What even is Quantum Computing?!

0:00 Trap. No, I have something more important. Oh! Did you trash your soul right dude? Bang her bro! My favorites are the red and the pink. Red! Obviously, you should listen to trash when it comes to stanks. Like the first time I ate it, like ate a red one, my eyes roll through the back of my head. I was like, I can't believe.

0:22 I'm sorry. What was the name of those treats? Dark treats. Dark treats. I was thinking they were dog treats. Yes, that's chewy on them. They're cutters, they're cutters. That's a dog brand, isn't it? It is. This is chewy. Chewy. By the way, chewy.com is a dog food. Yeah, yeah. You're right about that.

0:42 You're right about that. Yeah. Does say allergy friendly. The more I read the packaging, I'm wondering, trashed it. Have I been eating dog treats for the first time? They're delicious. Yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah! Anyway, sorry.

0:55 All right. Welcome to the stand-up. Today, we have with us Ginger Bill, an actual quantum physicist, teacher and trash dev. And we're going to be talking about things quantum AI. Because, you know, I don't know about you guys, but every single week, I see some breakthrough, quantum computers going to be absolutely amazing.

1:11 And this week, Google released a paper showing that it's able to do verifiable quantum algorithms, getting reproducible results through. And it's finally time, AI supremacy or quantum supremacy. Maybe quantum AI supremacy. From Google quantum AI. It's going to happen. We're going to see it in sense. None of us are smart enough except for Ginger Bill to actually understand what's happening.

1:32 We thought we would discuss this and bring on a real semi-expert. A quasi-expert. I know we're going to find a mic. To lay the foundation this week, Google released a paper, along with the marketing video showing that they've created something called quantum echoes, effectively a way for them to be able to take some sort of chip with 105 qubits on it, do some sort of perturbation is what they call it, make it vibrate, and then run the algorithm in reverse and boom, they're sucking out answers with a pretty high reproducibility and able to actually predict structures of molecules and all that.

2:07 Now for us, what does that actually mean? Is there any actual practical implication? Is 21 still the largest factor number by quantum machinery? We will find out! All right, Ginger Bill. Since all of us are actual online idiots, can you just please give us like the... Professional online idiots. Professor you, sir.

2:26 Sorry, I'm paid. I'm paid, so I'm not an amateur. Wait, Professor, I'm an amateur. It's fine. I have a professional here. Thank you. That's good. Made a couple dollars, deleted a couple posts, not a big deal. Oh no, oh no. Could you please give us like the high level of quantum computers, what this actually means, like something to kind of help us as idiots understand all these sensational articles we see constantly?

2:53 Right, I'm going to try my best here to explain any of this. Okay, so most people, if you've ever... I'm assuming the audience here knows that at least the basics of computers, right? They probably know quantum slip test as well, the double, whatever. Yeah, so I'm going to ignore all that. I'm going to go even more simpler first, just break it down.

3:14 You've got classical computers, right? It was called a classical, right? This is where you've got the same binary computers. Most are the B-Telvin, right? Yeah, all those stuff like all the glasses. That's where all the glasses strike on myself, but yes. I do like this as an artist and see mine. Oh. Now, we're going from there.

3:30 Classical computers, even like we've got binary things, you've got like the ones and zeros every day, where we call these bits, right? A bit, it can store two states either zero or one. If you've got two bits, you can store four states, zero, zero, one, one, zero, one and one, zero, right? I don't know why I didn't know that order, but that's...

3:45 You can see what it gets, and you get to three bits, you get eight, and you can see how many different states you get. Now, the thing is, in classical computers, when you've got something without many bits, it's either in one of the states only, right? If there are three bits, it can be only one of the eight.

3:59 That's it. Did you find the errors? I don't even know what they look. What do they even look like? They're in the phone. In the phone? Yeah, they're definitely in there. I just don't know how we've labeled them. I got it, don't worry. You've got to figure it out. We're running out of time, Brian. You've got to find them and meet me at the stand-up.

4:19 Roger. They're in the phone. It's so simple. Get all the context you need to debug your problem, because code breaks, so fix it faster with century. Right. That's the simple bit about, and then, basic classical computers. You've got bits. Now, when you get quantum mechanics, they have some fans here.

4:49 They're like calling them qubits or quantum bits. And what they do here, as I'm going to say a fancy word, which isn't actually that fancy at all when you actually understand it, is they say qubit is something that is a superposition of both the zero value and the one value. Now, the word superposition really confuses most people, because it just sounds silly.

5:11 But most people probably understand what it is. The one you always, like the textbook, when you're learning physics, is like waves. Right. If you can think of, imagine, it's extremely still pond. Right. And you drop like two stones at different parts of the pond. They'll produce ripples, right. When the ripples come to interact with each other, right, you've got two sets of waves coming to interact.

5:33 How those waves interact with each other add together to make one new wave. That's called a superposition of two different waves. That's it. That's what it means. You've added that literally they've come together, and that's a superposition. Now, the more abstract thing is when you talk about these qubit things, is that they are now superpositions of the zero on a one state or whatever they correlate to.

5:57 So in this case, it's a little bit more like, hey, this thing, this qubit, it's like we said, there's a 40% chance it's in the zero percent state, and a 60% chance it's in the one state, right. But that's it's like state currently. That's the state of that qubit. I'm going to make it completely static here as well, not to confuse any more and even further.

6:16 So you've got a basic, basic idea of what a qubit is. Now, the funny bit is when you have two qubits, right, let's make it a little bit more fancy. And with four, we had two bits, we have this four possible states. But when you have two qubits, you want to usually want to entangle them together. Now, this word entanglement is effective.

6:36 In that event, you can't pull them apart. Yeah. Okay, hold on. I need to know, is TJ correct? Is it entanglement like the Avengers? I actually have a major advantage. Neither the spy one or the superheros. What about like, love is blind entanglements. Like, what kind of entanglements are we talking about here?

6:55 Is interstellar a documentary? Sorry to tell you to each. It's a film by Christopher Nolan. That's my, he makes my favorite documentaries. I know. I really sometimes got, I like Tenon. That's a really good documentary, right? Sorry. Sorry. Sorry to trouble you can continue. You can't just drop the entanglements.

7:18 We're going without a salad like idiots. You got them. No, no, the word entanglements, it just means that these two, like, two states are intrinsically linked and you can't, they're not separable. It's a bad way of putting it, right? No, it's not. I don't know what it actually means, but okay, there's just two things together.

7:33 Yeah, yeah, like, they, like, it is, this is the weird bit about quantum mechanics. It's the entanglement bit, right? Superposition bit is not actually the weird bit. Yeah. That, you know, happens in classical, if you want to be like classical waves, right? They can be superpositioned. Right. And we dropped the two things in the lake.

7:48 The entanglement is literally, they just are like, joined together. You cannot separate them. There's no math model operation that can separate those two states. They're just now intrinsically linked. So what scientists have joined? Let no man separate is what you're trying to say. Is that the rule? Yes.

8:04 Okay. There's a caveat with them. I'm trying to dumb this down as much as I can. But it doesn't matter. Right. There's a reason. Right. So we've got some entanglement, right? You can entangle anything like really together. But they, in their sense, they have these qubits, whatever these qubits are represented by, right?

8:18 And then when they get entangled, they've got two of them entangled. Now, of this state, before we've had two bits, you'd have like the 001001 and 111 state, right? Now, together with those qubits, you're actually representing all four of those states at once with certain probabilities. So there's like, we say, 20% chance it's in the state A, 10% chance it's a state B, 30% chance it's in state C, and 40% chance it's in state D.

8:46 I had to add that open my head. Yeah. And they're kind of the probabilities it could be in that particular state, right? So now you've got all of them computing, like all stored at once. Now, when you add more and more qubits and things, you get more and more of these different states that are all being represented at once.

