[00:08.820 --> 00:10.400]  Thanks everyone for coming.
[00:10.480 --> 00:12.760]  So, my project is called Open Quantum.
[00:12.920 --> 00:17.880]  I think when we talk about quantum science, especially in the tech world, there's a really big focus on computing.
[00:17.880 --> 00:18.860]  That's what's accessible.
[00:18.860 --> 00:23.320]  You know, IBM has its API, other people have APIs, this is what you can play with.
[00:23.320 --> 00:29.860]  But really, the hardware that enables these applications still remains very arcane.
[00:30.520 --> 00:37.900]  I had the good fortune to work in a hardware quantum computing lab a couple of years ago, and my experience is that a lot of this stuff is really accessible.
[00:38.620 --> 00:48.720]  So today I just want to take you through what that might look like, how you can get started, and really my attempt to make it less intimidating and more accessible to anybody who wants to give it a try.
[00:49.880 --> 00:56.360]  So, here are three of probably the most futuristic projects that people are building in today's society.
[00:56.360 --> 01:03.340]  That's a quantum processor, this is NASA's deep space satellite, and this is a mind-reading helmet.
[01:04.120 --> 01:05.840]  What do all these things have in common?
[01:05.840 --> 01:11.020]  They all rely on the control of atoms, our understanding of atomic physics.
[01:11.340 --> 01:13.680]  That might not be something that you're very familiar with.
[01:13.680 --> 01:15.620]  So, what is atomic physics?
[01:15.680 --> 01:18.440]  You might think, well, atoms have a nucleus.
[01:18.620 --> 01:20.320]  It's something like nuclear physics.
[01:20.320 --> 01:21.800]  It's not nuclear physics.
[01:21.920 --> 01:23.780]  There's no nuclear processes going on.
[01:23.780 --> 01:27.740]  What we care about in atomic physics is the study of individual atomic systems.
[01:28.540 --> 01:33.020]  And really trying to understand atoms is what led to the birth of quantum mechanics.
[01:33.400 --> 01:38.940]  Trying to understand, for instance, the spectra of atoms, how they absorb energy, how they interact with light.
[01:39.240 --> 01:43.000]  And so, in 50 years, this was a purely theoretical field.
[01:43.800 --> 01:47.520]  Atoms move at about a thousand miles per hour at room temperature and pressure.
[01:48.620 --> 01:50.620]  So, pretty hard to look at.
[01:51.440 --> 01:59.140]  But in 1987, at Bell Labs, they invented this thing called the magneto-optical trap, which looks something like this.
[01:59.740 --> 02:07.560]  And what this allows you to do is to cool down and magnetically trap atoms at a collective temperature of about one microKelvin.
[02:07.560 --> 02:17.460]  So, these things are now moving so slow that you can actually isolate them and study them individually, which is the conditions that people think about when they talk about them theoretically.
[02:18.400 --> 02:20.420]  So, now what is quantum engineering?
[02:20.620 --> 02:25.700]  Quantum engineering I refer to as the ability to individual quanta of the universe.
[02:25.780 --> 02:30.140]  This is a very recent field, because quantum mechanics discovered last century.
[02:30.140 --> 02:32.320]  We've come a long way in a short period of time.
[02:32.480 --> 02:33.960]  What is a quanta of the universe?
[02:33.960 --> 02:38.200]  It's, you know, when you get down to nanoscopic levels, things become quantized.
[02:38.200 --> 02:39.500]  The states are discrete.
[02:39.960 --> 02:44.960]  We experience life in kind of continuity, like I can move my arm continuously through the space.
[02:45.020 --> 02:47.760]  But at the nanoscopic level, things become discrete.
[02:47.760 --> 02:48.980]  For instance, electrons spin.
[02:48.980 --> 02:54.220]  So, electrons don't actually spin, but, you know, we think about it as intuitive.
[02:54.220 --> 02:55.520]  They've got two spin states.
[02:55.520 --> 02:59.260]  Other things like flux qubits and superconducting computers have photon states.
[02:59.260 --> 03:01.540]  Photons also have two spins.
[03:01.540 --> 03:07.580]  But in atoms, we've got quantized energy states, which means that there's a set of energy levels.
[03:07.580 --> 03:11.620]  If you remember maybe high school chemistry, you've got like your passion ball and reliance series.
[03:11.620 --> 03:16.400]  These are the energy levels where atoms are absorbing and emitting photons.
[03:16.760 --> 03:17.920]  And there's a discrete.
[03:17.920 --> 03:18.640]  They're discrete.
[03:18.640 --> 03:19.400]  It's not continuous.
[03:19.400 --> 03:21.060]  They can only absorb at certain energies.
[03:21.740 --> 03:24.840]  And so, yeah, they move between these states by absorbing and then re-emitting light.
[03:25.060 --> 03:27.440]  So, if you can control light, you can control atoms.
[03:28.040 --> 03:29.140]  How does this work?
[03:30.220 --> 03:32.460]  How do we precisely control light matter interactions?
