Latest / In the Loop AI / This Week in AI: China's Lithography Heist // The Lunar Mass Driver // $15 Billion Deals

Transcript

This Week in AI: China's Lithography Heist // The Lunar Mass Driver // $15 Billion Deals · In the Loop AI

0:00
Josh: Imagine if bread kept all of humanity alive and there was only one oven maker.
0:04
Josh: And each one of those ovens cost $200 million and they only make about $40 to $50 a year.
0:10
Josh: And they're all based in the Netherlands.
0:13
Josh: This is the reality of our tech industry. Your iPhone, your laptop,
0:18
Josh: your smart tablets, your TVs, anything with electrical circuits is run by this
0:23
Josh: mysterious company that sits at the top of everybody called ASML.
0:27
Josh: Well, we've essentially built the civilizational Jenga tower.
0:30
Josh: They don't own 99.9%. They own 100% of the lithography market.
0:35
Josh: What is the lithography market? Well, it is kind of upstream of everything else.
0:38
Josh: And I think that's what we're going to start this episode talking about today
0:40
Josh: is this invisible monopoly that no one is really aware of that even exists.
0:45
Ejaaz: Everyone thinks that the most important company in AI is NVIDIA.
0:49
Ejaaz: And if you dig a little deeper, everyone says TSMC.
0:52
Ejaaz: But the truth is, this company, ASML, is the only linchpin.
0:56
Ejaaz: It is the most important company that the entire stock market is currently based
1:01
Ejaaz: on. And they're based in the Netherlands.
1:03
Ejaaz: And they do this really interesting thing, which is called EUV,
1:07
Ejaaz: which stands for Extreme Ultraviolet Lithography.
1:11
Ejaaz: It is a way to basically design very intricate patterns on chips that are used
1:16
Ejaaz: to train frontier AI models and inferencing, and even on your mobile phones
1:20
Ejaaz: as well, in a really specific way.
1:23
Ejaaz: It uses a bunch of laser beams, gases, and stuff like that. And it's an incredibly
1:27
Ejaaz: complex and expensive process.
1:29
Ejaaz: And this hasn't been something that they've been able to crack or any competitor
1:33
Ejaaz: has been able to crack for decades until this week,
1:38
Ejaaz: where it was revealed that some ex-ASML engineers that were hired by the Chinese
1:43
Ejaaz: government have been able to recreate a prototype of this lithography machine that they have.
1:50
Ejaaz: Now, Josh, here's some crazy stats. This single machine costs about $200 million
1:56
Ejaaz: to create and then to maintain even more.
1:59
Ejaaz: The company itself is worth $400 billion and they only have a couple of these machines.
2:05
Ejaaz: So if you do the math, if one of these machines get recreated,
2:08
Ejaaz: you've now taken over $200
2:09
Ejaaz: tens of billions of dollars worth of market cap and potentially even more if
2:13
Ejaaz: you can create AI chips that are of the same level as NVIDIA.
2:18
Ejaaz: And that's what China effectively has right now.
2:19
Josh: It's funny because you hear the $400 billion number and it doesn't even seem
2:23
Josh: that high for how existential this is.
2:25
Josh: Like truly, they are the total linchpin when it comes to creating any sort of electronic device.
2:32
Josh: And it goes, I wonder if that's part of the downstream reason why they haven't
2:35
Josh: been disrupted thus far, because the opportunity isn't in the trillions, but it's coming now.
2:40
Josh: And if it's not from China, it could possibly be coming from the US because
2:42
Josh: a lot of people are coming and trying to tackle this all at once.
2:46
Josh: So this is all amazing stuff. Ejaz, how were they actually able to pull this off?
2:50
Ejaaz: Well, this is the thing. No one really knows. And there's kind of like a gray
2:54
Ejaaz: area where we're like, maybe they stole the secret formula because Josh,
2:58
Ejaaz: this entire machine weighs 450,000 pounds.
3:03
Ejaaz: That is an incredibly heavy machine. And it is very, very complex.
3:07
Ejaaz: And there are tens of thousands of people that are responsible for different
3:11
Ejaaz: parts of this machine, the secret formula, which collectively makes up the makeup of this machine.
3:17
Ejaaz: China somehow have got their hands on it. So there are some rumors that are floating.