9:06 So I think like the latest computer, like I'm on computers, like, is it's over 600 qubits? So it's probably more than that now. I don't know. It's a big number in the quantum computer. Because it's two to whatever that number is. There's how many states there are. There's like more absolutely. Why would you need, so just help me understand even like a stupid question.

9:24 Why would I need something bigger than like 64 qubits or maybe 128 qubits considering it literally represents the probability of all values. Because that's the only bits they've got. Like in your computer right now, you've probably got like 64 gigabytes of RAM, right? That's a lot of bits. That's a lot of bits.

9:43 That's what they have memory-wise. That's all the bits. Yeah, but okay. Yeah. So yes, they can store a lot of states. In fact, they cannot just store two to the end. They do two to the two to the end. So it's really big. So 300 qubits is a fact equivalent to be more than atoms in the universe. Yeah. If I remember correctly.

10:05 But like CPUs process, like one operation at a time. Where's this? Yeah. Where these things are like, I don't know, saying on saying. Well, they're still doing one operation at a time. Okay. But the thing is because of their probability distribution thing going on here, right? It's kind of like is doing it all probabilistically distributed across all the possible states.

10:27 I got it. So we can do multiple, quote unquote, multiple things at once. Now for most problems, this is where like most people may not understand. Quantum computing has got some only certain cases where it's useful. Most of the time, classical, as we know at the moment, is better for virtually everything else.

10:42 Certain things are better with the quantum algorithms. So the most common one that most probably have heard of is sure's algorithm. This is the one where you can decompose an integer into its prime components. Right? Where? What do I actually listen? This can be done in polynomial time. Sorry. Prime components, ginger bill.

11:08 Keep up. The prime components. The prime again. Yeah. All right. Sorry. Keep going. Sorry. I'm trying to think as well. It's really hard to do when you're trying to actually think. You're dealing with the professional right now, okay? I'm not listening. I'm just trying to come up with jokes. Right? I had a joke about web browsers and we went right past it.

11:31 Hi, Holland T.J. Lay your web. Can I say it now? Yeah, since we're paused, I was going to say ginger bill said classical computers are better at web browsers. So I was wondering if that's kind of like ginger bill currently co-signing on the state of the web. So let's see. I'm putting the robust down, putting it great.

11:49 There we go. Looks good to me. Looks good. Looks fine to me. Yeah, I just, yeah, you got pierced. Yeah, I'll just merge it. It doesn't matter. It doesn't matter if we'll pass the tests. We'll fix it later. Never do. No, but yeah. So that's kind of the thing. So then there's the last component when it comes to quantum computing, if I can think of, which is effectively interference and decoherence, which I'm going to merge together just to keep it simple, which is not correct, but whatever.

12:18 Interference, you have to understand is that these quantum systems, if you've ever seen a quantum computer, they have to be really, really cold, like they're literally like fractions of a degree above absolute zero most of the time. The reason why you want to get them cold is you don't want any background noise interfering with it.

12:33 Because if you have background noise, it will make it noisy. And again, when these systems again, they're not discrete like your computer. They're probabilistic effectively. And if you add more noise to it, your result gets more noisy and noisier until you've got absolute noise at the end of those. It's useless.

12:51 So sometimes you'll have interference from these different probabilities, like sometimes they probably will amplify each other or cancel each other out as well. If there's even no background noise, that's another form of interference, which can be useful for certain algorithms. Again, if you have two waves on an ocean or a thing, if they're in sync, they'll get amplitude.

13:08 If they're out of phase, they'll cancel each other out. And then the last one is decoherence, which is tracking. So decoherence is the other one where you would say it's collapsing a quantum state into a classical state, right? Which is where you want to read it. You want to be able to read something in a classical state.

13:30 Is it in this state or another? Can you explain it in terms of monads? Because that feels a little confusing. Is this like the only thing I've ever done? It's dumbass out in the notes. Oh god, whatever the joke is, the endo funkters. I can't remember flipping it. Look, prime, they could be anywhere in the burrito until you open it up and look inside.

13:47 They're only in one spot in the burrito. You know how many times you close the burrito inside again? They're in the same spot. That's actually right. I understand it. Okay. Right? Yeah. See? Yeah. Well, yeah. Monads aren't really burritos either. Ginger Bill, okay. They're on computer construct. All right.

14:06 They're not actually a burrito. All right. There's not burritos in my machine, Ginger Bill. Stop. Who's the professional idiot now, Ginger Bill? Is this why you like procedural burrito? Other than on a GPU when it's red hot. Come on. You could maybe cook one on there. Okay, I'll give you that. I'll give you that.

14:25 And it's a good place to keep warm. It's a nice little warm room. There we go. It's actually the problem. Mm-hmm. All right, all right, all right, all right, all right. I have a quick question. I'm probably going to sidetracked this. That's great. We were not sidetracked right now. I'm very upset that you're doing this.

14:41 Remember that you've explained. It just sounds like it's so fragile. Yes. Good. So are they like bearing these things like undergrounds or just like no noise? No, so when you do, so to be very careful here, people who don't know I was, my specialty was not quantum computing. When I used to do it, it was quantum optics, which is still heavily related to this by the way.

15:02 So locally, I can black it enough. So I'm okay. So what they do is they don't have to put it underground. When you make things very cold in that sense, what you're doing is depending on which system you're in, if you say you're trapping, let's say an atom, let's just say you're trapping an atom, what you're doing is temperature at the end of the day is you're reducing the jiggly-ness.

15:22 That's something, right? Temperature is jiggly-ness. No jiggle. No jiggle. No jiggle. So the way to imagine when you're calling down an atom, right, imagine it's an elephant and you're shooting ping-pong balls at it, right? If you shoot ping-pong balls from every single direction, eventually, you'll slow down that elephant.

15:41 Yeah? Mm-hmm. Now, in this analogy, the atom is the elephant and the ping-pong balls are light, literally, photons, particles of light. Yeah? So the way you do this, to get it to work, I'm going to, oh, this is going to be fun explaining actually calling down atoms. This will be fun, right? I hope. I'm having a lot of fun talking to you.

16:06 I'm having a lot of fun talking to you. Right, right. You could look where to end the question with jiggle-ness, right? That's what temperature is. It's jiggly-ness. And what you're doing is you're trying to get rid of the jiggly-ness. Because when things are called, they slow down. Yeah, that's what the temperature is.

16:20 It literally is less, if you've got more jiggle-ness, it's halter. It's less jiggle-ness, it's colder. Yep. You can't ever go still. That would be absolute zero. You can't hit that. Right? Yeah, why not? That comes into quantum mechanics. And that's all to do with Heisenberg uncertainty. Okay. Well, you just check-mated me.

16:40 The time from breaking bad. Where does that come into this ginger bill? Yes. Say my name. No. No. No, so explain that. I don't know because this is like, oh, great. Explain any quantum mechanics to people. It's fun. So many new words. So many new words. Yeah. So what that means by the quantum, Heisenberg's end principle is that when you measure something, there's like a fundamental, something called action to the universe, and you can't go below it.

17:09 So when you're measuring like the position of something, right? If you get below this limit, the product that you measure to get this action is the momentum, right? Then the uncertainty that you're measuring the position in, the uncertainty of the momentum would have to skyrocket. So actually, that means it's now jiggling even more.

17:29 Or if you got that down, you don't know where the heck this atom is anymore. Right? But we can't observe it. We can't send light. You're observing it. It's going to hit us. You're observing it. You think the uncertainty gets spread so far. You can actually observe this uncertainty principle in the macro world.