[03:32.460 --> 03:35.040]  So, this top picture is a neon sign.
[03:35.480 --> 03:37.540]  Neon is an imprecise way of controlling light.
[03:37.540 --> 03:43.840]  We put high voltage electrodes across a tube filled with gas, and you ionize it, and it emits light.
[03:43.840 --> 03:45.460]  You know, this is kind of brute force.
[03:46.760 --> 03:49.100]  Atoms absorb photons at very specific energies.
[03:49.800 --> 03:56.520]  You can only be in a very small band called the bandwidth around a resonant energy to get it to absorb.
[03:56.880 --> 04:00.440]  And when you do that, the photons are then converted into the kinetic energy of electrons.
[04:00.440 --> 04:07.460]  They move up in orbitals and then eventually decay back down to the ground state.
[04:08.040 --> 04:11.960]  And so, the way that you can do this on command is with something called a frequency-stable laser.
[04:12.220 --> 04:18.740]  And because the energy of a photon is related to its frequency by constant, these are essentially the same thing.
[04:19.860 --> 04:27.580]  And so, if you can tune a laser to exactly the resonant energy of an atomic transition, now you can control all these atoms as you wish.
[04:28.840 --> 04:31.360]  But, you know, still, why do we care?
[04:31.360 --> 04:34.740]  Is this useful for anything beyond just understanding our world?
[04:35.940 --> 04:38.780]  So, here's kind of one motivating phenomena.
[04:38.800 --> 04:39.920]  Radiation pressure.
[04:40.060 --> 04:41.160]  So, light has momentum.
[04:41.160 --> 04:42.340]  You can push things around with light.
[04:42.340 --> 04:46.740]  So, maybe from your favorite science fiction novel, it's how a solar sail works.
[04:46.740 --> 04:54.760]  You put, you know, this big solar sail out in space and ambient light from solar emission will push it along.
[04:55.340 --> 04:58.840]  If you can precisely control light, you can precisely control atomic motion.
[04:58.840 --> 05:03.920]  So, this enables, for instance, a really interesting phenomena called optical tweezers.
[05:04.060 --> 05:08.240]  In this bottom picture, I don't know if you can see it, but there's a little particle being levitated.
[05:08.240 --> 05:11.300]  For whatever reason, Sharpie is actually really easy to do this with.
[05:11.300 --> 05:14.640]  If you have a Sharpie, a lens, and a laser pointer at home, you can do this.
[05:16.300 --> 05:23.000]  And what's happening there is the laser is transferring momentum to the particle in such a way that it's counteracting the gravitational force.
[05:23.200 --> 05:32.180]  So, this is a really DIYable experiment where you can verify that, A, light has momentum, which is kind of unintuitive because in school you're taught that momentum is mass times velocity.
[05:32.180 --> 05:33.080]  Photons have no mass.
[05:33.080 --> 05:34.160]  So, how does that work?
[05:35.460 --> 05:36.980]  I'm not going to get into that right now.
[05:36.980 --> 05:42.540]  But, this won the Nobel Prize in 2018.
[05:42.540 --> 05:45.740]  So, fairly recent, you know, all these things are really new and exciting.
[05:45.920 --> 05:50.400]  Ways that we can improve our control of the universe, which I think is fucking sick.
[05:52.140 --> 05:55.180]  So, another phenomenon that builds off of this is called Doppler cooling.
[05:55.180 --> 05:56.800]  So, you may have heard of the Doppler effect.
[05:57.240 --> 06:01.580]  You know, every time an ambulance drives by you, it gets higher pitch when it's moving towards you.
[06:01.580 --> 06:03.300]  And then as it goes away, it gets lower pitch.
[06:03.320 --> 06:05.720]  That's because the wave front is getting compressed and then stretched.
[06:05.860 --> 06:08.220]  This applies to any wave and light is no exception.
[06:09.080 --> 06:16.640]  So, atoms moving towards a light source see those photons at a higher energy because the wave front is collapsed.
[06:16.640 --> 06:18.920]  The frequency increases and the energy increases.
[06:19.260 --> 06:26.180]  So, if you have a photon that is tuned below the transition energy of the atom, it'll see it as higher energy.
[06:26.180 --> 06:28.380]  So, it's blue shifted if you're familiar with astronomy.
[06:28.780 --> 06:34.040]  And what that means is that the only atoms that will see your light are the ones that are moving towards it.
[06:34.040 --> 06:40.160]  So, you can do this to kill off the kinetic energy of atoms that are moving towards you by just shooting light at it.
[06:40.160 --> 06:46.180]  Again, you have to very precisely tune this energy, slightly below the transition frequency, but you can cool them down in this one direction.
[06:46.180 --> 06:50.480]  I mean, temperature, like cooling, is a measure of average kinetic energy of an ensemble.
[06:50.860 --> 06:54.580]  But really what you're doing is taking off kinetic energy piece by piece.
