Latest / In the Loop AI / The Quantum Computing Moment: The Last Tech Frontier

Transcript

The Quantum Computing Moment: The Last Tech Frontier · In the Loop AI

0:00
Josh: I came across this post that asked the question, what comes after AI?
0:03
Josh: And I was thinking to myself, AI has been my entire universe for a long time.
0:06
Josh: What does actually come afterwards? And I thought robotics, we have space.
0:09
Josh: We also have this thing called quantum computing. And I was like,
0:11
Josh: huh, it's been a while since we've talked about quantum computing.
0:13
Josh: What's going on there? It turns out the answer is quite a bit,
0:16
Josh: starting with the president of the United States who just signed two executive orders,
0:20
Josh: accelerating the rate of progress that we have around quantum computing,
0:23
Josh: as well as personally taking the balance sheet of the United States of America
0:27
Josh: and using it to invest privately in a few of these quantum companies.
0:30
Josh: So this episode, we're going to unpack the quantum opportunity,
0:32
Josh: what's going on, give you an update on the state of things, as well as where
0:36
Josh: you can find optimal places to put your money in the case that you are bullish
0:40
Josh: on quantum as a category and you think it is going to be one of the next upcoming trends.
0:44
Josh: This, of course, none of this will be financial advice, although if you've been
0:46
Josh: copy trading the president of the United States for a while,
0:48
Josh: you've been doing pretty damn good. So that's what compelled us to want to make
0:51
Josh: this episode. We were personally curious, and hopefully this curiosity can be
0:54
Josh: shared with you guys who are listening and you can get something pretty positive out of it.
0:58
Ejaaz: You know, it's so interesting. I've always thought of the United States government
1:02
Ejaaz: as being like a customer of innovation.
1:05
Ejaaz: So like they buy the rockets, they buy the weapons, they invest in the labs,
1:09
Ejaaz: but they never actually take like an equity upside in any of these companies.
1:13
Ejaaz: And recently they've been doing that with AI and now they're doing it with a
1:16
Ejaaz: new technology, which is not technically new, but is coming into its kind of
1:20
Ejaaz: like chat GPT type moment, which is quantum technology itself.
1:23
Ejaaz: They signed two executive orders yesterday. One is focusing or directing federal
1:28
Ejaaz: agencies to work very closely with the companies that are building this new
1:31
Ejaaz: quantum technology, work with the academics and understand how this works with
1:35
Ejaaz: the sole goal of building a quantum computer
1:38
Ejaaz: by 2028, which supposedly is going to solve all our problems.
1:42
Ejaaz: And then order number two is directing all the agencies, the governments,
1:45
Ejaaz: the companies that exist and run the world today and the country of the United
1:48
Ejaaz: States of America to harden their systems against potential quantum attacks.
1:53
Ejaaz: Quantum computers, as we're going to explain in a short second,
1:56
Ejaaz: could be used very maliciously if placed in the wrong hands.
2:00
Ejaaz: Now, Josh, I know you've been looking into how these quantum computers work
2:04
Ejaaz: and how they're different from classic computers.
2:06
Ejaaz: But the way I understand it is it's very different from a classic computer.
2:10
Ejaaz: Typically, a classic computer operates in bits. So it's either zero or it's either one.
2:15
Ejaaz: And it's a bunch of zero and ones. And this computes a bunch of different algorithms,
2:20
Ejaaz: which run on the AI chips that we speak about all the time on this show,
2:24
Ejaaz: or any kind of software that you run on your app or on your phone or whatever that might be.
2:28
Ejaaz: Now, the difference here is with the quantum computers, they have this thing
2:31
Ejaaz: called qubits, which means that they could be a zero and a one at the same time.
2:36
Ejaaz: So it's this weird thing called supervision. I don't know. Can you explain this?
2:39
Josh: Yeah and and every state in between so it's basically not any state until it's
2:43
Josh: observed and that state exists in this thing called the superposition which
2:47
Josh: means it's both of them at the same time and both of them not at all it's this
2:50
Josh: very weird concept drop your head around but applied to a practical sense,
2:54
Josh: we are trying to create as many qubits as possible which you could think of
2:58
Josh: as the modern-day equivalent of a bit but the futurized version and the thing
3:02
Josh: that that unlocks because they exist in the superposition is.