17:46 It's not like a small level thing. You can actually see it in the big world. I haven't got demonstrated already to do this. Just find something on YouTube. We probably exist to find a little thing. But that's why you can't get down to real absolute zero. Because effectively, you're reaching those certain limits.

18:02 So when you get this cold, I'm just curious, I'll never explain. Quick, quick, quick question. Because I'm sure people watching us are going to want to know. Can this thing explode? I'm just like, hello, everybody up? A quantum computer? I don't know, because if I had a quantum computer in my attic and then things got hot, would it blow me up?

18:22 Great question. It depends on the quantum computer, right? Because it depends on... If I remember correctly, there's different ways of making quantum computers. You can do like atoms, ion traps, there's photon-based ones, all this, right? Oh, yeah, you're right. It depends on what they represent, right?

18:36 Of course. Of course. You clearly know this, right? You're in the know. Which brand of photon one do you like? Which brand of photon one do you like? I knew that there's different types. Stupid me. The show answers, yes, they could blow. Because you've got lasers to deal with. And these are not like...

18:52 Ooh. Like the big lasers. Also, this is clearly not what I'd wear in a laser lab, because a tie would be literally a health hazard. Because you go over a bit and do that and like, ah, dammit! I've now burnt my tie and killed myself. That's how they find stuff out there. That's pretty much how microwave's got invented.

19:09 Yeah. Yeah, the microwave. Someone is like, you notice he's a chocolate bar melted in his back pocket, didn't they? Luckily, didn't microwave his ass. But, yeah. It's true. Just the chocolate. That's real history. Yeah. You can look at that. Let's get back on track. What's the radiation that develops in a lot of people's basements that they have to get test on?

19:28 Microplastics. No, it's radon. Radon. Radon. Radon. Radon. Radon's a literary element, right? Dude walked through some sort of checkpoint and he was just like so nuclear that it was setting off all sorts of alarms. It turns out he should have like, he was probably going to die pretty soon, but it just turns out that he discovered that there's this thing that happens in people's basements.

19:53 Like real story. It's not just basements because it usually happens where near coal mines. If you've got a lot of coal deposits, they have to worry about radon gas. Like I live near my dad when my uncle and my grandpa's were all coal miners, right? So they always had to be careful of that as well. Did they have the black lung?

20:08 No, none of them are fine. They were all fine for that. Okay, that's good. Yeah. Yeah. All right, sorry. Okay, let's let's begin. But again, this is the whole point. This is the standard. But I'm never going topic. Yeah. This is good. This is good. I'm actually loving this because I'm actually learning something also.

20:23 So I've got to think about it after a question. There's a minor thing I've missed out. You've got a massive magnetic field by the way as well. You don't just shoot lasers at it in all directions. You have a magnetic field as well, which the way I'm trying to figure out how to get. It does magnetism just like help hold it.

20:40 Piggas magnetism is so strong compared to no gravity and other stuff. So what you're doing. No, no, this is a trick. What you're doing with the magnetism in a really strong magnet field is when you had to have put this only certain wave like certain frequencies or wave lengths or colors that photons can get absorbed by atoms, right?

21:01 By the electrons in the atoms. They have to be exactly like certain frequencies. If you add an electric field, I'm trying to done this down, it can make those ranges a little bit fuzzier. An easier to hit. It's about hypnosis. Amocyst levels. Got it. Yeah. Amocyst, effectively. Amocyst. Because without it you'd be screwed.

21:22 Right? Yeah, like me playing called duty. I got it. I'm following. That's why if you went into the control, you got the amocyst. It's fine. But the superior race clearly uses a mouse. Yes, right. Final like that. But that's kind of what you're doing with the magnetic fields. But if I remember also correctly, some of the early cubic ones were also like MRI, like effectively MRI machines.

21:44 Effective is how they worked. Which is a similar thing. You have to have the massive magnets. But instead of like they do, they would have been like photons but not necessarily lasers. They would have done like radio waves. Which is how an MRI machine works. Effectively. You've got giant magnetic field.

22:00 You shoot radio waves at it and then the hydrogen atoms in the water will effectively remit those photons as bits of light against you can see it. I've heard any of these words in Southern college. Yeah. Trash, you're supposed to be doing research for this episode. This is research. This will be on the test.

22:19 I was watching videos. Sorry. Five-year-olds. It was really like, this was like for a 10-year-olds. This is beyond that. Trash, watch. Right. But that's it. That's my crash course in quantum computing for you, don't guys. Okay, so I've got a question about them. So I get it. The thing that's very different about it is instead of ones and zeros, we have probability distributions.

22:44 I feel like that's kind of like a big part of it. And then we can sample from those probability distributions like a lot of times. But it's better if we can make the probability distributions be more probable to give the answers that we want. So that's where we have to write completely new algorithms.

23:02 My understanding is, I have a SHA-256 algorithm. I write it for regular real computers that actually work and can go in consumers' places and aren't like a government conspiracy. Like birds. I can do those. I write it at my house. I write it down. If I wanted to solve SHA-256 for quantum computers, we have to write a completely new style of algorithm.

23:25 That is all based around quantum gates and stuff like this. They have higher level programming languages. Maybe we should go there and then I have no problem. No, I don't think so. Yeah, I think those different gates for people who are listening, maybe logic gates, you've got like AND or NOT those kind of things, right?

23:42 They're logic gates. That's how all computers are built. Guys, we don't have to see how they're all built. We don't see how they're all built. We're not doing any gatekeeping around here. Right, no gatekeeping. We're only or gates. Or gates only. I like Zores a nice. Yeah. I like Zores a nice. But they're quite probably should say that.

23:58 You have more gates. You have these different things entirely. Yes. And they don't even have the same logic. So one of them is, this is like a effectively that happens in like matrices, but they call it an odd the mod or had a mod. If I'm not going to do the French pronunciation. So the had a mod. Thank you.

24:14 Well, aren't right. Because screw French pronunciations. I don't even get them wrong anyway. That's not the correct pronunciation in my accent. And I'm like. We don't even like you. You weren't born in France. So it's fine. What can I do? I'm trying my best. They hate me anyway. So it's fine. Just by yeah.

24:31 So we've got had a mod thing. So had a mod is. Had a normal had a mod product in matrices is you just multiply each element by itself. And that's it. So it's like a dot product effectively, but for matrices. But the had a mod product for quantum mechanics. Not quantum computing is effectively. You're changing.

24:48 You're using this to change the basis vectors. The basis of the thing you're dealing with. Track basis vectors are like x, y and z. They're they point in. Yeah. And then you're mixing the whole thing up. So instead of like having it this way. You're changing your basis to look like this. Or something.

25:05 Just rotating it around somewhere effectively. And that's the what the had a mod product does, right? And then you can use that to make probability distributions look different, right? That's my understanding. Yes, you can do it. Again, it's a simple, again, other things to add together. And most of these different gates are just different types of matrices at the end of the day.

25:28 And they do different things. So that had a mod. Again, I'm doing it wrong. It's effectively just as a slight rotation kind of thing. There's currently three whiteboards on TJ's camera right now. Scratch, can you see this? Trash, can you see this? Yep. Okay. This is like x and y, right? But really you want to do that.

25:46 You want to say that this one is like one or something like this at this one. That'll do. That'll do. Oh, right. That's out there, man. I got my cats. I did my research. I'm coming in. The bra as well. By the way, they are electrical bras and cats because direct was an autistic thought is funny. No. That's actually going to stay.

26:07 That's 100% going to stay and it's ruined now. Sorry. I'm actually going to have to call Michelle and get a new one. Okay. So you start and you've got this guy. This represents a state vector for a qubit. Maybe. Yes. Let's go. Yep. This is good. Okay. I'm going to see where his knowledge ends because he's doing good.