3:22
Ejaaz: One tweet here says, my sources on the ground says they stole the source,
3:26
Ejaaz: like they used a bunch of other versions in the past decade.
3:29
Ejaaz: And it goes on to explain how they've probably got moles.
3:33
Ejaaz: And yeah, I mean, like people that they've planted within this company throughout
3:36
Ejaaz: several years that have been able to take these secrets. It sounds very tinfoil hattie.
3:40
Ejaaz: But to kind of out and look at the wider picture.
3:44
Ejaaz: China, if true, has built an effective
3:47
Ejaaz: working prototype of extreme ultraviolet in less than five years.
3:51
Ejaaz: And he compares it to the American venture capital space who have been trying
3:56
Ejaaz: to crack the same problem for over a decade now, Josh, and no one's been able
4:00
Ejaaz: to crack that 100% monopoly.
4:01
Ejaaz: So the fact that we've finally made this breakthrough is awesome.
4:04
Ejaaz: But the fact that it's also a Chinese venture that's been able to do this is kind of bearish.
4:10
Josh: Yeah, well, I would say there is another competitor in the ring here.
4:14
Josh: It's not just China that's trying to take ASML's market.
4:17
Josh: It is also a company called Substrate, which recently raised a bunch of money.
4:21
Josh: They were kind of announcing what they're doing.
4:23
Josh: And their take on it is trying to reinvent.
4:26
Josh: If you think of these fabs as printing presses, they're trying to reinvent the printing press.
4:31
Josh: And their bet is that they are going to take an advanced x-ray lithography,
4:35
Josh: like we mentioned earlier, but turn it into particle accelerators.
4:37
Josh: This is really crazy, technically complicated thing.
4:40
Josh: But I do want everyone to understand that they're working on it too.
4:44
Josh: And it appears as if this monopoly is kind of being broken into two.
4:49
Josh: One, of course, is China. One is the United States. We each have a leading company
4:53
Josh: working on this technology to dismantle the singular monopoly.
4:57
Josh: So who is going to win that race? I'm not sure, but it certainly seems right
5:01
Josh: that you should be diversifying if we are so reliant on this singular,
5:06
Josh: company for every single chip that we make across the board.
5:09
Josh: Like you said earlier, NVIDIA is not the bottleneck.
5:13
Josh: The TSMC is not the bottleneck. It is this weird company in the middle of nowhere
5:16
Josh: that makes these chips. That is the actual bottleneck.
5:20
Ejaaz: It just blows my mind how dependent the entire stock market is, Josh.
5:24
Ejaaz: How your and I's investments are. The world is dependent on these two companies,
5:29
Ejaaz: or in this case, this one singular company.
5:31
Ejaaz: And just to highlight how important of an
5:34
Ejaaz: opportunity this is if a company or country
5:37
Ejaaz: is able to crack the secret formula you have
5:40
Ejaaz: so many other companies that are focused on this samsung in fact has been
5:43
Ejaaz: at this for over a decade with no luck they made some recent announcements this
5:48
Ejaaz: week where they're creating a new fab that is going to focus on building two
5:53
Ejaaz: nanometer cpus remember that's cpu not gpu so they're still about five years
5:58
Ejaaz: behind getting to the same level that this company, ASML, is.
6:01
Ejaaz: So if China has, in fact, cracked this, it would be a miracle.
6:05
Ejaaz: We still don't know how they've done it, but crazy to see.
6:09
Ejaaz: Moving on, Josh, I saw this tweet by Elon Musk this week, and I had no idea about it.
6:16
Ejaaz: And you said, we have to talk about this on the show. Have to.
6:18
Ejaaz: I love it. Walk me through it.
6:20
Josh: Okay, this is called a mass driver. And a mass driver is not even a new idea.
6:25
Josh: This was proposed in the 1970s at MIT.
6:29
Josh: And one thing I found about sci-fi is that oftentimes, I mean,
6:32
Josh: it could be wrong, but when it's right, it's very right. It's just right on the wrong timescale.
6:36
Josh: It seems as if that scale and that time window is finally getting close to where we are now.