3:05
Josh: Is the interesting part. When you think about encryption, the thing that underlines
3:09
Josh: all of the internet, everything digital, it's the way that we have privacy,
3:12
Josh: it's the way that we have security that relies on this thing called prime number factorization.
3:17
Josh: And basically it exists by multiplying two very large numbers together,
3:21
Josh: but then reversing the process to figure out the numbers that were multiplied
3:23
Josh: to get that large number is really difficult.
3:26
Josh: And traditional computers have to spend millions to billions of years trying to decrypt that.
3:31
Josh: And because it takes so long to do that they're incapable of decrypting standard
3:35
Josh: login information for example when you type in your password it's encrypted
3:38
Josh: using this prime factorization what quantum
3:40
Josh: decryption can do because it has these qubits that exist in all these positions
3:43
Josh: at once it can do this factorization math very quick and it can reverse the
3:48
Josh: process of let's say a password that's been encrypted
3:50
Josh: in a matter of seconds maybe even less so it it kind of breaks down the underlying
3:55
Josh: security of the internet i guess we could say and that is one of many.
3:59
Josh: Second order effects that happens from encryption.
4:01
Josh: I think we have a lot of kind of progress to be made around biology,
4:05
Josh: taking these large swats of data like we have with our DNA makeup and making
4:08
Josh: sense of it. There's a lot of really interesting use cases and applications.
4:11
Josh: It's just been difficult because we haven't been able to figure out these things
4:14
Josh: called fault tolerant qubits, basically bits on a computer that work.
4:18
Josh: And we're measuring these by the tens and the dozens.
4:21
Josh: We're not talking about terabytes, which are trillions of bits, but we are on our way.
4:25
Josh: So it very much feels like that very early stage of the exponential where
4:29
Josh: we're back in the day we used to measure bits before we even got to a kilobit
4:32
Josh: of information and then we got to a megabit and then terabit and then petabit
4:36
Josh: and then we does the whole thing we're at the the bit stage where we can measure
4:40
Josh: them individually instead of by the dozens or hundreds and
4:43
Josh: That is where we're getting in now with this executive order and President Trump.
4:47
Josh: He's buying real early into this exponential curve here.
4:50
Ejaaz: And I think like, you know, you mentioned the number that were in the tens and
4:53
Ejaaz: the dozens, and that might lead you to think that like we haven't made much progress at all.
4:58
Ejaaz: But I was reading like, you know, if you had 300 of these qubits,
5:01
Ejaaz: you could basically simulate every single atom in the universe,
5:04
Ejaaz: theoretically. We're not quite there yet, but practically you could.
5:07
Ejaaz: The other thing I was thinking about is, okay, if I'm listening to this and
5:11
Ejaaz: I'm still confused by everything you just said, what are the main applications
5:15
Ejaaz: that you could potentially use for a quantum computer that is supposedly going to be built in 2028?
5:19
Ejaaz: Well, as I understand, there's two main applications. One is simulating molecules.
5:23
Ejaaz: So in the sense of creating new drug discoveries where you can cure certain
5:29
Ejaaz: ailments or cancers or whatever that might be, quantum computers theoretically
5:33
Ejaaz: could fix this in a matter of hours if applied in the correct way.
5:37
Ejaaz: But also So materials that we don't even know could exist today,
5:40
Ejaaz: combinations of chemistries and compounds or whatever that might be,
5:44
Ejaaz: can be solved by quantum computers.
5:46
Ejaaz: So anything from a molecular simulation side of things can be resolved and solved
5:50
Ejaaz: by a quantum computer once it's functionally accurate.
5:54
Ejaaz: The second application is cryptography. Cryptography basically hardens every
5:59
Ejaaz: security system in the entire world right now. It underpins.
6:02
Ejaaz: It's the substrate behind Bitcoin and a lot of cryptographic security systems in the world.
6:08
Ejaaz: Now, if you have a quantum computer, you can basically surpass or break through
6:13
Ejaaz: this barrier, which is why executive order number two from Trump signed yesterday
6:17
Ejaaz: basically said, hey, any like major system that runs and operates a bulk of the world's GDP today
6:23
Ejaaz: needs to start preparing and becoming quantum resistant, including all the crypto
6:28
Ejaaz: and blockchain types of stuff.
6:29
Ejaaz: So I guess now's a good time to kind of get into what the executive orders are
6:34
Ejaaz: and what they actually say.