[06:55.940 --> 07:02.600]  So, here, you know, how do we start turning these interesting physical phenomena into engineering devices?
[07:03.080 --> 07:05.500]  The first one I want to talk about is atomic clocks.
[07:05.760 --> 07:09.820]  So, since about 1968, the second has been defined by an atomic clock.
[07:10.520 --> 07:12.380]  This is done at NIST in Colorado.
[07:13.080 --> 07:17.140]  The current clock they've built doesn't lose more than a second in 15 billion years.
[07:18.400 --> 07:20.820]  And the way this works is all based on cold atoms.
[07:20.840 --> 07:27.980]  You pull these atoms down, which I'll talk about in a bit, and then you throw them upwards in this vacuum tube, and you excite them with microwaves.
[07:28.160 --> 07:33.320]  And you scan through these frequencies, and the frequency that excites them the most is the resonant frequency.
[07:33.320 --> 07:39.780]  And, you know, frequency is just something that you can change with time, because it's, you know, inverse time.
[07:39.780 --> 07:43.020]  And that's the frequency that determines what the second is.
[07:44.860 --> 07:48.120]  Another example is this mind control helmet, which I really love.
[07:48.220 --> 07:53.540]  So, external magnetic fields are something that shifts the transition energy for an atom.
[07:53.840 --> 08:05.220]  And so, if you put your precisely tuned laser through a cell of atoms, as the magnetic field increases, pure atoms will absorb photons at the resonant frequency, because the resonant frequency is changing.
[08:05.980 --> 08:12.280]  And so, when you measure the fluorescence of that glass cell, you can, you know, do some math to estimate what the magnetic field is.
[08:12.280 --> 08:15.580]  And so, what these people have done, it's a startup called Kernel, I'm not affiliated with.
[08:17.160 --> 08:25.720]  They put it right on your brain, and when your neurons fire, I'm also not a neuroscientist, generates a magnetic field, and you can read that out just sitting on your head.
[08:25.720 --> 08:30.080]  So, this is like, they market it as a high-resolution MRI.
[08:30.080 --> 08:32.520]  I don't know if it works, but the idea is really sitting.
[08:32.920 --> 08:34.080]  Also looks sitting.
[08:35.020 --> 08:38.460]  And then the third thing, which probably most of you are here for, are quantum computers.
[08:38.620 --> 08:45.760]  So, in this apparatus, you've got atoms that are cooled and then individually levitated with this optical tweezer phenomenon I described.
[08:46.020 --> 08:51.000]  Then you've got another laser system that can excite those trapped atoms into a higher energy level.
[08:51.120 --> 08:57.780]  And really what we do, just mathematically, is you encode the ground state as a logical zero, and then you encode that excited state as a one.
[08:59.140 --> 09:08.140]  And atoms can be, because these are quantized quantum states, you can put them into superposition by irradiation with this control laser for specific durations of time.
[09:08.280 --> 09:11.780]  And you can also do things like entanglement through more sophisticated techniques.
[09:11.800 --> 09:18.260]  So, in this way, you can use an array of neutral atoms to precisely engineer and control quantum gates.
[09:18.260 --> 09:19.700]  You can run a quantum algorithm.
[09:19.800 --> 09:22.060]  There are five startups in the world trying to do this.
[09:22.860 --> 09:26.500]  I'm forgetting about as far as every other quantum startup, but they're trying.
[09:28.060 --> 09:36.260]  And so, my project was, you know, having worked on one of these neutral atom quantum computers, what is the bare minimum required to get started with a neutral atom experiment?
[09:37.080 --> 09:40.480]  Because my job at this lab was in fabrication.
[09:40.480 --> 09:44.840]  And as I was fabricating, you know, I came in as an engineer with not much knowledge of physics.
[09:45.400 --> 09:49.300]  And the stuff they were having me build just didn't seem that hard to comprehend.
[09:50.940 --> 09:55.640]  And my project on the side was, well, can I make this 3D principle?
[09:55.640 --> 09:56.600]  Can I make this accessible?
[09:56.600 --> 09:57.780]  Can I make this cheap?
[09:58.080 --> 10:02.520]  Because I think quantum mechanics is something that is traditionally regarded as very arcane.
[10:02.520 --> 10:07.080]  You know, it's like, you use it to describe something that is hard to understand.
[10:07.300 --> 10:11.220]  And that's because it's this really abstract theoretical field, except now it's not.
[10:11.220 --> 10:13.900]  Because now we have the ability to control quantum phenomena.
[10:14.260 --> 10:21.000]  And when you've got motivating examples, when you've got something you can build towards, at least for me, I think it becomes so much easier to learn something.
[10:21.160 --> 10:22.340]  You know, you land with a goal.
[10:22.340 --> 10:24.600]  You land with something that you want to put into the world.
[10:25.360 --> 10:27.160]  And it's really exciting because it's so recent.
[10:27.160 --> 10:30.200]  There are all these different experiments, different devices that you can build.