6:41
Josh: So on screen you're seeing a visual of this crazy thing called
6:44
Josh: a lunar mass driver it was so fascinating and he does
6:47
Josh: when you were first reading it you were like whoa this is this is real and i think that
6:50
Josh: was the reaction that i had too basically what
6:52
Josh: is it's a giant solar powered electromagnetic catapult
6:56
Josh: on the moon so you can fling cargo into space without
6:59
Josh: rockets or fuel so if you imagine building this like
7:02
Josh: really long track on the moon with superconductor coils you
7:05
Josh: load it with chunks of lunar material and then it uses electricity to
7:08
Josh: yeet them into space at 5300 hundred miles an
7:11
Josh: hour with no rocket engines no fuel it's fascinating
7:14
Josh: and why can we talk about this again all of
7:17
Josh: the most interesting conversations that we're having recently are downstream of
7:20
Josh: the fact that starship has been able to get actual mass
7:23
Josh: to orbit for a low cost or the assumption is
7:26
Josh: that this will happen in the next year so the idea is that you can go put this
7:29
Josh: on the moon why the moon well because it has low gravity it has one-sixth the
7:33
Josh: gravity of earth so you need to spend a lot less energy to throw stuff off there's
7:36
Josh: no atmosphere there's a ton of sunlight for energy and naturally the question
7:40
Josh: I ask myself is, okay, well, why should I care about this? Why does it matter?
7:44
Josh: And the reality is the economics of making this a reality are pretty staggering.
7:49
Ejaaz: Yeah, I was looking into the math before we came on the show,
7:53
Ejaaz: and the Falcon 9, the old rockets from SpaceX, costs about $3,000 per kilogram
8:00
Ejaaz: to get mass into lower Earth.
8:02
Ejaaz: Starship targets $100 per kg, which is a massive, massive reduction.
8:07
Ejaaz: And so if you were to use a lunar mass driver powered by solar and running continuously
8:12
Ejaaz: with zero propellant, well, an early study in 1979 estimated that it would cost $1.
8:18
Ejaaz: So the point here is that the cost of bringing mass back and forth to Earth
8:23
Ejaaz: into low orbit space, it would come down to minimal, would come down to zero, which is just insane.
8:29
Ejaaz: So then the question becomes, well, what does that timeline look like?
8:32
Ejaaz: And the post goes on to describe that we're roughly going to target the mid
8:36
Ejaaz: 2030s to have something like this functioning.
8:39
Ejaaz: So, Josh, this is something that's going to be pretty much realistic and achievable
8:43
Ejaaz: in our lifetimes and not far off when we're old men.
8:48
Ejaaz: Relatively when I was quite young, within like the decade, which is just insane to think about.
8:52
Josh: Yeah, we filmed this episode about SpaceX and their IPO last week,
8:55
Josh: and I would highly recommend watching it because the idea is that SpaceX going
8:59
Josh: public and getting the resources required to go and build an outer space really
9:04
Josh: changes the fabric of reality that we live in today.
9:06
Josh: Like these sci-fi concepts that have been ideated on for the last 60 years almost
9:12
Josh: are able to be put in reality from at least a first principles perspective,
9:16
Josh: where if you're able to get this amount of mass payload to orbit on the moon, ideally,
9:22
Josh: you can turn that into one of these mass drivers that then offsets the global
9:25
Josh: economy in a way that is the thesis for this universal high income,
9:29
Josh: where if you can get this abundance of resources back to Earth cost effectively and efficiently,
9:35
Josh: well, that really disrupts a lot of the financial systems we have today.
9:38
Josh: And it gives us this level of abundance that is almost inconceivable to someone
9:43
Josh: who doesn't read sci-fi on a regular basis.
9:45
Josh: So to me, this is exciting because one, the math checks out and it's really just a matter of
9:50
Josh: making the right set of decisions over the next decade or two to turn this into
9:54
Josh: a reality where we do have a moon base and then eventually a martian base and
9:57
Josh: we have ai supercomputers in space and this world of space is just starting
10:01
Josh: to be unlocked and it's going to unlock so many unbelievable ways that we can take it.
10:06
Ejaaz: I i like your point on uh abundant energy being the key to unlocking future progress josh i was
10:13
Josh: Listening the only thing that matters yeah.