26:31 Okay. Chat. Okay. Chat. Chat. If you're out here, comment. Pog. Teach whiteboarding. Pog. Thank you. Okay. So then we can do a Hadamar thing. We can reflect it across an axis. Then it looks like this. I think I can't remember exactly the matrix, but it does something like that. Does something like that?

26:49 Yeah, whatever. Far less confidence in that next line. You're so right. You're so right. You're so right. Thank you. You're totally correct. That's worse. I think you're thinking about it too much. That's close enough. Yeah. But then what we can do, this one's not necessarily that helpful for finding out a better answer.

27:14 We can also rotate them. I don't remember how we do that, but there's rotating gates that you can put stuff through. So imagine trash. You can get this arrow closer to this one. Eventually you could make it point it only ket two. So then the chances of getting the right answer are higher and higher and higher.

27:32 And then we've solved the problem and I can hack your password. I don't know if that's right. How close am I ginger bill? I mean, not the password hacking bit, but like you did it literally. It's just linear algebra, man. It's just linear algebra. It's just linear matrices. Yeah, but so are AI just linear algebra.

27:49 Okay. Yeah. Everything's linear algebra at the end of the day, man. Okay. Everything's a linear algebra or weakles. So this is how you guys are describing it and it just makes it really confusing. Or at least this is how all quantum computers here. Here's my whiteboard. Okay. Now it's my time. That's a blackboard.

28:04 Everyone describes everything quantum and I'm very confused by it. Is that you first start off with a state and you somehow feed it a problem. Right. You feed the quantum all the little gates, these little problems. And somehow it's able to take that problem. And then when you look at it, the probabilities are just aligned in the giant cloud space that is two to the end or whatever it is.

28:24 Possibility is in here. It's just the correct end. It's the correct two to the end or two to the two to the end. Space is just like, here's your answer. It's these 105 bits. And you're just like, how does it go to the correct answer? I don't understand that leap where it goes from probably all the answers.

28:41 But there's one once you observe it and it's the correct one. Let me try. Let me try ginger bill and then you can tell me where I'm wrong. Go ahead. Go ahead. Okay. Okay. All starts with the flux capacitor. Okay. Okay. I'm kidding. I'm kidding. I'm kidding. Let me try and answer it. Are you with you?

28:59 Okay. So the box that they put it in, for me, this is the thing that I think I'm getting but I don't for sure. It has a custom algorithm to solve the question that you are trying to ask, right? If they have to build just like an FPGA, build programmable gate array, that's literally we build a custom computer to solve the problem that you want to put in.

29:27 I built an FPGA that could play the game of life where you would press buttons and it would move you on stuff and then you would press another button you would go, or no, not game of life, shoots and ladders. Sorry. I got confused because I know prime loves game of life. I did. I could play shoots and ladders and it would be like, you landed on 13, boom, you jumped all the way to 27.

29:44 And it was just, it was but it was just an FPGA. It could only do that one thing. We will build quantum computers for now because we don't have a programming language for quantum computer that can solve particular problems. Maybe this will chip from Google can be reprogrammed to do other things or it has lots of gates or something that can let it do that.

30:05 But like, so you have to build a custom quantum algorithm that can do basically a matrix transformation that takes some input and puts it into the output that you want. The thing that's different about it is it can do not like a lot of them at once, but it can do, I don't know how to describe it. So I don't know.

30:26 There's my closest thing. It's not. Where am I going? Once. That is the kind of mathematics you can do. Yes. Thank you. Right. It's just different is the better way of describing it. Yeah, it's hard to explain because it's like, it's like, it's like, that answered it. Okay. No, but the black box is not a, like I feel like when people say quantum computing, they're like as soon as we get the quantum computer, passwords are dead.

30:55 And you're like, we don't have a SHA 256 Reverser quantum computer algorithm yet, right? So we would have to still find a way to express that transformation in quantum computing gates or some higher level construct that could do that, that then could do the math to solve that question for you. And generally speaking, my understanding, maybe I'm wrong and this one ginger bill, it's not that it solves it instantly, but that then it's not like it's over one time when we figured out.

31:32 But it's that there's some instead of being like SHA 256, and it's going to take a trillion years to solve it. It's like, now you could do it in like a day or to use the canonical example of quantum algorithm with a shores algorithm, right? Which is the one where you find the prime factors of an integer on a quantum computer.

31:49 You can factor like an integer N, right? And shores algorithm can do this in polynomial time. In fact, I think it's proportional to the log of N. The time it takes was on a non-contin computer. Effectively, I think it's like much, much slower than log N. I don't think it's exponential, but it's like maybe N factorial or a couple of some like that.

32:11 Which is, thanks for the exponential. I thought that the factorial was bigger than exponential. It is, but it's close enough on the same order. Okay. I'm being lazy with my terms here. But in general, that means it's very, very slow, but the quantum one, they can do it in again, log N, which means it means it's actually a lot, lot faster for this particular algorithm.

32:32 Is how it does it. And there is a way of doing it, but it has to be through the quantum algorithm. And the way you're doing it is your effect to be doing through this estimation thing. And a lot of them are, when you were saying askin' like the algorithm stuff, I'm doing, I'm trying to remember, it was algorithm off the top of my head, which is hard.

32:49 And I'm also looking here, just even then, it's still not helpful with Wikipedia notes because Wikipedia is written for PhDs. Yeah, but I'm afraid to hear a lot of it, it's just... Yeah, but that's just going on. You can read it. Yeah, this is... This is a lab coat. This is actually my lab coat, but made into a Halloween costume.

33:07 So yeah. This is clearly... This is Halloween. There you go. Your career is now a joke. Yeah, it's a joke. But... It's not my career anymore. Anymore, unfortunately. I don't know. I've added a lot of thought in it. Okay. But yeah, it's kind of one of those things where you just do a lot of the math and you can break it down into these quantum logic gates and such this way.

33:30 But it just to figure out what that algorithm is. So the question is, how does it do this in the log N time? Well, it's kind of... It is kind of doing more of these estimations at once, which is why those probabilities work. And the ones that don't work effectively, they get filtered out. Another way of thinking about this through is if I can...

33:52 You know when you do audio and you've got an EQ, right? Yep. We do band pass filters, aren't you? Yeah. Good point. You're effectively doing that. That's how you filter out the stuff you want. So if certain things come in and they're like the well off the spectrum, they get filtered out. So although not probably the right answer, nowhere near, they'll be much lower down.

34:14 Where the ones that might be closest to the right answer, they'll get more amplified up. Right? This is like a Fourier transform, the thingy, a quantum Fourier transform specifically, but like it's that kind of thingy, my bobb. It's really hard to explain because I'm not going into real complicated mathematics.

34:29 I don't want to do that. Thank you for not doing that. Thank you for not doing it. I would have liked that. We'll do that on separate episodes. Watch this video in the comments. I'm like, I'm trying, guys, I'm trying. The thing is not what you say. Though about quantum mechanics, quantum computers. Somebody in the comment is going to say, well, they've already solved it.

34:48 They've been lying to us all along. So somebody's going to be mad in the comments, no matter what. Yeah. True. You don't have to be mad. True. No matter how well you explain things. We need to go dumber right now. I don't know. I actually don't know if we can. Follow up question. Follow a question for the speaker.

35:06 Yeah. Four shores. It's not that we put in a number at the front, but into our mystical quantum computer box. And then in one shot, we're guaranteed to get out the prime factors, right? That's right. Okay. So then it runs multiple times with a new state vector. Just like if we wrote a really dumb classical computer example of like, I'm going to find the prime factors by just like literally like guessing random numbers.