[10:30.300 --> 10:32.000]  You know, maybe you come up with a crazy idea.
[10:32.280 --> 10:33.360]  No one's ever done it.
[10:33.900 --> 10:35.160]  Maybe you want to try it.
[10:35.180 --> 10:38.180]  No, it doesn't cost a million dollars and a PhD to get started.
[10:38.180 --> 10:40.460]  You can start prototyping on this in your apartment.
[10:40.460 --> 10:42.240]  I did this in my apartment for a year and a half.
[10:42.760 --> 10:43.760]  No one got mad at me.
[10:43.760 --> 10:44.860]  Didn't burn anything down.
[10:45.360 --> 10:46.980]  So what does this look like tangibly?
[10:48.060 --> 10:50.880]  The platform I'm building is called the magneto-optical trap.
[10:51.100 --> 10:54.060]  And so this is the foundational platform for most of atomic physics.
[10:54.060 --> 10:58.720]  And that's because, again, atoms at room temperature and pressure are moving at thousands miles per hour.
[10:58.720 --> 11:01.320]  You need to make them slow before you can do anything interesting with them.
[11:01.320 --> 11:03.040]  Otherwise, you just fly away.
[11:03.640 --> 11:08.140]  And so the way this works, very high level, is that you release hot atoms into a vacuum chamber.
[11:08.560 --> 11:11.460]  And then you've got three pairs of counter-propagating beams.
[11:11.520 --> 11:14.600]  Each one of them is doing the stop-liquid process I described.
[11:14.600 --> 11:22.940]  So because you're killing off kinetic energy in each direction, in three orthogonal directions, you push them into this point right in the middle.
[11:24.260 --> 11:35.700]  And then there's some, you know, the magneto part is a magnetic field that ensures that they stay in the central location, which I can describe to people who are very interested later.
[11:35.740 --> 11:37.340]  It's got a little more physics.
[11:37.340 --> 11:46.080]  But really what you can do is you can get about a million to a hundred million atoms that have a collective temperature on the order of a few micropole.
[11:48.040 --> 11:49.800]  Or like the coldest things in the universe.
[11:51.340 --> 11:53.700]  And from there you can do things like, again, quantum computing.
[11:53.820 --> 11:56.640]  If you've heard of Bose-Einstein condensates, it's a really exotic state of matter.
[11:56.640 --> 11:57.900]  This is how you make them.
[11:59.200 --> 12:02.040]  It's such a new and exciting field of physics.
[12:04.080 --> 12:06.520]  And so everything I'm building is open source.
[12:06.720 --> 12:12.900]  You know, I started machining a lot of metal and I realized that if I do that, no one is ever going to replicate it because it's just a pain in the ass.
[12:13.460 --> 12:17.160]  So what I've been designing are all these little 3D printed things.
[12:17.160 --> 12:21.220]  You might think like, well, can, you know, printed plastic really do quantum science?
[12:21.560 --> 12:22.840]  The answer is yes.
[12:22.980 --> 12:24.900]  And I was as surprised as you are.
[12:24.960 --> 12:27.820]  So this is my laser design.
[12:28.160 --> 12:33.360]  I've had probably six or seven at this point, but this one was cheap and simple enough that I could bring.
[12:34.880 --> 12:49.040]  The way it works is that a diffraction grating at a very stable, tunable angle reflects light at a very specific frequency that depends on the angle back into the internal cavity of a laser diode.
[12:49.300 --> 12:53.800]  And what this lets you do is you can now stabilize your wavelength to tens of femtometers.
[12:53.800 --> 12:58.380]  So when we talk about wavelength, you usually say something like 532 nanometers, 780 nanometers.
[12:58.380 --> 13:00.680]  You only go to that third significant digit.
[13:00.680 --> 13:02.940]  Now we're going six more because that's what you need to do.
[13:02.940 --> 13:05.480]  Five or six more to be able to control atoms.
[13:06.680 --> 13:08.760]  Also, the diode has to be temperature stabilized.
[13:08.760 --> 13:12.800]  So there's a little PID loop with a Peltier cooler, a thermoelectric cooler.
[13:13.120 --> 13:17.020]  This prevents shifts in the length of the cavity by thermal expansion.
[13:17.380 --> 13:29.040]  And finally, there's a little piezo actuator in that corner that allows you to control that angle to micro nanoradians because those have such precise spatial control.
[13:29.360 --> 13:32.700]  And then you can PID lock this grating angle to another signal.
[13:33.460 --> 13:36.060]  All these parts are either 3D printed or commercial official.
[13:36.060 --> 13:38.800]  The total bill of materials it's like less than $200.
[13:38.940 --> 13:44.600]  When we were buying these things in the lab I was working in, we were buying them from anywhere from $8,000 to $20,000.
[13:44.760 --> 13:48.040]  So this is really like, you know, low cost.
[13:48.640 --> 13:52.800]  You can even do interesting things with this, like look at Raman scattering.