10:15
Ejaaz: I was listening to this interview with Demis Hesibis on the Google DeepMind podcast this morning.
10:20
Ejaaz: And one of his key takeaways was, or one of his learnings rather,
10:23
Josh: Was he plays AI models like a game.
10:27
Ejaaz: And this game is to solve certain sub-routes is what they describe, or what he describes.
10:33
Ejaaz: And he describes a sub-route as something like solving AlphaGo.
10:36
Ejaaz: And if he solves that game, then the AI model is able to solve other games.
10:39
Ejaaz: And he said, he was then asked, what's the most important sub-root to solve
10:44
Ejaaz: right now, Demis? And he answers with one word, energy.
10:47
Ejaaz: If you can create abundant energy, you can create an infinite type of progress and world.
10:53
Ejaaz: And on the topic of Google and on the topic of energy,
10:58
Josh: We have an update.
11:00
Ejaaz: Yeah, bring it back to Earth. There we go. All the puns. We have some new news
11:04
Ejaaz: from a Google-owned company.
11:06
Ejaaz: Waymo, the self-driving car company is in discussions to raise more than $15
11:11
Ejaaz: billion at a valuation near $100 billion, which seems pretty cheap to me,
11:16
Ejaaz: but we'll get into that in a second.
11:18
Ejaaz: But the most important part is being led in a round by its parent company, Alphabet, aka Google.
11:27
Ejaaz: I don't know what to think about this because I thought Google was loaded,
11:31
Josh: Josh. You would think so, right? That a company with hundreds of billions on their balance sheet,
11:36
Josh: who owns the entirety of the company would want to just double down on their
11:40
Josh: investment, but they have not.
11:42
Josh: They are looking to raise another $15 billion.
11:45
Josh: That signals to me perhaps a lack of confidence.
11:49
Josh: Ejaz, do you have a different take? Because to me, I'm like,
11:51
Josh: well, we just recorded an episode about this. Waymo is the clear loser here.
11:55
Josh: Does Google recognize that and are they hedging their bet? Or is there something
11:58
Josh: else going on behind the scenes?
11:59
Ejaaz: Okay, so I'm trying to put my head into the head of Waymo's product division and his pitch deck.
12:07
Ejaaz: What's he saying to justify a $15 billion raise?
12:10
Ejaaz: Well, I'm guessing he's going to say, let's spend $10 billion of these dollars
12:14
Ejaaz: to purchase more Waymo cars so we can expand our network.
12:18
Ejaaz: The issue with that is each Waymo car costs $150,000 versus a Tesla CyberCab,
12:25
Ejaaz: which will cost $30,000, maybe even less, because everyone else that buys a
12:29
Ejaaz: Tesla can effectively add to the RoboTaxi network.
12:31
Ejaaz: So I don't see a justification for the $10 billion.
12:34
Ejaaz: Now, if I put my tinfoil hat on, Josh, I'm thinking the only reason why they're
12:39
Ejaaz: taking external capital is to one, de-risk the venture itself,
12:42
Ejaaz: and two, suck the capital out of competitors.
12:46
Ejaaz: So instead of the venture capital money going to competitors,
12:49
Ejaaz: it would go to Waymo to fund their expansion.
12:53
Ejaaz: I don't really see how this is going to be sustained.
12:56
Ejaaz: And, you know, listen, it's a great 2X for anyone that got involved in the 2024
13:00
Ejaaz: round, but I'm not really kind of impressed by it, I guess.
13:04
Josh: Boring. I mean, Waymo's got a long way to go if they want to compete with Tesla.
13:07
Josh: And that's just the harsh reality of it, where there will still be a business,
13:11
Josh: but is it capable of growing?
13:12
Josh: I do not know. But that is not the only big fundraise we have this week.
13:15
Josh: We also had news in our circular economy section of the show.
13:20
Ejaaz: It's circling. It is circling like a bunch of vultures.
13:23
Josh: Vultures ready to feast on an infinitely expanding bubble that has grown by
13:28
Josh: another $10 billion this week as it relates to OpenAI and Amazon.
13:33
Josh: Ejaz, you have been very bullish on Amazon. So please explain the bull case
13:37
Josh: for this deal, at least. Why are they making this deal?
13:41
Ejaaz: I am going to give you the bull case and then I'm going to give you the slight bear case. Okay.