6:35
Josh: Yes, these two executive orders were just signed this week. Order one launches
6:39
Josh: a national effort to produce a quantum computer capable of performing important
6:42
Josh: scientific calculations.
6:44
Josh: It also aims to develop quantum-enabled sensors and networks within the next
6:48
Josh: five years. Okay, that seems pretty quick.
6:50
Josh: Order number two directs federal agencies to transition their computer systems
6:53
Josh: to quantum cryptography, post-quantum cryptography standards, by 2031.
6:58
Josh: And it positions the U.S. government to lead wider adoption of these stronger
7:01
Josh: security standards to address future quantum computing threats.
7:04
Josh: That seems conservative. because at the rate that progress is being made for
7:09
Josh: them to transition all of their computer systems to cryptographic
7:13
Josh: resistant quantum by 2031 I don't know we might get there a little bit beforehand
7:17
Josh: but those are the two things it's noteworthy that these are being signed because
7:20
Josh: that means it's at least front of mind for enough people to get these on executive
7:24
Josh: orders to get these passed get these bills signed
7:26
Josh: and I think it signals the fact that this is very real this is happening and
7:29
Josh: there's a tremendous amount of progress being made in the space.
7:32
Ejaaz: Now the last time that the government took such action for a specific technology was AI.
7:39
Ejaaz: And if you remember, it was also followed very soon after by an actual investment
7:45
Ejaaz: from the government into equity stakes in different companies in AI, right?
7:49
Ejaaz: So they made the Intel investment where they acquired, I think,
7:52
Ejaaz: 10% of the stock or the company itself.
7:56
Ejaaz: They did something similar about three weeks ago when they announced the CHIPS
8:02
Ejaaz: Act, where they invested $2 billion across nine different quantum-based technology companies.
8:09
Ejaaz: Most of them are private. There are some public.
8:12
Ejaaz: And we'll get into what those companies are. But if I were to zoom out and think,
8:16
Ejaaz: why is the government making this investment now specifically?
8:19
Ejaaz: I think they've been rattled by the whole AI thing. I think they've seen how
8:23
Ejaaz: quickly AI models are progressing. thing.
8:25
Ejaaz: They've shut down Fable because it can basically access any of their security
8:29
Ejaaz: systems. And they're thinking, hmm, we need to get ahead of this other technology,
8:33
Ejaaz: which could potentially pose an even worse threat.
8:35
Ejaaz: And we're going to be more proactive. We're going to invest in the companies
8:39
Ejaaz: that are actually building this foundational layer before they go public.
8:42
Ejaaz: And we're going to be able to effectively control what they build and how they
8:46
Ejaaz: build it and get access to the thing before they release it publicly. Now,
8:51
Ejaaz: I'm in two minds about whether this is a good or bad thing.
8:54
Ejaaz: On the good side, I like that, you know, a government is taking an equity stake
8:58
Ejaaz: in the GDP and companies that are prospering, that are innovating in their own domain.
9:03
Ejaaz: But I'm also like, are you trying to nationalize this technology before it even
9:07
Ejaaz: comes to fruition? And I feel like that has bad effects on the whole capitalist side of things.
9:12
Ejaaz: Either way, I think it's cool. I think it's interesting. I think the timelines
9:16
Ejaaz: are a little conservative only because you have AI models now that you could
9:19
Ejaaz: probably chuck at this problem and it'll accelerate quantum chip development
9:23
Ejaaz: probably pretty rapidly at this point.
9:26
Josh: Yeah, listen, I'm happy for them. They're buying it early, but I'm interested
9:28
Josh: in how I can buy in early. I think that's what I want.
9:30
Ejaaz: To talk about.
9:31
Josh: And to get there, we probably want to start with the big ones,
9:34
Josh: the ones who have solved this problem called fault tolerance.
9:37
Josh: There's this problem with these quantum bits because they don't really exist
9:40
Josh: how you want them. They're very noisy and they're impossible to measure as a
9:44
Josh: result. So therefore you get a lot of noisy results as you test these things.
9:47
Josh: They don't work. there's this thing called fault tolerance where you kind of
9:50
Josh: pair together these noisy qubits and it gives you a local qubit that local qubit actually reads out
9:55
Josh: what we call fault tolerant outputs and that is the measurement that these companies
9:59
Josh: are using to determine how powerful their quantum chips are
10:02
Josh: the two companies leading this right now are the two you would expect these
10:05
Josh: are not anything surprising this is google this is microsoft they have been
10:09
Josh: leading the race for a long time google has this roadmap on their website actually
10:12
Josh: which i thought was phenomenal
10:14
Josh: and it kind of walks through the progress that they plan to see along with dates that they have
10:18
Josh: And as we click through these, I mean milestone 3 is coming along It's an order
10:23
Josh: of magnitude improvement.