[13:54.340 --> 13:59.940]  The next part of the experiment is atomic spectroscopy, which is... I'm not going to try to lift that up.
[14:00.480 --> 14:02.720]  So you've got this frequency stable laser.
[14:02.720 --> 14:04.520]  Now you need to... so that's precision.
[14:04.520 --> 14:05.420]  Now we need accuracy.
[14:05.420 --> 14:09.380]  We need to lock it to the frequency of, well, in my case, rubidium atoms.
[14:09.400 --> 14:13.620]  You can choose a different color and pick your favorite atom, like potassium, like strontium.
[14:14.880 --> 14:17.460]  And this requires some precise alignment.
[14:17.460 --> 14:19.480]  You might think, well, how do you do that with a 3D printer?
[14:19.680 --> 14:24.060]  This relies on a notion of optical alignment called kinematic mounts.
[14:26.140 --> 14:33.660]  They're basically a way to restrict the degrees of freedom between two mating surfaces in a way that is highly repeatable.
[14:34.560 --> 14:42.540]  If you buy these from an optical provider like Thorlabs, it'll cost you about $50 to $100 per piece, but you can actually make on a 3D printer.
[14:42.540 --> 14:53.780]  It takes a little bit of dexterity, but now you can control the angle of these beam paths to, again, milliradian, naniradian level.
[14:55.360 --> 14:59.500]  Some wave plate rotators, these are also 3D printed.
[14:59.500 --> 15:01.180]  This allows you to control the polarization.
[15:01.300 --> 15:08.240]  These are things that are important for getting the laser to the right specifications that allow for cooling.
[15:09.060 --> 15:12.060]  As an aside, I'm publishing all the documentation.
[15:12.060 --> 15:17.580]  This will be available in much higher detail, but I want to skip the nitty-gritty physics for this talk.
[15:18.640 --> 15:32.100]  And so the way this works is you have this ability to tune a frequency, and now here you've got a vapor cell, which I didn't bring because mine is from an old Soviet surplus atomic clock, but I'm finding a new provider for all of you guys to use.
[15:34.380 --> 15:48.520]  And you pass the light through the vapor cell, and there's a photo detector there on the left, and you're measuring the absorption of that photo detector through the vapor cell.
[15:48.760 --> 15:55.400]  And basically, as you get closer and closer to the resonant peak of this transition, the reading from that photo detector will drop.
[15:55.400 --> 16:06.620]  And the reason for that is because you're exciting the atoms in the vapor cell now, and as you excite them, they'll absorb that photon and then, through spontaneous emission, re-emit it in kind of a different direction.
[16:06.620 --> 16:08.140]  So it's never hitting the photo detector.
[16:08.260 --> 16:12.380]  So as you get closer to the resonant peak, you just see a smaller and smaller voltage reading.
[16:13.100 --> 16:14.680]  This is very, very simple, straightforward.
[16:14.680 --> 16:19.080]  Anybody who's studied a semester of EEE knows what a phase-locked loop is.
[16:19.080 --> 16:21.220]  You can write this like I did in Arduino.
[16:21.220 --> 16:22.380]  I'm not an electrical engineer.
[16:24.020 --> 16:26.040]  So this is really straightforward.
[16:26.160 --> 16:29.060]  Almost all of this is 3D printed except for the photo detector.
[16:29.860 --> 16:36.320]  And this allows you to stabilize and then lock your laser to the fundamental transition frequency of your video.
[16:37.300 --> 16:40.200]  The next step is building a vacuum system.
[16:40.200 --> 16:42.660]  So this is kind of where it gets a little tricky.
[16:42.660 --> 16:44.700]  It's not the most accessible thing.
[16:45.020 --> 16:51.400]  Unfortunately, you need ultra-high vacuum to prevent collisions and oxidization with your reading vapor, also just to keep it in one place.
[16:52.120 --> 16:58.080]  Because again, if you're trying to cool something to a microkelvin, anything that bounces into it is going to impart kinetic energy.
[16:58.340 --> 17:00.320]  Orders of magnitude above that temperature.
[17:00.960 --> 17:03.100]  Roughly 10 to the negative 9 Torr.
[17:03.920 --> 17:06.740]  So this is something that I'm actively trying to figure out.
[17:06.740 --> 17:13.420]  And how can you make this DIY-able, something that a high school, community college, a hacker in their basement can do.
[17:15.040 --> 17:16.320]  Getting somewhere.
[17:16.820 --> 17:20.260]  But this is kind of high-level publishing the schematics for this.
[17:20.260 --> 17:21.120]  What do you need?
[17:21.260 --> 17:24.220]  Even just knowing what you need, I think, can be a little arcane.
[17:24.340 --> 17:29.300]  Because again, in these labs, I mean, this field of science is about 20 to 40 years old, depending on UBS.
[17:29.300 --> 17:32.700]  So the only way to learn, I think, is to go join a lab.
[17:32.700 --> 17:35.180]  And they're not really the most welcoming of people most of the time.