13:46
Ejaaz: So the bull case here is OpenAI needs a load of compute to train their AI models
13:53
Ejaaz: to compete with the likes of Google Anthropic, which by the way,
13:56
Ejaaz: they are still losing to.
13:58
Ejaaz: No surprises there. If you were to guess or predict in January who would win,
14:02
Ejaaz: you would say OpenAI and now everyone is shocked.
14:05
Ejaaz: So one way to achieve that is, okay, what if I told you this,
14:10
Ejaaz: Josh? What if I told you that you could get the same frontier level intelligence for 50% off?
14:16
Ejaaz: Would you take it or would you leave it?
14:18
Josh: I'm taking that in a heartbeat.
14:20
Ejaaz: Okay, that is effectively what Trinium 3 chips, Amazon's latest AI chips,
14:24
Ejaaz: can afford you if you were to kind of enable that via Amazon. But there's an issue.
14:30
Ejaaz: OpenAI is losing $12 billion per year now. It's probably going to ramp up even more next year.
14:34
Ejaaz: And so they need to figure out a creative way to get their hands on their chips
14:38
Ejaaz: without spending the necessary money.
14:41
Ejaaz: Insert this idea. Amazon invests $10 billion in OpenAI, which they're probably
14:46
Ejaaz: going to spend that money on Trinium 3 chips.
14:49
Ejaaz: So this, which brings me to my bad case, the circular economy of these deals is insane.
14:54
Ejaaz: Why? Because it's going to boost their valuations even more.
14:57
Ejaaz: Josh, I don't know if you saw, but rumors broke this morning that OpenAI is
15:00
Ejaaz: going to get valued at $800 billion and they're raising another $100 billion.
15:03
Ejaaz: So I'm just like, what is going on here?
15:06
Josh: Everyone wants in. And the merry-go-round continues to spin.
15:10
Josh: The musical chairs continue to go around and we will see where everybody stops.
15:15
Ejaaz: Yes. Cue Michael Burry meme. Cue Michael Burry meme.
15:18
Josh: For those not watching, it's Michael Burry, and it's a quote with the Amazon
15:22
Josh: investment would help OpenAI afford commitments,
15:25
Josh: including from Amazon Web Services, which is ironic because Amazon Web Services
15:30
Josh: is just GPUs and cloud servers and more compute.
15:34
Josh: But Ejaz, you did mention a 50 percent improvement in your model.
15:38
Josh: And for users of Google, whether you're a builder or a consumer,
15:42
Josh: your model has just gotten 40% better and 50% cheaper overnight with Gemini 3 Flash.
15:49
Josh: The best model in the world just got better. This one appears to be the best
15:54
Josh: pound for pound model on Earth by a fairly large margin. And you have pulled
15:58
Josh: up on screen the ARK AGI prize chart, which I love to look at because it's just clear.
16:03
Ejaaz: It's a favorite chart, everyone.
16:04
Josh: It's such a good chart because it shows like the Pareto frontier.
16:08
Josh: And we talked about this previously where there is these trade-offs between
16:11
Josh: cost and quality per token.
16:14
Josh: And what you're seeing with that green line that goes straight up is that Gemini
16:18
Josh: 3 Flash is really on the most vertical trajectory out of any of these.
16:22
Josh: So you'll see it falls right in line with GPT 5.2 high, but it costs less than,
16:29
Josh: what is that, a full order of magnitude less per token?
16:34
Ejaaz: Yes, it does. It's crazy cheap.
16:36
Josh: It's crazy. Yeah.
16:38
Ejaaz: I think the cool story about all of this, Josh, is the process of distillation,
16:44
Ejaaz: which is, okay, you spend tens of billions of dollars to train a frontier model.
16:48
Ejaaz: And then what you can do is you can distill all the intelligent components of
16:51
Ejaaz: that model into a cheaper, more, like less costly model, and then scale that
16:56
Ejaaz: out to whatever product that you have. And who has the most product distribution?
16:59
Ejaaz: Google does. So if you can surface that to a bunch of different people,
17:03
Ejaaz: let's say through, what's their number one product?
17:07
Ejaaz: I'm racking my brain. Oh, of course, it's Google search with 5 billion users.