10:24
Josh: Then milestone 4 is another order of magnitude 5 and 6 are also.
10:28
Josh: Full order of magnitude improvement. So each step of this process that they're
10:31
Josh: planning to do, maybe every couple of years is going to be a full 10x improvement.
10:36
Josh: And the same is true with Microsoft. So those are the two companies that we
10:39
Josh: know are creating these quantum computers, they're the furthest ahead,
10:43
Josh: they have the most resources allocated, that is kind of the top level of the stack.
10:47
Josh: If you go further down the stack, there's another couple of options here.
10:50
Josh: And those use a slightly different way of going about this quantum race.
10:55
Josh: So you could think of it as like three different ways of doing quantum computers.
10:58
Josh: There's the traditional superconduction, which means you require a lot of cooling.
11:03
Josh: That's when you see those crazy big computers.
11:05
Josh: They cool a tiny little chip and it freezes it to the point where it removes
11:08
Josh: all the noise from the system.
11:09
Josh: Then there's a second thing called trapped ions, which is where there's a second
11:13
Josh: set of company. And then a third, which is called neutral atoms.
11:16
Josh: It gets a little technical there, but let's go into the stack just below Microsoft
11:20
Josh: and Google. You guys, do you have a few options over there?
11:21
Ejaaz: Yeah, Google and Microsoft have been kind of like forging the path at least
11:26
Ejaaz: publicly for everyone to kind of like see.
11:28
Ejaaz: One thing I couldn't help but notice is the updates that these two companies
11:32
Ejaaz: who are basically leaders in this field at this point have primarily been infrastructure updates.
11:38
Ejaaz: So the updates are like, if you look at the news here that I'm showing from
11:41
Ejaaz: Microsoft's recent major runner to quantum chip news is basically,
11:46
Ejaaz: hey, yeah, we increased the architecture by a thousand fold and then when you
11:49
Ejaaz: look into like what it actually does you're like
11:52
Ejaaz: I don't really know. It's just a cool science experiment that we could run in our lab.
11:55
Ejaaz: And like, we brought the error rate down and we're like, great.
11:58
Ejaaz: Like, what are we actually using it for?
12:00
Ejaaz: So we dug into what companies are actually, one, public or two,
12:05
Ejaaz: private, but actually building something cool and uniquely available that retail
12:09
Ejaaz: might be able to play around with eventually.
12:11
Ejaaz: And the number one company that came at the top of that list is that age-old
12:15
Ejaaz: company, that age-old computing company called IBM.
12:19
Ejaaz: And the US government- That's crazy. Yeah, it's a dinosaur. Exactly.
12:22
Ejaaz: And the US government had a $2 billion investment vehicle, which they dispersed
12:27
Ejaaz: three weeks ago that I mentioned earlier.
12:29
Ejaaz: And 50% of that $1 billion went into IBM.
12:33
Ejaaz: So you might be wondering, why on earth was it going into IBM?
12:36
Ejaaz: Well, IBM has actually been cooking up a quantum product for a while now.
12:40
Ejaaz: It's called IBM Quantum, and it serves over 210 organizations.
12:45
Ejaaz: And they've been working on their qubit chip, essentially their quantum computer for a while now.
12:50
Ejaaz: They have it in a very basic form. It's not really widely accessible to the public.
12:54
Ejaaz: But I was wondering, what are they using this for? Well, there's two main applications right now.
12:58
Ejaaz: One is being run by the Cleveland Clinic, and it's being used to simulate a
13:03
Ejaaz: specific type of protein that can treat certain ailments that their patients experience.
13:09
Ejaaz: They're focusing on building a specific protein that is around 300 atoms in
13:13
Ejaaz: size. So we're talking about literally the most microscopic of microscopic types
13:17
Ejaaz: of molecules. But the idea is it'll form a foundation to build antibodies that
13:22
Ejaaz: they can serve up to their clients.
13:23
Ejaaz: Again, it's nowhere near at the clinical trial level, but it's something that
13:27
Ejaaz: they've been playing around with. And apparently it's been very useful for them.