35:39 Or I'm going to go start at zero. And I'm going to go to, well, obviously not zero. I'm going to start at one. I'm going to go all the way to here and see which ones. In fact, I'm going to start. And then we like just keep guessing until we find them because it's a search problem. Yeah. Right. And so there's just better search.

35:55 Or there's a better way to branch prune is what you're saying is that you. Yes. Because when you think anything is about statistical right? Because again, with quantum systems, there's a lot of noise due to the stuff. Oh, early classical computers had a lot of noise to them as well, by the way. It's just that nowadays we've gotten so good at this.

36:11 In fact, well, you're not really. But we've gotten better at it nowadays that you don't get the noise. Sometimes you hit it like, especially like the newer RAM stuff where they're so crammed together, like the transistors are where they start to get like you get error happening with the like the RAM, like even in classical because quantum effects start happening.

36:28 Right. They effectively things get noisy. The electrons jump between everything and let's say electrons be known. I mean, everything gets fuzzy. So you get, that's why you need like error correcting memory a lot at the time. To get rid of that. But when in quantum, it's like is fundamentally, yeah, everything's just noisy.

36:45 And you're trying, if you do multiple times, those multiple runs will then clean things up. So think of it as effectively as like a path tracing or something like that. And then our, as we do more iterations, our state vectors probability ends up looking more like. Instead of being like, you know, we have a 0.3, 0.4, 0.2, that's 0.1.

37:14 We like end at something like, oh, we're pretty sure it's, this one is nine, or I guess it would be like 5, 5, 0.0, 0.0, and then like 0.45. Like if we had two slots that we needed to go to be like, oh, well, that's actually these don't go to 0 though. They'd be like one or something like that. That's very small, yeah.

37:34 Because we can make them 0, that would be not right. But we can make them really small. And then we're like, oh, when we get out of the thing at 0, absolute 0 can't do it. Absolutely not. Actually, this is the half of blinding. Right. Yeah, I'm going to mind it. It's your turn. It's your turn. Yep. On certainty, he's actually intrinsic to the world.

37:56 It's not like the quantum thing. It literally is what is the level of your measuring tool? If you've got like a rule, yeah. Significant figures. Yeah, yeah. You've always got literally, like if your ruler goes to, I don't know, in, say, millimeters, the error on that's going to be like half a millimeter, right?

38:13 When you can't be any more accurate than half a millimeter. What are millimeters? Okay. Rice to 30 seconds of an inch, right? Oh, thank you. 30 seconds of an inch. Yeah. If you have 30 seconds of an inch or 16th of an inch, let's say 16th of an inch on your tape measure, right? Because they usually go down to 16th.

38:27 Yeah. You have a 30 second error. Okay. We've got a 30 second error, right? On any of your measurements. So fundamentally, your error is on your measuring stick. You can't get rid of this uncertainty. Your uncertainty is fundamentally there. This is not like a quantum thing. This is just a measuring stick thing.

38:44 And you, you could hype that it, get, get less and less and less to make it more accurate. But you get more accurate in the thing that you're measuring. The thing that quantum mechanics brings in is that it's actually, there's a, something associated with it. You would start losing that information as well.

38:58 You can't go, you can get more accurate on one thing. But once you get very, very close to this fundamental limit, the thing gets fuzzier, right? The other thing goes, the uncertainty of the other thing. And this, if it's position, momentum goes, if it's time, energy goes, right? So they're the kind of the different things you're uncertain to.

39:16 So that's another thing I want to very, very clear. The uncertainty principle is everywhere. You do it in real life. Like it's not just. I got a good example. Yeah. Yeah. From real life. If your kids are playing in a different room, they act different than if you're in the same room with them under observation, the children act differently.

39:37 That's the same thing with electrons. I don't know if that's actually a good example. So if I go look at an electron, it's going to slow down. It's going to, it's going to play really nice. It's going to have some like nice with its brother and sister. So I just, they're now probably looking at like a bunch of electrons and you're acting differently now.

39:55 Yeah. They're actually like they did. Let's do anything. They don't have that weird castle. What? What else like this? Me! Oh, hey, hey, dad. We're just hanging out. It's totally nobody. I'm just going to start walking around my house. Hey, I see you. See you over there. Okay. I've got a follow up question about the next part of quantum computers.

40:16 I don't understand ginger bill. What are people planning on actually using them for besides spending a lot of research money? That's a very good question. Thank you. And the answer is, I think it's just spending a lot of research money currently. Yep. But a lot of it is like, again, they want to break encryption.

40:34 A lot, like again, sure, the algorithm is a good example of this. If they can break a lot of basic like RSA encryption, you've broken it. Can I have one more? No, I know there is some encryption which is like quantum, like, safe already. I just... A liptical hash. Yeah. A liptical curve. Sorry. A liptical curve madness.

40:55 Some of them stuff is like, you know what I'm saying, impossible to break quantum computers. They just don't have an advantage anymore. So a lot of this is... It's because it's like, you need step by step process and the resolution of each step is now. Oh, that's really smart. I don't know if that's true, but I think it's something along those lines where there's like a dependency problem that becomes really difficult for it.

41:18 Yeah. You know, there's certain other algorithms because again, it's all probabilistic. Some of the algorithms also do like, you know, like a random walk, have you ever heard of a random walk before? Monte Carlo. Monte Carlo. Monte Carlo, yeah. But a random walk for people who don't know, imagine a drunk man in a field, he will walk randomly eventually, but eventually he will get out of the field after a certain amount of time.

41:40 Give him enough time trying to... Pull me in a field right now. Right now. I'm going to see. Two dimensions, by the way. That's in two dimensions. And then when you get to three, it's fine. But the thing is, there's these called quantum walks, right, which is slightly... It's not a random man in the field at this point, it's a very quantum tiny little thing.

41:58 Not really, but it's just like, it does quantum stuff, right? And there's a lot of algorithms that help when you do quantum walk, things, right? So that's another one to use with quantum things and they can make it go much, much faster with quantum algorithms than they can do with classical. I'm just trying to think of anything else, which I can't describe.

42:18 Well, here, maybe to help jog your memory, I did draw this picture of trash. He is very confused right now. That might be the thumbnail. I tried to give you anime eyes, I'm not very good at drawing, so I could do it. Those are terrifying. Right into my soul. All right, all right, so let's... I want to talk about the Willow thing, because obviously the example that they gave was that they're going to help do predicting of various structures of some sort.

42:53 I think it was... It's not protein structures. It was like structures for specific organic compounds, if I remember correctly. I can't remember exactly what was their examples. But does that mean, based on everything you've just said, that the Willow chip really truly is like a single-served chip, that it can only do this predictive thing?

43:15 Or are they reprogrammable, like, field-programmable gatorade? I don't know, fully, but I think they're probably single-purpose chips, currently. Wow. Okay, and these chips, you have to also run at, like, negative 271. Yeah, yeah, yeah, like, very, very cold. And what is it, negative 486 Fahrenheit, or whatever it is, like that?

43:36 No, we don't use those units. Okay, Celsius is okay. We can use Calvin, though. Yeah, it's like, like, one-fail to close, yeah. Much less than one, actually, but yeah, yeah. Yeah, it is. I was going to call that out, though, I didn't want to. Yeah, very cold is what I'm saying, yeah. Because they want to minimize the noise.

43:57 They want to minimize the noise, they have to get it very cold. They want no noise and thing like that. Okay. So when you say they have the research, that means every single problem that potentially contains quantum supremacy, they have to first, like, hand-construct. Not just, like, these are not FPGA's where they're more general purpose.

44:15 They actually have to do, like, A6 for every single problem, except in the quantum way. I think so again. Which is one of the comments will tell me I'm completely wrong, and they're probably right being completely wrong. Because again, this is not my field. I know I've got the lab coat. But he's a compiler to have these days.