17:12
Ejaaz: That would be an attractive proposition.
17:16
Ejaaz: And guess what? With this update or a story that kind of went under the waves,
17:21
Ejaaz: what we're going to highlight here is this domain called google.ai,
17:26
Ejaaz: which now effectively embeds Gemini 3 Flash
17:31
Ejaaz: into Google search. Now, there's a few steps involved here.
17:35
Ejaaz: But the simple fact that you can now go to Google search accessible to 5 billion
17:39
Ejaaz: different users and get access to AI mode with frontier level intelligence at
17:45
Ejaaz: a fraction of the cost, I just don't see how Google loses this at all.
17:49
Ejaaz: They have the moat, they have the distribution. We've said this multiple times
17:52
Ejaaz: before. We've done a Google bull case episode. You definitely go check that out.
17:56
Ejaaz: But now they've combined the frontier model with this josh like this is like
18:00
Ejaaz: they're like cannibalizing themselves in a way
18:02
Josh: Yeah we're starting to get the answer to what do you do in
18:05
Josh: the case of the founder's dilemma where what happens when innovation eats
18:08
Josh: your core product and that's what's happening and google has decided to double
18:12
Josh: down and fully lean into it and we are going to see how that plays out i mean
18:17
Josh: this is google.ai this is not google.com so their core search product is still
18:21
Josh: intact but i imagine and this is a transitionary period where they are testing,
18:26
Josh: there is currently the AI mode button in Google.com's homepage.
18:30
Josh: And it's only a matter of time until they're able to, until they commit to fully switching over.
18:35
Josh: And then you have to ask the question, well, how are they going to support this business?
18:39
Josh: And is it just going to be an effort to bleed out the results from all these other companies?
18:44
Josh: Because OpenAI needs to make revenue.
18:46
Josh: They must make a ton of money. And in order to do that, they're going to need
18:49
Josh: to start implementing advertising and sponsorships into their products.
18:53
Josh: Google has a very different business model. They are a behemoth.
18:56
Josh: They have a tremendous amount of cash on the balance sheet that they can use
18:59
Josh: to subsidize all of this compute and all of these AI needs that people need.
19:04
Josh: So they have TPUs that are hyper-efficient. They have the best,
19:07
Josh: most efficient model in the world.
19:08
Josh: And they have no need to monetize urgently because they have such a huge cash balance sheet.
19:14
Josh: So OpenAI and other companies that need to compete on this run are going to
19:17
Josh: start to have to make some trade-offs here. Yeah.
19:21
Ejaaz: You might be thinking as well, why doesn't Google just switch search completely to AI if it's so smart?
19:29
Ejaaz: And the simple answer is they haven't been able to embed advertising and search
19:34
Ejaaz: rankings into the model effectively yet.
19:36
Ejaaz: So the same kind of experience where you see sponsored posts at the top of your
19:39
Ejaaz: Google search results page, you can't quite do effectively with AI models.
19:43
Ejaaz: It's a completely different beast.
19:44
Ejaaz: And both companies, Google and OpenAI, will need to figure this out eventually.
19:48
Ejaaz: I think all roads lead to ads. But the question is, at what time does that make sense?
19:54
Ejaaz: If Google can hold out to your point and OpenAI kind of comes in first,
19:58
Ejaaz: Google wins the moat because people just want to use an ad-free product. It is crazy.
20:03
Ejaaz: But Josh, on the topic of OpenAI, did you see this news?
20:07
Ejaaz: They've introduced...
20:09
Ejaaz: Gpt chat gpt image 1.5 which is their answer to google's nano banana pro
20:17
Josh: And as everybody knows we myself i'll
20:20
Josh: speak for myself absolutely adore nano banana pro it is one of my favorite models
20:23
Josh: of the year it is an excellent image generator and as i was looking through
20:27
Josh: this announcement i was hopeful that this would be a better replacement it's
20:31
Josh: built into the chat gpt app that we know and love and i think the reality of
20:35
Josh: it is it is just not that good i mean in the video So if you scroll down just a little bit.
20:40
Ejaaz: Or we start right here
20:41
Josh: With this example. So for those who are watching, close your eyes.