13:30
Ejaaz: The second company is in the complete opposite end of the field,
13:33
Ejaaz: which is HSBC has been running quant algos using their version of a quantum computer.
13:40
Ejaaz: So quant, quantum, there you go, to basically come up with new algorithms for trading.
13:45
Ejaaz: Now, the last time I checked, HSBC hasn't made all the money in the world.
13:48
Ejaaz: They haven't beat Jane Street. They haven't beat Citadel. So I think we're okay for now.
13:53
Ejaaz: It's interesting to see the kind of applications and interest that,
13:56
Ejaaz: or rather the variety that a lot of these companies are experiencing for IBM's
14:00
Ejaaz: products specifically.
14:01
Ejaaz: You've got Mercedes-Benz that's using this, Exxon Mobil, Bosch is using it.
14:05
Ejaaz: So anywhere and anyone across the hardware layer, the hardware stack,
14:09
Ejaaz: whether it's cars, whether it's drills, whether it's motors,
14:12
Ejaaz: whatever that might be, is playing around with this quantum stuff to try and
14:15
Ejaaz: figure out how they could advance their own science and materials technology.
14:19
Ejaaz: And then the other companies that kind of like top of this list that isn't publicly
14:23
Ejaaz: invested in by the government but we're spoken about is google they've been
14:26
Ejaaz: spending a lot of time working on this
14:28
Ejaaz: uh they have their chips i think sundar pichai last year announced a major breakthrough
14:32
Ejaaz: and then you have a few other companies which we'll get into now
14:35
Josh: Okay so we can help with biology but i'm still interested in where it's going
14:38
Josh: to start to break things and destroy things because it feels like that will
14:41
Josh: be then too large to um ignore
14:44
Josh: and it seems like i have a few numbers that i found during research today that
14:47
Josh: we need about 4,000 to 6,000 of these bits,
14:52
Josh: in order to start actually breaking the standard encryption.
14:56
Josh: And we need even less to break elliptical curve encryption. Now,
15:00
Josh: you don't need to know what elliptical curve encryption is.
15:02
Josh: All you need to know is that is the encryption that underlines Bitcoin and modern
15:05
Josh: messaging and a lot of that type of encryption.
15:08
Josh: That needs 1,200. The current numbers today, where we're at,
15:12
Josh: Google Willa, which was announced, I believe, at the end of 2024,
15:16
Josh: early 2025, has 105 of these.
15:19
Josh: IBM has 120 of these, scaling very quickly to 360.
15:24
Josh: And when they talk about the way that these systems improve,
15:27
Josh: they are orders of magnitude.
15:28
Josh: They will continue to 10x, which means we are maybe one to two full orders of
15:33
Josh: magnitude away from really having like a pretty significant impact on standard encryption.
15:38
Josh: So we have our biological impact. We have the encryption impact.
15:41
Josh: We also have, I'm sure, a lot of mathematical use cases that we don't understand.
15:45
Josh: Within these companies, there are more that we haven't talked about.
15:48
Josh: The first one is Rigetti, which I actually haven't heard of, EGES.
15:51
Josh: What is Rigetti and why is it an interesting quantum company?
15:55
Ejaaz: Well, Rigetti has been around since 2021 and there's actually a really good
15:59
Ejaaz: analogy to explain what they do.
16:00
Ejaaz: So in the AI world, when you create all these chips and you want to serve up
16:03
Ejaaz: this wonderful compute to AI labs, there's this intermediary player that basically
16:08
Ejaaz: helps you do this, they're known as a neocloud, or in the traditional sense,
16:12
Ejaaz: it's Amazon Web Services.
16:13
Ejaaz: They run the- No, we have neoclods for quantum. Yep, yep, yep.
16:16
Ejaaz: We have neoclods for quantum. That's essentially what Rigetti does.
16:19
Ejaaz: And they own the entire stack, which is really interesting and separate from
16:23
Ejaaz: like the typical cloud services.
16:25
Ejaaz: They design the chip and they fab the chip.
16:28
Ejaaz: So as you can see, like a vertically integrated stack like that,
16:31
Ejaaz: it's kind of similar to what Google does with TPUs on the AI chips,
16:34
Ejaaz: and they serve up the actual AI model themselves.
16:37
Ejaaz: That's the equivalent of what Rigetti is doing. But at the very small market cap of $7.1 billion.
