical computer can’t do. The biggest challenge in  making quantum computers is something called decoherence:  a physical process that makes quantum objects lose their quantumness. When a qubit decoheres, it goes from  being in some kind of superposition state to a regular old zero-or-one bit.

And once that’s happened, it’s  almost impossible to change it back. And what causes decoherence? Well, a better question would be  what doesn’t cause decoherence.

Quantum systems are notoriously sensitive. Basically anything causes them to decohere … getting hit by a single particle of  radiation, air molecules bumping into them, or their surroundings getting too hot… And by ‘too hot’ I mean, they can decohere if they get hotter than the  deepest depths of outer space. So the quantum objects that  people actually study in labs tend to be on the scale of individual atoms.

Or at most, small molecules. Any bigger, and it’s far more likely to decohere. This is also why we don’t  see weird quantum effects like heads-and-tails coins in everyday life.

At human scales, decoherence washes away all quantum effects in ludicrously short times. Making a quantum computer is an  endless fight against decoherence. It takes a lot of work to both  get enough qubits together, and keep them stable enough  to be used in computation.

Remember how modern microchips  can have billions of bits on them? Well, right now, the most  advanced quantum computers have about a thousand qubits. And they’re extremely unreliable.

Now, this sensitivity can vary, depending  on what you’re using to make your qubits. This is a new field, so researchers are  trying to figure out what works best. But the current leaders in  the quantum computing race use a superconducting qubit.

These are micrometer-scale chunks of silicon, made using techniques borrowed from  the regular microchip industry. And a micrometer is huge in  the world of quantum mechanics. So given their relative size,  these superconducting qubits are a lot easier to manipulate in a lab  than qubits made of individual atoms.

But as you might suspect, there’s a trade-off. They’re also way more sensitive to  things that can make them decohere. So to keep them functioning properly,  they need to be kept unimaginably cold.

Which finally brings us back  to our original question: why do quantum computers look like that? It’s not wrong to say that some quantum computers do in fact look like funky chandeliers. But there’s some nuance, here.

Because that huge chandelier-like device isn’t actually the quantum computer itself. After all, if a qubit is a  microscopic piece of silicon, it only needs a microchip-sized  platform to stand on. So the actual quantum computer is just one chip, small enough to fit in the palm of your hand.

Meanwhile, the giant chandelier  is the cooling device. It’s called a dilution refrigerator, and it can make some of the  coldest places in the universe. The technology behind it is  actually more than 60 years old.

And while there’s a bunch of research being done to improve these things, many  labs use off-the-shelf versions. So, yes, if you have a spare  million dollars lying around, you too can impress your friends with  your very own dilution refrigerator. But despite being off-the-shelf, it uses  some pretty neat physics to do its job.

In order to pump heat out of the working area, a dilution refrigerator cycles between  mixing and unmixing two different isotopes of liquid helium, both of which are only about 4 degrees warmer than absolute zero. The first isotope is your  run-of-the-mill helium-4, with two protons and two neutrons in its nucleus. The second is helium-3,  which only has one neutron.

The actual cooling process relies on  one of many quantum mechanical quirks that kick in at such low temperatures. Basically, when you mix  these two liquids together, they wind up absorbing a lot of  heat from the stuff around them. Including the stuff that can cause a  quantum computer to stop working correctly.

So by repeatedly mixing and  unmixing these helium liquids, you can continue to pull in heat,  and keep your computer cold. As for that tiered, chandelier-y shape:  with each layer, you get more cooling. The top layer is pretty close to room temperature, but the bottom layer is just a fraction  of a degree above absolute zero.

So that’s where you put the quantum chip. Meanwhile, all the wires you see sticking  out of things are there to control input and output signals to and from the  qubits, to actually do computations. Now technically, not every kind of quantum  computer needs a dilution refrigerator.

Some, for instance, are based  on photons…particles of light… and don’t need this sort of cooling. So not every quantum computer actually  does look like an expensive chandelier. And scientists may eventually  develop the technology in such a way that these huge coolers  aren’t really necessary, anymore.

But right now, even with all this cooling, qubits are still super prone to decoherence. If you want even more jargon from  the quantum computing community, many people say that we’re now in the noisy, intermediate scale quantum era, or the NISQ era. That means we can get enough qubits  together to do some things we want… that’s the ‘intermediate scale’ part of the name… but we haven’t fully tamed  the decoherence effects.

That’s the ‘noisy ’part. The era beyond that is where quantum computers could have many more uses outside research labs. It may only come when we have  tens of thousands of qubits… roughly ten times more than  we can currently manage… and relatively decoherence free.

However, some experts say  that that’s an underestimate, and only once we hit the million-qubit era can we actually do interesting stuff outside the lab. Like maybe designing new  medicines by searching huge arrays of molecule combinations super  quickly, or making more efficient, advanced versions of “artificial intelligence”. But as exciting as that  potential future may sound, before we wrap up, I should  offer a word of caution.

There’s a lot of hype around  quantum computers these days, with countless startups  trying to sell their products as the vanguard of a quantum revolution. But right now, we simply don’t know for sure how useful quantum computers will be in the future. For one thing, there’s a lot we still don’t understand about regular, classical computing.

In recent years, there have been  several instances where companies claimed to have solved a  problem using a quantum computer that would have taken a classical  computer thousands of years. But every time, other researchers  quickly showed that the problem could be solved classically  in a much shorter time. So it actually could be done on a  regular supercomputer…or even a laptop.

And that’s kind of a cool result. It proves that research into quantum computers can help us learn how to make  classical computers work better. Also, designing quantum  algorithms is notoriously hard.

And if you want to actually  use or sell that algorithm, you also need to prove that a classical  computer couldn’t do it better or cheaper. Very few cases have been proven to have  this so-called ‘quantum advantage’. So while it may be true that  quantum advantage will revolutionize whole areas of science, commerce,  the economy, security, and more… It may also be the case that quantum  computers will be worth the effort for only a handful of individual,  highly specific use cases in some industries and scientific research fields.

So be very sceptical of anyone  who tells you quantum computers will change everything, or that  they’re right around the corner. But despite the caution,  there’s still cause for lots of excitement in the near  future of quantum computing, with countless researchers across the world exploring all the avenues  in this emerging technology. And one day, maybe you could  hold a quantum computer in the palm of your hand…without it also being  stuck to the bottom of a giant refrigerator.

Although you probably won’t be using it  to speedrun Angry Birds at the airport. [♪ OUTRO]