20:46
Josh: If you're not, what we're looking at on screen, and I don't even know if you
20:49
Josh: should scroll down any further than this.
20:51
Ejaaz: Is a post from Mr.
20:53
Josh: Sam Altman himself, shirtless as a firefighter in front of a Christmas tree.
20:58
Josh: And if you scroll down even further, it's actually a holiday calendar.
21:01
Josh: And it's an example to reflect the quality, I guess, of the new 1.5 image model. It's good.
21:08
Josh: It's not great it would appear as if nano banana is better after
21:10
Josh: going through a bunch of examples nano banana pro still is good but what's
21:13
Josh: cool about this is if you don't use google if you're not familiar with their
21:17
Josh: uh work flow if you don't use gemini if you only use chat gpt you have a pretty
21:23
Josh: good imaging model because this is built right into the app and what you'll
21:25
Josh: notice it you guys is it's really good at faces so we're seeing on screen a
21:29
Josh: picture of sam and ilia kind of a little sad at a party uh,
21:33
Josh: it's pretty good it shows faces in a world where they didn't show faces before
21:37
Josh: and they're fairly accurate so it's an attempt it is by no means a smackdown
21:42
Josh: on the king nano banana pro but hey it's pretty cool and if you use chat gpt
21:46
Josh: your image gen model just got much better listen.
21:49
Ejaaz: Call me a google bowl but like i don't think it's that impressive like we're
21:53
Ejaaz: looking at 1.5 gpt 1.5 on the left it just looks so much grainier like like
21:59
Ejaaz: the example on the ride by nano banana looks way more accurate.
22:03
Ejaaz: Like you can see the physical expressions on Ilya's face as well.
22:06
Ejaaz: It looks more like Ilya as this post claims.
22:10
Ejaaz: I don't know if this is a zinger. And what's important about this is Sam initiated
22:15
Ejaaz: Code Red a week and a half ago.
22:18
Ejaaz: And since then, he's done something really impressive, which is divert all resources
22:22
Ejaaz: towards number one, creating the best model, and number two,
22:25
Ejaaz: creating the best image model.
22:26
Ejaaz: Now, he succeeded on number one, releasing GPT 5.2
22:29
Ejaaz: last week which broke all frontier benchmarks again surprise surprise um but
22:33
Ejaaz: this image model he's kind of flopped so i i don't know if we've reached a point
22:38
Ejaaz: where open ai still has the edge that they that they uh once had at the start
22:42
Ejaaz: of this year um i i don't know it still leaves me uncomfortable i don't believe
22:46
Ejaaz: that open ai has kind of nailed it just yet yeah
22:49
Josh: They haven't but if you are again if you're a user go try it out it's better
22:52
Josh: than it was before and if you're being too stubborn to go use gemini well this
22:56
Josh: is the best image model on the block for you uh now i guess we can head over
22:59
Josh: to a little more frontier science-y stuff.
23:04
Ejaaz: Josh, how am I going to live forever?
23:07
Josh: Okay, perhaps...
23:09
Josh: I cannot tell you the answer to you living forever, but I can give you an answer
23:12
Josh: to your organs surviving longer if
23:15
Josh: they ever leave your body for whatever circumstances that might require.
23:19
Josh: There is a brain-computer interface startup called Science Corporation,
23:22
Josh: which we love, founded by Max Hodak, who is the former Neuralink co-founder and president.
23:27
Josh: And they decided this week that they were going to announce they're moving into
23:31
Josh: organ preservation and life support tech.
23:33
Josh: Basically, it's like the hardcore biomedical engineering mindset used for brain
23:37
Josh: implants is now being pointed at a different problem, which is keeping organs
23:41
Josh: alive outside of the body for way longer.
23:43
Josh: And I had to learn a lot before talking about this because right now,
23:46
Josh: I didn't realize that organ transplants are in a pretty brutal race against the clock.
23:50
Josh: Normally, they last like a day, sometimes less if you have lungs or kidneys.
23:55
Josh: It's very expensive to transfer these from place to place. A lot of time it
23:58
Josh: was done via like private jet or really like whatever fastest way you can get
24:03
Josh: it from one place to another because the person receiving it is normally in
24:06
Josh: very bad shape and the person giving it,
24:08
Josh: the organs are just not capable of living that long outside of it.