[17:36.020 --> 17:39.680]  So I just want to make sure that everyone knows what they have to do to get started.
[17:40.120 --> 17:44.060]  In science, it seems like there's a lot of friction points that have nothing to do with actually learning things.
[17:44.060 --> 17:52.060]  It's like, you know, knowing what you have to do to get started, what tools you need, or picking screws on McMaster, picking resistors on DigiKey.
[17:52.060 --> 17:54.020]  I don't think those should be friction points.
[17:54.040 --> 17:57.280]  You know, the things that should make your brain hurt are trying to learn physics.
[17:57.780 --> 17:59.740]  And I want to take care of everything else.
[18:00.200 --> 18:02.100]  So you've got this vacuum system.
[18:02.100 --> 18:03.520]  Now you've coupled light in there.
[18:04.060 --> 18:07.420]  I've got this really nice guy in Austria who makes rubidium dispensers.
[18:07.420 --> 18:09.560]  You just heat them up, it releases rubidium vapor.
[18:09.560 --> 18:14.340]  Now you've got a frequency-stabilized laser locked to the transition frequency of rubidium.
[18:14.440 --> 18:16.120]  And you've rubidium in a vacuum chamber.
[18:16.120 --> 18:17.200]  So we're almost there.
[18:18.120 --> 18:20.300]  Now we just need to cool and trap it.
[18:20.620 --> 18:26.260]  And so a typical magneto-optical trap, as I described earlier, requires these three pairs of counter-procreating beams.
[18:26.500 --> 18:31.520]  And for the average person, this requires like a lot of beam splitters and waveplates and mirrors and a lot of alignment.
[18:31.960 --> 18:37.580]  Don't underestimate optical alignment, because you just have to turn these knobs until something is precise to about tens of microns.
[18:37.760 --> 18:40.360]  This can take days for someone new.
[18:40.360 --> 18:41.260]  It took me days.
[18:41.260 --> 18:42.480]  It's really painful.
[18:43.100 --> 18:48.160]  And also waveplates are about $200 each, so you don't want to have like 10 of these in the system.
[18:49.440 --> 18:53.380]  So a single beam configuration will save you a lot of money, complexity, and space.
[18:53.380 --> 18:58.360]  And that's what I've implemented, where you've got these three mirrors in a tetrahedral configuration.
[18:58.360 --> 19:01.420]  You've got your beam coming down, and it reflects off these beams.
[19:01.420 --> 19:05.540]  And they bounce back up based on the angle into this trapping zone, this little point.
[19:05.780 --> 19:07.000]  That's where the atoms will get pushed to.
[19:07.000 --> 19:10.680]  You can do some geometry, some math, it works out.
[19:11.180 --> 19:14.160]  You've got this little stable point, and this is an image from a paper.
[19:14.160 --> 19:19.180]  This bright spot is actually about 10 million ultra-cold rubidium atoms.
[19:19.980 --> 19:24.380]  And all that they've done is press these mirrors up against a pane of glass.
[19:25.120 --> 19:26.880]  Nothing crazy going on.
[19:27.300 --> 19:31.900]  The other thing you have to do is align your magnetic fields with that bright spot.
[19:32.500 --> 19:49.840]  This also requires some precision, but I think on the scale of complexity building systems, especially compared not only to academic usual atom systems, but also, again, compared to superconducting qubits, this is something that anybody with a 3D printer and a lot of patience can do,
[19:49.840 --> 19:53.800]  which I think are just more widely applicable resources.
[19:55.900 --> 19:59.800]  And then when you've got it cooled and trapped, this is what you end up with.
[19:59.800 --> 20:04.320]  So this is not an image from my experiment, unfortunately, but from the lab I worked in.
[20:04.460 --> 20:07.260]  Just really one of my favorite scientific images ever.
[20:08.320 --> 20:16.480]  You see our cooling beam passing down the middle of this vacuum chamber, and then right in the center there's something that looks almost like a little fingerprint smudge.
[20:16.900 --> 20:20.600]  And what that is is about 100 million strontium atoms.
[20:20.800 --> 20:22.200]  And we've just pushed it to this point.
[20:22.200 --> 20:23.100]  And you can see it.
[20:23.100 --> 20:26.620]  You don't need some fancy electronic detection equipment to see it.
[20:26.620 --> 20:27.880]  You can just bend down.
[20:27.920 --> 20:30.940]  Don't burn your eyes out, but you can look at this.
[20:32.240 --> 20:40.240]  And it's kind of mind-bending to just bending over your table and seeing something that has a collective temperature of a microkelvin sitting in front of you.
[20:40.560 --> 20:42.740]  And then knowing that this has so much potential.
[20:42.740 --> 20:45.980]  You can trap it, you can apply a Hadamard gate, whatever you want to do.
[20:45.980 --> 20:49.480]  You've got quantum matter sitting in your room.
[20:51.680 --> 20:55.840]  So really, yeah, the goal of this project... I take a lot of inspiration from the RepRap projects.