44:32 Oh, but do you know any reason why, like, generalized programming languages can't or currently don't exist for this? Like, how is there not just some, um, I think they're probably, oh, but they're probably the equivalent of assembly at the moment. We haven't had that revolution to get to, like, the equivalent of, like, sea or Pascal for computers, yeah, or the algorithm.

44:55 Or odor, right? Not because the thing is modern, like, normal programmers think they're all classical. They're based on classical logic, right? Get ANS, NOTS, ORS, ZORS. That kind of thing. So they do have some stuff, though. I just, yeah, they probably have some stuff already, which is why they represent all the gates.

45:13 But it's effectively, like, drawing circuit diagrams at the end of the day is what they're doing. No, they have an open-source Python, uh, circuit, but it is basically, like, what you're saying, though, yeah, it's, as I said, it's effectively assembly language level. Right. Not, uh, you construct the circuit, it looks like.

45:31 Yeah, you construct the circuit with Python. It's not like Python is being ran on these machines. Yes, right. Python to construct the thing. So it's kind of like, basically, you're just writing a compiler, um, yes. And that's why. So I think the reason why they haven't found this yet, because also it's like, they don't need it.

45:47 Like, think back in the day, well, that was von Neumann. Not a joke, von Neumann believed assembly language was two, there was not programming. He said, only a real, like, person should literally write the machine code. Like, he believed, like, oh, yeah, you should, if you're not doing real programming, let's do machine code.

46:02 I do love that von Neumann was, uh, was like, effectively just like a Twitter rage baiter back in the day. Just like, yeah, that's not real programmers. So I would do, dude, I want to say students, but like one of his postcards saying, you know, you shouldn't be making assemblies. Just write the machine code.

46:19 Come on. I also did, by the way, just look up Willow, it according to some searching, less than 20 millicalvins is where it needs to run at. Pretty cool. That's very cold. So is there, is there, is there a future where people can actually run quantum computers or is this forever in Sergey Brins basement?

46:40 Like, you just can't, you have to be this ultra ultra wealthy company could be able to have any access to this in the next 50 years. So the answer is, most likely most people's basements for the next few decades, unless if they get like, uh, with say, photonic based ones better, because usually they can work multiple temperatures or whatever, or if they found out a room temperature semiconductor.

47:06 That oh, well, X was really hopping with that. When was that last year? Not said that the doctor's superconductor, right? Sorry, superconductor. Sorry, yeah, superconductor. Sorry, yeah, it got a bus in the ass. My second doctor's are going hard right now at room temperature. Yeah, I'm sorry, I'm talking about right.

47:20 So very soon semi superconductor. So I'm not going to, oh, I'm saying this. Get a superconductor, right? And because at the moment, most of them, there are some superconductor based quantum commutings, computers out there. But again, you have to have them very, very cold. Interesting. I'll get them in temperature.

47:35 Why people are so stoked about, I thought it was just like about making trains go super fast. Oh, no, no, no, no, yeah. But the thing is about, if you have zero resistance, zero impedance, so you can literally have no endgillis. Yeah, you could have everything. It's amazing. They already have conductors on trains, bro.

47:51 And they're super in my world. Okay. There's just, you'd be amazing. Like the amount of endgillis you get from power station is in saying, but I'm not doing this joke. I'm not playing a long teach. You made the same face trapped at me when he's been listening to you. Also, you're just like, uh, like I just watched it to find my zero for a second.

48:12 All right. Teach that was pretty good. That was pretty good. Okay. I literally thought that's the joke you were making. I thought you were saying the trains thing for doing the conductor. No, no, I'm just being serious because I thought it was just about like being able to transfer things super fast and super easy.

48:29 I thought you were setting me up for it. Like, yo, all right. I'm bumping over here. Like be ready to spike this. Okay. So is it really just like, okay, I'll just put my cards on the table. It seems to me like a crazy conspiracy to try and break the internet and be able to get back doors and everything.

48:52 That's why people are super interested in it. Like I don't. That's the main one, but there's like, there's loads of other things like good. The other one they were talking about is like, yeah, molecules and such, right? So this isn't the exact paper that they're talking about in there. The actual paper that Google published at the moment, which is, I would say the title because it sounds fancy anyway, observation of constructive interference at the edge of quantum ergodicity.

49:16 They did use the term ergodic in there and I had to look up what ergodic meant and then when I looked it up, I had no idea what it meant. It's like, yeah, it's fancy work. I read the definition. I was like, can't figure that one out. Yeah, I wouldn't worry too much. They're hot for those that don't know.

49:30 Let me just give you, let me just give you like a, just a quickie in this paper horn. I have it up here right here. This is, this is the thing with ergodic. It goes like this. As entanglement growth with size, cis or a whole, as entanglement grows with system size or evolution time, the resulting dynamics are often ergodic.

49:49 And I was just like, I don't know what that word means. I was trying to read the paper. And ergodic is relating to or denoting systems or processes with the property that given sufficient time, they include or impinge on all points in a given space and can be represented statistically by a reasonably large selection of points.

50:08 That was way to me. Again, go back to the drunk man in a field. Eventually, every single point in the field, if he doesn't escape it, right? Yeah, room. That's another way of explaining that one. I can follow that one. OK. Yeah. Yeah, yeah, yeah, yeah, yeah. Bring it down to the drunk man and the whole level.

50:26 I'm going to start saying that long the time. The entire board. OK, I know. Correct. Yes. Yeah, that's what it means. Yeah. That's all it is. And when they say state, in this case, or a space, when they say space, they don't mean literal like a space in a field, they literally mean like, oh, it could be like this three bits.

50:40 Like, could it be in all three bits, different states of that? That's all that means it's like a very abstract conception of what a space is. The possible, like, probability space, all the, I love like he's saying words. And it is just I think I'm like hard on my clarity at this point. Yeah, I'm so used to believing, yeah, all these words are two many words.

51:01 It makes sense. Yeah, I'm like, I know anything. You started saying a bunch of smart things. I was like, I was watching videos too. I guys, I'm pretty much a quantum expert. I like everyone else on Twitter. I spent one day researching it and I knew everything. My, these are my, these are my notes. Superposition entanglement can be both values at once.

51:24 Has to be cold. Close enough. Close enough. That's a lot of what's my summary of every book I learned. And then we went far beyond that. I was like, yeah, I can't be in this conversation anymore. Trash, those are good notes. Honestly, most every, like, every quantum competing article should start with those because even just hey, you're going to have to as far as we can tell, keep this at basically zero.

51:48 You're like, oh, so not everyone's going to have one in their basement. Like is it already like you're just not going to happen? At least for a while. I'm going to start saying erotic. That's a big learning trash with that one and two. But that is just a fancy word. They just stupid physicists using stupid words is all I'm saying.

52:05 Okay. By the way, I also brought props today because I actually have got some molecules to exactly. Yeah, I have loads of these qubits. What you got there? Yeah, right. Oh, no, I've just broken this one. Oh, no, where do I put that back? See, if that was sion, we have to go up if that was a, this one.

52:22 I think some people might know this. This is just glucose. It's sugar. That's just sugar. I had no idea that was just. Dude, trash, you should know that one. Oh, I need it. I can sue me. I've been getting from this stream. It's adrenaline, right? I'm just making some props. No, nice. So there you go.

52:36 There's adrenaline. Okay. So then explain why do quantum computers matter for seeing the shape of these? So a lot of these things is that not those particularly there, but it was say it was something like protein specifically because a lot of like medicine is based around figuring out proteins and their shapes of them.

52:56 Because they may know the chemical like representation, but they don't actually know how it's oriented. Now, I know there's a lot of people doing that with like manually doing this and such, but if you could get a quantum algorithm to actually just figure it out, that's much better, right? With an AI is doing a really good job at the moment of this.