16:44
Ejaaz: Now, we were looking at like the stock performance over the last or over its
16:49
Ejaaz: existence, and it's up 122%.
16:52
Ejaaz: It hasn't been performing too crazy on the news. I think like it probably reacted
16:57
Ejaaz: pretty well from the government investment about three weeks ago.
17:00
Ejaaz: But the idea is these companies are entirely small. They haven't proven their
17:04
Ejaaz: ability to scale yet because there is nothing to scale. The major players,
17:08
Ejaaz: Google and Microsoft, hasn't actually built something yet that is worth noticing.
17:12
Ejaaz: Now, the second company that I want to speak about that the government also
17:15
Ejaaz: put in another $100 million check-in is this company called D-Wave.
17:20
Ejaaz: Now, if we scroll to like the old time, it's up a similar amount.
17:23
Ejaaz: It's around 141%, and it was formed around the same time that Rigetti was as well.
17:30
Ejaaz: Now, the whole form of D-Wave is, okay, so let's say you create this quantum
17:34
Ejaaz: chip, and let's say you're an enterprise that wants to use this thing.
17:37
Ejaaz: You need to heavily optimize everything from the chip to the actual application
17:41
Ejaaz: within your enterprise to make it make sense.
17:44
Ejaaz: The irony is we're seeing the same thing here. You have all these fancy AI models,
17:47
Ejaaz: but all these enterprises are like, I don't know how to connect this to my team
17:50
Ejaaz: or to my database or to do the thing that i wanted to do that's essentially
17:54
Ejaaz: what d-wave quantum does
17:56
Josh: Yeah i think i want to note something that i see in both of these charts as you're pulling them up
18:01
Josh: they were very dormant until when when does the volume actually start like when
18:04
Josh: does it pick up around the end of 2024 yeah and
18:07
Josh: that synchronizes with the announcement of google willow which signaled to the
18:12
Josh: world i believe that quantum computing is serious they have achieved a,
18:17
Josh: Like a meaningful amount, a triple digit amount of these fault tolerant qubits.
18:20
Josh: And there's a clear trajectory to scaling that.
18:23
Josh: And I think that moment in 2024, when you see both of these charts start to
18:27
Josh: move, although it hasn't been unbelievable gains, it's been a much different
18:30
Josh: than the dormant four years prior that these companies have been trading.
18:33
Josh: It feels very much like the early days of AI. And I was just listening to this
18:37
Josh: conversation with Fei-Fei Li.
18:38
Josh: She's one of like the godmothers of AI. She was one of these three critical
18:42
Josh: components that led to current day AI that we have using this thing called ImageNet.
18:46
Josh: And it reminds me very similarly of the story that she was telling where it
18:50
Josh: was very much a research project for a very long time and a lot of people didn't
18:54
Josh: take it too seriously they saw the signs but they never thought it would work
18:57
Josh: they thought it would be years decades away
18:59
Josh: and the reality is is that all it took was a single breakthrough,
19:03
Josh: with product market fit as it relates to something like chat gpt to really start
19:06
Josh: this race so it feels to me now like we are in the research and development phase
19:11
Josh: of quantum where a lot of these companies are spending lots of money the government
19:14
Josh: is participating they're investing a lot they're showing some signs of life
19:17
Josh: these companies are not completely dormant we have some volume here it is being traded
19:21
Josh: and there will be a moment in
19:23
Josh: time in which that switch flips how long is that going to be i don't know
19:26
Josh: but now feels like that early time where okay they have a hundred this year
19:30
Josh: maybe they get a thousand two years from now maybe they get ten thousand two
19:34
Josh: years from then at what point.
19:36
Josh: On that scale? Does it actually start to meaningfully impact the world?
19:39
Josh: And I think it's similar to AI where there will be a moment in which it very much does.
19:43
Josh: Like if they are able to start decrypting the internet, that's like a huge deal.
19:47
Josh: And there very much should be a race to solve this. And I will assume this will
19:50
Josh: be like a very large government thing because the downstream implications of
19:54
Josh: what these quantum computers can do is huge.