[20:56.200 --> 20:58.860]  That's what one of the first RepRap models looked like.
[20:59.200 --> 21:02.340]  This open source project really drove down the cost of 3D printing.
[21:02.340 --> 21:09.100]  20 years ago, 3D printing was something that was restricted to R&D labs in industry or academic labs.
[21:09.100 --> 21:10.400]  It cost about $20,000.
[21:10.400 --> 21:13.300]  Now you can go on Amazon and get a number 3 for $150.
[21:14.060 --> 21:23.420]  And really, a big part of that, well, besides the patents expiring, were people who were willing to just go and hack some shit together, throw it on the internet, let people iterate on it.
[21:25.200 --> 21:31.180]  And people tend to think that, well, when someone's trying something, it makes it less intimidating, right?
[21:31.180 --> 21:32.680]  You have something to build on top of.
[21:32.800 --> 21:33.900]  And that's what I want to do here.
[21:33.900 --> 21:36.680]  So this is a field of physics that, again, is recent.
[21:36.860 --> 21:42.200]  It has a lot to do with quantum mechanics, things that people associate with high complexity.
[21:43.280 --> 21:46.160]  I want to make it as transparent as possible.
[21:46.740 --> 21:52.740]  So everything's going on the internet, all my CAD files, all my circuit schematics, all my PCB files, the firmware.
[21:53.580 --> 21:58.580]  And, you know, it's me hacking it together in my free time.
[21:58.580 --> 22:01.360]  So I would love it if you guys could help me iterate on it.
[22:01.460 --> 22:08.240]  The idea is to provide this development platform to explore all these really interesting phenomena that really previously weren't accessible.
[22:08.240 --> 22:13.820]  Before, you know, this decade, it's just not something that citizen scientists, hackers could do.
[22:13.900 --> 22:15.240]  But, you know, we're at that time.
[22:15.240 --> 22:21.440]  And I think it just takes people putting this information out into the world to really get the ball going.
[22:21.900 --> 22:24.900]  So, again, it's almost entirely 3D printed.
[22:25.580 --> 22:31.120]  There are as few machine parts, as few, like, things that you can't buy on Amazon out there.
[22:31.700 --> 22:36.020]  The idea is to make it as easy to replicate, as cheap, as un-intimidating as possible.
[22:36.360 --> 22:40.560]  Ideally, I want to be providing these kits to universities, to high schools even.
[22:40.900 --> 22:45.060]  Because at least I know for myself, I schooled learning quantum mechanics for several years.
[22:45.060 --> 22:55.060]  If I had something like this to provide me with tangible grounding to work towards, it would have a much more enjoyable and much more efficient process.
[22:55.700 --> 23:00.200]  My final goal is to iteratively reduce the cost to about sub 5,000.
[23:00.200 --> 23:03.320]  I've spent 15,000 of my own money.
[23:03.920 --> 23:05.940]  Also supported by the Hacker Initiative.
[23:05.940 --> 23:07.160]  They've given me a little grant.
[23:07.160 --> 23:08.300]  I'm very kind of them.
[23:09.600 --> 23:19.660]  But that's the price point that I see where anybody can spend a little bit of cash or, you know, a significant amount of cash, to do something that never really has been done before.
[23:21.080 --> 23:25.260]  So if you're interested, any contribution is appreciated.
[23:25.260 --> 23:26.760]  There's a lot of engineering stuff.
[23:26.760 --> 23:28.780]  Again, I'm not really any kind of engineer.
[23:28.780 --> 23:30.220]  I just dabble.
[23:30.440 --> 23:35.400]  So if you're an electrical engineer, if you're a mechanical engineer, I'd love to talk.
[23:35.400 --> 23:40.160]  There's probably something you can figure out that has been giving me a lot of pain.
[23:40.800 --> 23:42.040]  Any donations?
[23:42.040 --> 23:43.940]  Anybody who wants to replicate it, please talk to me.
[23:43.940 --> 23:50.100]  I'd love to see more ultracold gases sitting in apartments, not in labs.
[23:51.240 --> 23:54.560]  If you want to hire me, consult.
[23:54.860 --> 23:55.780]  I do lasers.
[23:55.780 --> 23:56.620]  I build lasers.
[23:57.960 --> 24:00.700]  You know, I make materials, whatever you want.
[24:01.100 --> 24:07.240]  But yeah, any kind of interest you have in the project or any parts of the project, please send me an email.
[24:07.240 --> 24:10.220]  We have a website at open-quantum.org.
[24:11.340 --> 24:14.280]  Yeah, I think this is the platform.
[24:14.800 --> 24:20.440]  I would really love if anybody wanted to go and had some crazy idea that they wanted to do with it.
[24:20.700 --> 24:31.080]  For me, it's just been this process of getting out of the way, having it something that my students can play with, something that's repeatable, something that is low cost.
[24:31.080 --> 24:33.220]  So thank you all for coming.
[24:33.220 --> 24:34.640]  I really appreciate it.