53:14 I was one place AI didn't really get up. Yeah, because it's a fact to be a pretty scared statistical. This ergodicity thing, figuring out all this is a searching algorithm and quantum stuff is really good at for searching for things. That's where that quantum space is all doing. It's like it's doing all the different searches around, find it probably worthy.

53:31 So if you can find the right orientation and like mapping of a certain molecule, then it helps. Like again, here's his adrenaline. Look, there's a floppy tail here. And imagine this was a fancy thing. Like sometimes it could be the orientation like, oh, this is the lowest energy level in this orientation, rather than this one, which is very like high energy or something.

53:49 So you try to figure it out and that makes sense. Yeah, I'm kind of doing it, but I got props. OK, yeah, OK, so let me just walk it back to make sure I'm understanding. We've got molecules. They exist. We believe in molecules. I'm following so far. The shape they are actually makes a difference because like we do stuff like inhibitors and receptors and all this other crazy stuff.

54:12 They got to like match up next to each other like in physical real space. Like it actually matters whether it's a curly queue or whether it's a straight line. Yes, for the medicine to work. Otherwise, it like doesn't block the thing or it does block the thing. Yeah, OK, so then we're saying we got to figure out what they look like.

54:29 Because that actually matters for us to try it. And we got to try like a lot of different kinds. Maybe adding one little more little guy here or doing one little guy over here. Something else is going to change the shape a lot. What's the most likely shape that they're going to be so that they can do the medicine.

54:43 Anything or whatever, right? That's a good answer. OK, so then we're saying we have this really big search problem. That follows a bunch of probabilistic rules. That seems like a good example of where maybe we could find or narrow the search space. Like in the bouncing ball example, which we didn't get to show KC today, but maybe we'll get a separate clip of showing KC that and just put it into the episode right here.

55:06 I don't know. They're like, this is 13,000 times faster for this random thing. So you're saying, OK, we could search for the shapes of the molecules. Like in this example, 13,000 times faster than we would with old classical computers. So we could test like a lot more different kinds of drugs faster and more cheaply.

55:23 And then we could make new medicine. Yeah, this is another poll point of the computers. Yeah, it's one of those where there is like, it's a fact to be most of these are searching algorithms. Got it. And you think look at all most of the quantum algorithms are all sorts forms of searching algorithms. Right.

55:38 OK, so that means that the, you know how there's like the, I donate my computer to medical science. And it goes and it does a bunch of protein looking and all that. Yeah, is that effectively because I mean, because even if Willow was a thing, it only runs in Larry's basement. Therefore, they donating your computer to science is still producing an absolute term. Yeah, more even though it's 13,000 times faster than a super computer.

56:01 It's just like, well, we have a million people doing it. So it's like, this will always win because it's just so much easier. Yeah, the moment at the moment, but the whole point is if you get these bigger, what faster and everything, because they're 13,000 times one is they're talking about this specific algorithm.

56:14 Which I think on their particular paper. But in like the articles I was reading, not the paper itself, they'll call it quantum echoes. What it means is they're like, it's how fast you can scramble this data, then unscramble it. That's effectively the algorithm they're talking about, like a scrambling unscramble, like you want eggs, you scramble it, then you figure out how to unscramble the eggs.

56:32 And then they may able to do it 13,000 times faster than the biggest super computer at the moment, apparently, which is impressive. But again, it's this very specific algorithm they've designed it for. Yeah. And if you need to change the algorithm, you'd have to refab the chip and everything. So there is like limitations and time and change and unknowns in the problem.

56:54 So every time you hit a roadblock, you're like, all right, back to the classical. It's because we've got to do a year and I actually need to make a failure for those chips. Like how many chips they have to be made to make you make them work? Because I know like 50% of entanglements ended in separation.

57:09 So I'm not sure how along these chips actually last four. Yeah, I mean either. I mean, too much. Is that a real statement that you just said? One post, one post, and I'm trying to see if she made it to the core. She's still broke. Oh, oh, okay. As soon as he said it, I just got so confused. But I was just like widened on that stuff.

57:28 Did I miss something? Why is it only in the US? This is crazy. Dude, that's pretty much the plot of the dark forest trilogy. Yeah. It is kind of sad that all my quantum knowledge comes from the dark. It comes from that. The three body problem. How do you think that's a good book? It's a good book. Yeah, but is this also like a Christopher Nolan document, a documentary or is it real?

57:55 It's also a document. Which one's we had the book? No, not the show. The show is terrible. The show is I learned it. I've only seen the show. I've only seen the book. No, I like the show because I didn't read the book. Dress you have to say that, bro. The show has to be even more confusing without the book.

58:14 It's like super like I was watching. I'm like, what the hell is even happening here? Anyways, okay, we got to get back on topic. We can't be talking about that that show. I don't even know what the topic should be. I was like, so we're going from this. Okay, I've got I'll do another one then. What are the odds that we have a like or maybe a better answer or a question would be what is the timeline you expect?

58:42 Because so what I found very funny is Google is like, bro, this is the first time anyone's ever even done anything with a quantum computer. Yeah, that's like what they said. I was like, this is the first time anything's ever been done. And it's way faster than a classical computer at solving this thing that only quantum computers would care about, which I was like, okay, sure, fair enough.

59:05 So like, let's say, I mean, they may throw a billion dollars at something to make it help them find a particular drug. So that does in my mind, that doesn't really count. That's just like saying we spent a bunch more money on something to do what we could have done with regular computers. Like imagine how many how many AWS lambdas you could rent for a billion dollars, right to run your thing.

59:29 So like, where do you have any guess or any feeling of when like a commercial application, not a research project like a commercial application is like, oh, we rented out Google's willow chip for a week. So we could run drug tests and it was cheaper for us than the equivalent, you know, classical computer.

59:47 Is there like a rented out, right? Because wouldn't that be the problem in the drug test would have to be like unique and bespoke in hand. They have some stuff to run some, like some amount of reprogram ability onto the chips. I think, but it's not like a you can't write like a for loop. You're putting like the gates and connecting them.

1:00:05 Yeah, they probably have some stuff. So I think it's not literally full refrap every single time. But I don't have no idea what the timeframe is like I say is that's that's that's my answer. I have no idea. I was in an interesting talk in 2010, which was like the the future of computing and it was at MSU and he was talking about by 2025.

1:00:28 The government was saying that you should be like more quantum can secure. And by 2028, you should be like, you should not have any RSAs left was kind of the time frame. And so now we're kind of arriving to this time where it's looking like things are actually kind of falling into place where yeah, maybe in three to five years something could be, you know, they could start hack and, you know, maybe only like one K RSA, even though we're at 4K, but that's not that's only going to take an extra couple bits before we're at the next level.

1:00:52 And so I was I was very curious if if you believe in his hypothesis, which the guy standing there was he made a really great hypothesis, which was if the government tells you by this time, you should be secure. The government already has that technology and it's just like a way ahead of the curve. Do you think that the government actually has good quantum technology at this point?

1:01:14 It's just not really, you know, releasing area 51 space lasers is shit or is it just like or is it all fake? That's a good question actually. Yeah, sometimes we don't have technology before the public does a long time before, yeah, but the question is, is do they have this? And I sometimes the government also just has like random advice, which is like pulled out of nowhere.

1:01:38 Like these sometimes do this, like just sell it or we'll make sure up to date, like yeah, make sure it's by 2028 guys. I don't know, it's like they're still running cobalt, make sure you update everybody by 2028. Yeah, but they might do, this is the thing, if they do, we aren't going to know. And if I did know, I wouldn't be able to say anyway, because I'd be probably under the official secrets act.