19:56
Josh: So when I see these charts, when I look at these companies, I'm imagining a
20:00
Josh: five to 10 year investment that I'm going to put in to seed this thing
20:03
Josh: to be early in the research development and hope that one of these will actually wind up being
20:06
Josh: the core weave neo cloud it will be the i mean will the next google be google
20:11
Josh: probably possibly it seems like they're the furthest ahead
20:14
Josh: but that's kind of how i'm looking at this right now it's like these are research
20:17
Josh: projects that are showing some signs of life and some will win and the ones
20:20
Josh: that do it looks like they're going to win pretty big.
20:22
Ejaaz: Yeah i mean the coolest part is you aren't just looking at small random companies
20:28
Ejaaz: that you've never heard of that are working on this on these technologies they're
20:32
Ejaaz: important companies google and Microsoft are name brands.
20:35
Ejaaz: And the good part is they've been investing in this for well over a decade.
20:39
Ejaaz: And Google, you know, somehow manages to be the first for all of these different
20:43
Ejaaz: technologies. They were essentially the first that invested so much in AI.
20:47
Ejaaz: And they're the same doing it on the quantum side of things as well.
20:49
Ejaaz: So just as like a general basket bet, usually their companies do very well as
20:54
Ejaaz: in terms of exposure to these newer technologies.
20:56
Ejaaz: In terms of the convergence of AI and quantum right now, and the government
21:01
Ejaaz: making significant investments in both of these technologies.
21:04
Ejaaz: I don't think it's a coincidence by any chance. If I think about these two technologies,
21:09
Ejaaz: I kind of think about them as very rudimentary tools, which you can kind of
21:14
Ejaaz: like wave at any problem.
21:16
Ejaaz: And theoretically, they could fix it. And we're nowhere near that right now.
21:20
Ejaaz: Air models hallucinate. They get a lot of things wrong.
21:23
Ejaaz: And then on the quantum side of things, we don't even have anything to use right now.
21:26
Ejaaz: But the idea is directionally, over that five to 10 year time span that you
21:29
Ejaaz: mentioned, Josh, I believe we'll see these two technologies be the most important
21:35
Ejaaz: and impactful on humanity as a whole.
21:37
Ejaaz: And what that yields, what breakthroughs and innovations, material sciences,
21:42
Ejaaz: molecular compounds that that creates, we don't know yet and we don't need to
21:47
Ejaaz: necessarily know yet, but it's definitely two tracks that we need to pay very close attention to.
21:52
Ejaaz: And I have a feeling, a strong feeling that both of these technologies will
21:56
Ejaaz: be very heavily intertwined. If you are an AI model provider and you want a
22:02
Ejaaz: smart computer to kind of like regulate and guide the AI model.
22:06
Ejaaz: Quantum computer sounds like it kind of like knows what it's doing.
22:08
Ejaaz: It understands the molecular simulations of a bunch of how the world works,
22:12
Ejaaz: the physical reality. It's interesting you mentioned Fei-Fei Li,
22:14
Ejaaz: like she's building an amazing world model, which I believe is gonna be the
22:18
Ejaaz: next iteration of AI models.
22:20
Ejaaz: What's one great unlock about a world model? It understands physical reality,
22:24
Ejaaz: gives the LLM an understanding of how the world actually works,
22:28
Ejaaz: sees and feels through the human kind of set of sensors.
22:32
Ejaaz: And I believe quantum computers could potentially be that to AI models.
22:35
Ejaaz: Now, a lot of that is hearsay right now.
22:37
Ejaaz: I have no proof for all of this. I'm not a quantum expert, but it feels generally
22:41
Ejaaz: that's where the world is taking a step towards.
22:45
Josh: Yeah. And in terms of tangible timelines, we have today, we have dozens to 100
22:49
Josh: qubits. By 2027 to 2029, the expectation is to get hundreds of these.
22:55
Josh: And then by the early 30s, the expectation is for thousands of these.
22:59
Josh: And those definitions and the timelines, as we know, they get very fuzzy.
23:02
Josh: Like you mentioned, AI is very closely intertwined. A lot of these things kind
23:05
Josh: of converge on themselves.
23:07
Josh: They speed up expectations. And we know this. We've seen this before where we,
23:11
Josh: we, everyone was like, no, AGI isn't going to come for decades.
23:14
Josh: And like, here we are right on the doorstep.
23:16
Josh: The estimates have been a little weird. Back in 2019, it was estimated that
23:19
Josh: it would take about 20 million qubits to break the popular encryption RSA 2048
23:23
Josh: that the internet runs on, and it would take a couple hours.