[24:36.340 --> 24:47.840]  Thanks to the Hacker Initiative and the Z-Lab at Columbia where I got started in all of this.
[24:49.260 --> 24:50.000]  Questions?
[24:50.000 --> 24:50.620]  Yeah, questions.
[24:50.620 --> 24:51.560]  Anybody has questions?
[24:51.560 --> 24:52.860]  Do you have your slides up?
[24:52.860 --> 24:54.240]  They will be, yes.
[24:54.520 --> 24:59.740]  If you don't have the hard vacuum, how far can you get?
[25:01.840 --> 25:05.000]  What spectroscopy setup if you don't have a vacuum system?
[25:05.000 --> 25:07.120]  So you can find one of these glass cells.
[25:07.120 --> 25:11.420]  I've got a provider who just got back to me who's going to provide this at low cost.
[25:11.420 --> 25:16.880]  And you can see fluorescence from atoms in an invisible or in a glass cell, transparent glass cell.
[25:16.880 --> 25:20.760]  And this is already kind of mind-bending because it looks like just an empty cell there.
[25:20.880 --> 25:23.600]  And as you get closer to resonance, you start to see this glow.
[25:24.760 --> 25:29.920]  And that's kind of one way to verify for yourself that quantum mechanics is real, that atoms are purple light and red light.
[25:30.440 --> 25:32.620]  What temperature will those atoms be?
[25:32.640 --> 25:35.400]  So you're not actually cooling them, they're just absorbing light.
[25:35.400 --> 25:37.480]  So they'll be more or less ambient.
[25:39.700 --> 25:43.100]  What's your roadmap for people to track?
[25:43.100 --> 25:49.520]  Yeah, so we're pretty much done with the spectroscopy setup, which I've got sitting right here, if anybody wants to come see it after I stop talking.
[25:50.320 --> 25:52.540]  I have a vacuum system working.
[25:52.660 --> 25:56.440]  Vacuum systems are actually pretty cheap on eBay, if any of you are really motivated.
[25:56.840 --> 26:00.240]  And I'm trying to find out ways to make this easier to print.
[26:00.280 --> 26:03.600]  So printing and electroplating it, for instance, seems to work some of the time.
[26:04.660 --> 26:12.660]  But I'm publishing those schematics online and then just setting up convenient control firmware because everything is kind of packed together to get it peer-repeatable.
[26:12.660 --> 26:15.520]  You flip a switch and you get a cold guess.
[26:16.080 --> 26:18.720]  Hopefully by the end of this year.
[26:21.920 --> 26:26.620]  I always ask the boring questions because I'm not a professor, but I'm a professor at a community college.
[26:26.880 --> 26:32.580]  And that's something that we are trying to get the community to recognize this technology.
[26:32.580 --> 26:34.660]  So I congratulate you for all that you've done.
[26:34.660 --> 26:45.100]  My question is, have you developed some curriculum that kind of maps introductory courses in quantum using open quantum projects?
[26:45.100 --> 26:48.120]  Yeah, absolutely.
[26:48.120 --> 26:50.700]  So I've written up probably half of the curriculum.
[26:50.700 --> 26:54.440]  I teach kind of ad hoc to students who are interested in my work.
[26:54.440 --> 27:00.420]  I've written up... I teach engineers, so I would claim to be an engineer more than a scientist.
[27:01.060 --> 27:04.620]  So everything I write is through the lens of practicality.
[27:04.620 --> 27:06.860]  Like, what do you need to make to make these things happen?
[27:06.960 --> 27:12.520]  I think going from it from the other way, like understanding theory to practice, would be interesting.
[27:12.580 --> 27:14.240]  So yeah, it's just a tough career.
[27:27.820 --> 27:29.480]  So you have to expand the beam.
[27:29.720 --> 27:31.960]  You make it really wide, and then it's balancing off.
[27:39.900 --> 27:41.820]  Yeah, so it goes like right in the middle.
[27:44.000 --> 27:46.180]  Yeah, you expand it and then you call a meter.
[27:50.420 --> 27:53.940]  Do you have any... are there any safety concerns for doing this near a polymer?
[27:54.120 --> 27:59.520]  So you need a class three laser, so I probably should also put a laser safety course on my website.
[28:00.060 --> 28:08.200]  You know, they're not powerful enough to blind you instantly, which I think is a great threshold to be below.
[28:09.060 --> 28:14.620]  If you left it running, you forgot about it, and it was pointed at something wood or plastic, you would probably burn it.
[28:15.260 --> 28:18.840]  So that's a very good point that I need to figure out how to deal with.
[28:18.840 --> 28:20.280]  You know, lasers are not regulated.
[28:20.840 --> 28:21.720]  People do worse things.
[28:21.720 --> 28:22.760]  That's my justification.
[28:24.120 --> 28:29.920]  All right, well, thanks all for coming.
[28:29.920 --> 28:35.460]  I'll be around for a while longer to talk about this, answer any questions, whatever you guys want.