1:02:00 So which I have already signed before for a previous job, but like yeah, yeah. Change of bill confirms government conspiracy online.net right now. Secret secrets are no fun. Look, if I knew, I probably won't be able to tell you. And if I did didn't know, I still got a great point. I got from my tinfoil head on.

1:02:20 We'll be tinfoil. You're helping us understand quantum computers right now. Exactly. Well, you know, here's how many people have, like there might be an actual state of knowing. It's just your knowing of it is, you've only got a probabilistic under-knowing of it. That's it. That's it. Okay, I got an organic keyboard right here.

1:02:39 That's amazing. Can I buy some ergodic office for sure? Trash, trash then. Today's like, I need some ergodic supplies. The manager says, okay, trash orders a bunch of alcohol and goes get to drunk in the field. Just walk around in my backyard. That's a way to let fit in. You're going to get so ergodic right now.

1:02:58 This is probably going to get an ergodic this weekend. I've been pinching all the points in a given space. I don't have any other questions. I think we need to know about quantum. I would just like to say that I left with some new knowledge. But I mostly left with the same questions I've always had. Yeah.

1:03:20 Okay, so this technology really, because we never even got to the fact that there's people that are throwing out terms. I've seen this happen, especially with a lot of the clickbait, like I'm going to trick my uncle into believing something. Videos, which is just like, just wait until the quantum AI revolution happens.

1:03:36 I assume that's not a real thing in any sort of capacity. No. And that everything they're saying is just words that are too, I just assume they're just two words next to each other that sound very fancy, but have no actual value or meaning. Pretty much. Okay. I hear that. If you ever heard the word quantum talking about anybody who's not a physicist, just assume they're bullshitting you.

1:03:58 Pretty much always, right? Do you think ginger bill, though, since AI is like a big matrix, multiplication thing of, you know, probability distributions? Like that there could be potentially, like someday, like a quantum AI. Maybe they could. Maybe they would do that in the day, but I mean, anytime soon at the moment, assume it's yes.

1:04:20 Yeah, that makes sense. And TJ, is it really just probability distributions or hear me out? Is it just a really confusing compression algorithm? Which one? AI. Yes. It's very lossy compression algorithm. I think it's very lossy compression algorithm, actually. I think AI is, I would say it's more of a, it maps ideas into higher dimensional spaces, and then it points in certain directions towards those things, which that's why it makes me feel more like quantum stuff, because it also has a state vector.

1:04:55 Like, and that's why words get associated with each other, because they're near each other in the same space. But, I don't work at OpenAI, so Sam Altman hasn't told me if we've achieved AGI internally or not. I'm just doing this for the love of the game. You tell me, don't worry, you get sort of a direct came saying that already.

1:05:16 Yeah, true. Yes. All right, well, I guess I'll have to tell my uncle about quantum AI being alive. Hahaha. Which by the way is how I heard about all this, just like, do what the hell is all this quantum AI business? I'm like, you know, I don't understand those. I'll call my individually, I struggle with let alone together.

1:05:37 I'll call my closest thing to a quantum physicist friend, and we'll work on this together. Thank you for answering the age old Facebook question, so that should really help. I'll report back on my mom's wall right now. Mom, we don't have to worry. Hahaha. I forgot they called it a wall. Hahaha. That's a good point.

1:06:00 I don't even think I could have done that. I'm still have right. I've not been on Facebook in over a decade, but you still have to like, people say that what their status is, and such like that or any more, or no. People still use that thing. I don't book. I'm log into this right now. I'm going to say on, whoops, on a podcast.

1:06:18 I did try it. I did try like I logged back into Facebook because I just got so curious. Just to like see, it does it still exist. And I went back there and I realized the old things that I set myself up as are still like what I'm defined as on Facebook. Oh, see. So it says I'm in a domestic partnership with my wife.

1:06:35 And what's it called? And I'm still a rocket scientist at Jesus. Jesus incorporated it is where I work as a rocket scientist. Here you go, guys. On a podcast. On a podcast. Like podcast. Oh, right. You can't still do it. Wow. Okay. Dangerous game pulling up your, your Facebook. I was pretty sure it was fine.

1:06:59 Fine. Wow. Okay. Well, I don't have anything on there that is cancelable. I've got lots of embarrassing posts from me and my then girlfriend now wife from high school. Oh, I miss you so much. The 4 p.m. Like we haven't seen each other in an hour. Have to post it on each other's walls. I haven't seen each other since your high school math class together just an hour earlier.

1:07:22 Yes. So this is babe. I'm in sports class right now. I just really spiked a volleyball. It's not about you. It's tough. It is painful to read. I did download my Facebook export a while ago and just poked around inside. What is it? It's just like a dream. You can download everything. Trash, everything.

1:07:42 All your DMs, all your posts, everything. And let me tell you, you do not want to do that, Trash. You are going to. I don't want to do it. The biggest loser of all time. I want to take that. And then I want to do a little fine tuning. You're going to train a little bit on all your ads. Prime 10 years ago.

1:07:58 10 years ago. Oh, it's all your self is prime 10 years old. Teen Prime GPT. Prime book. Prime book. Yeah. Okay. Well, this has been absolutely fantastical. So thank you very much, Bill, for being on here. You know, I would say that the biggest takeaway, hopefully everybody has is that all those posts and articles that you see all the time are just a bunch of, probably just a bunch of hogwash for at least the next decade, is what it sounds like.

1:08:25 You never know. Maybe there'll be some amazing breakthrough that will change the entire field completely. But if the field is just iterations, we're just going to be waiting a long time until we see anything fantastic. Thank you, everybody, for being a part of the stand-up ginger bill. As always, you are fantastic.

1:08:40 Ginger bill made Odin because C sucks. Check out OdinLang.org. That's it. That's it. Okay, good. That's like a Rob Burgundy. No, I could remember if it's. Don't come or don't. Don't come or don't. I'm Charles Barkley. You're not. I'm Charles Barkley, and I'm an asshole. What does he say? I mean, don't come up as well.

1:08:59 As well. So, yeah, don't all go.com, it doesn't matter. Oh, nice. You got both of them. Okay. Nice. And then, of course, we have Tige. I don't know if you know who Tige is, but he produces courses for boot.dev. Boot.dev slash Tige, correct? Use promo code BTW for both of us. Well, I was going to get there.

1:09:17 I was going to get him. Oh, okay. I'm nice. You're Marmelon. Oh, with a twist. Yep. All right. I am prime. Of course, I'm just going to order the boxes. I see. I am prime. I also have courses on boot.dev. Boot.dev slash prime. But if you like Tige and I equally, you can do boot.dev slash by the way, for 25% off.

1:09:34 And both Tige and I will get paid. Lastly, we have with us trash dev. Trash dev. I don't have anything I can promote of you other than your Twitter. Go check out Trash's Twitter. Where he will promote anything and everything. So just slide a DM and be like, hey, I'm going to go currency. I'm going to cheat him.

1:09:50 But he will cheat the guy. He works for Yummer's buddy. Yeah. Send him some Yummer's. I'm seeking a food sponsorship, like seriously. That's what I think that's what it's calling not even kidding. On his little shill for food. I will let you have a food sponsorship on this podcast. Add infinitum with no money attempted to grab into it.

1:10:09 If you could bring your own. Completely separate. Bring your own sponsor. B-B-Y-O-S. That's the end. Make sure it's human food sponsorship as well. OK. No, no bully sticks and stuff. No bully sticks, exactly. I don't think that intro made it under the recording. So they're not going to know why that's true.

1:10:29 It's great. Yeah, they're just going to hear that. I'm going to like sticks. All right. Thank you very much for joining us. This is the standup. Bye, everybody. Boot up the day by potent errors on my screen. Turn a little car. And have a good day.