23:26
Josh: By last year, a Google researcher showed that it could be done with under 1
23:30
Josh: million qubits in under a week.
23:32
Josh: And then new error correction schemes are probably going to push that number even lower.
23:36
Josh: So we're getting this accelerated rate of progress that I think we saw in AI,
23:39
Josh: and that's going to push things in a really interesting place and maybe cause
23:43
Josh: a little bit of panic because the downstream implications of this are huge.
23:46
Josh: But I think that is, that's the update.
23:48
Josh: That is the all encompassing kind of state of the union on quantum as it relates
23:54
Josh: to just the companies involved and how to invest in them.
23:58
Josh: Thankfully, a lot of them are public. Thankfully, a lot of them are not even trading that high.
24:01
Josh: I'm not sure if they will for quite some time, but those are the players to
24:04
Josh: look out for. Those are the companies that the government is investing in.
24:08
Josh: Those are the companies that are showing signs of life on the chart for the first time in history.
24:12
Josh: And those of the companies that are seeding the roots to this new exponential
24:16
Josh: growth curve that we have in the world of quantum. It's much earlier than something
24:19
Josh: like AI, but I imagine we'll see a very similar trajectory.
24:22
Josh: And that's something excited to be in. So personally, I mean,
24:26
Josh: I'm not in any quantum exposure right now. Maybe I should. I mean,
24:29
Josh: I guess Google accounts technically.
24:31
Josh: But what do you think, Ejaz? Is this something you're interested in?
24:33
Ejaaz: So one, yes, something I'm interested in, but only to the sense where it's like
24:37
Ejaaz: a decade hold at this point.
24:39
Ejaaz: The other thing I really wanted to say is it's kind of nice that a lot of these
24:43
Ejaaz: companies that were invested in are essentially public. Like there's a bunch
24:48
Ejaaz: of private companies, but
24:50
Ejaaz: IBM has been around for a while and also these other small cap companies like
24:54
Ejaaz: Rigetti and D-Wave, which, by the way, I'm not recommending investing at all.
24:58
Ejaaz: I own nothing in it. Disclaimer, this is not investment advice.
25:02
Ejaaz: It's just interesting to see that they are publicly available for the retail
25:06
Ejaaz: to get access to if they believe in this thesis and if they believe those companies
25:09
Ejaaz: are going to be able to fulfill the vision that we kind of like laid out on this episode.
25:13
Ejaaz: With all of that being said, I'll have a small back in quantum.
25:16
Ejaaz: I do believe I've been bullish on it for a while now. But to be honest,
25:19
Ejaaz: like one, a lot of the technology itself kind of escapes me,
25:22
Ejaaz: I need to do a lot more learning.
25:24
Ejaaz: And two, I think AI is still going to be the focus for a while now.
25:27
Ejaaz: But that's where the asymmetry when it comes to investing kind of lies at the
25:31
Ejaaz: moment. But I believe that is it for this episode. That's the quantum episode.
25:36
Ejaaz: If I had one single prompt for the audience, it would be if you're one of the
25:41
Ejaaz: 210 organizations that are working with IBM on their quantum product,
25:45
Ejaaz: what are you doing with it?
25:46
Ejaaz: Like, or if you know something about this that we haven't mentioned on the show,
25:49
Ejaaz: I would love to hear more about what you guys are kind of like thinking about
25:53
Ejaaz: how these quantum computers will be applied because right now it's not too clear
25:56
Ejaaz: and maybe that's just the simple answer at this point.
25:58
Josh: As always, thank you guys so much for watching. If you enjoyed this episode,
26:01
Josh: please don't forget to share it with a friend who you think might also enjoy.
26:04
Josh: We are also looking for sponsors. If anyone wants to support the show,
26:07
Josh: if you want to contribute to the balance sheet of limitless that allows us to
26:11
Josh: keep the lights on and do this show forever, that would be very much appreciated.
26:14
Josh: We have gone very long time without interrupting the content with ads.
26:18
Josh: It is time that we're going to need to include just a little bit.
26:20
Josh: So if you know anyone, if you have anyone, if you are someone who is interested
26:24
Josh: in supporting the show, we are all ears. You know how to access us on Twitter.
26:27
Josh: You can message in the comments below, whatever it may be, just reach out and
26:30
Josh: let us know. We appreciate it. But as always, thank you so much for watching
26:33
Josh: and we will see you guys in the next episode.