24:11
Josh: So what they shared today is this breakthrough in how they're able to use this
24:16
Josh: fancy new device to actually preserve organs for longer to make them last for
24:20
Josh: days and hopefully soon weeks instead of just hours.
24:23
Josh: And i found it to be really fascinating just on the the
24:26
Josh: scientific frontier front because again they're using ai in
24:30
Josh: a industry that has not really used ai very much in the past to create these
24:35
Josh: really cool new contraptions and you could kind of see an idea of what the photo
24:38
Josh: looks like in this post here that can actually help save people's lives and
24:43
Josh: make a big difference i mean traditionally it cost 250 000 for a machine that
24:47
Josh: does this plus tens of thousands of dollars per use.
24:49
Josh: And what they're planning to do is get that cost down to $10,000 and even less.
24:53
Josh: And I think it's just a really fascinating exploration into what's possible
24:56
Josh: in the world of science now that we're applying AI to a lot of these difficult problems.
25:01
Ejaaz: Josh, to give you a kind of crazy example to demonstrate how important this
25:05
Ejaaz: thing is, in a past life, I studied biology at university.
25:10
Josh: That's right. You're a pro on this.
25:11
Ejaaz: Well, I had a crazy opportunity to perform heart transplants between murine mice.
25:19
Ejaaz: So these are like test lab mice. It was a pretty crazy opportunity.
25:22
Ejaaz: I signed up for it, somehow got approved because I was studying biology at the time.
25:27
Ejaaz: I love genetics and specifically regeneration of cells. So I qualified.
25:31
Ejaaz: And the TLDR is I performed four of these things and it failed 75% of the time.
25:40
Ejaaz: The craziest part, the mouse I was transplanting the heart into was right next to me.
25:46
Ejaaz: It was literally five inches away. And the fact that you had that failure rate
25:49
Ejaaz: just goes to show you how crazily complex it is to keep these organs alive and
25:55
Ejaaz: sustainable. Just a fun anecdote.
25:57
Josh: It's really challenging. If you scroll down a little bit, we could kind of see
25:59
Josh: what the device looks like in full.
26:02
Josh: It's interesting. The thesis behind this was that, like, why can't you ship
26:06
Josh: organs long distance like luggage would be?
26:09
Josh: And what they wanted to do is create this backpack size system.
26:11
Josh: So their prototype that we see on screen, It has integrated sensors for oxygenation
26:16
Josh: and flow pressure or temperature.
26:17
Josh: And it's this closed loop system that automatically adjusts depending on its outside environment.
26:23
Josh: And it's really cool because traditionally it required something much larger
26:26
Josh: than this. That was much more expensive. That was single use.
26:28
Josh: That didn't allow these to survive for much longer.
26:30
Josh: So just an interesting thing worth checking out. We'll add a link to it in the
26:33
Josh: description if you want to read more.
26:34
Josh: Just something worth paying attention to that health and sciences,
26:38
Josh: thanks to AI, are becoming a lot more exciting.
26:40
Ejaaz: It has been a crazy week for science in general, or rather sci-fi.
26:46
Ejaaz: Sci-fi has had a big win this week.
26:49
Ejaaz: Yeah, yeah. We've got the mass drivers. Elon's launching satellites to train
26:52
Ejaaz: AIs. We're harnessing the energy of the sun.
26:54
Ejaaz: We have got this crazy device that can extend the lifespan of organs.
26:58
Ejaaz: And on top of it all, down on Earth, we have one singular company that has been
27:03
Ejaaz: discovered, which is upholding the entire world right now.
27:06
Ejaaz: Pretty pretty insane um that is the end of another roundup on the limit another
27:11
Ejaaz: big week another another huge week huge week we have some
27:15
Josh: Good weeks coming up too.
27:16
Ejaaz: We have some great weeks coming up uh as is
27:18
Ejaaz: normal uh of the end of the year we've we've got to
27:21
Ejaaz: do some reflection josh you and i've been doing a bit of reflection recently
27:24
Ejaaz: um and we have filmed a banger of an
27:27
Ejaaz: episode which will tell you about the top wins
27:31
Ejaaz: and most importantly the biggest losses in ai
27:34
Ejaaz: this here and then of course you know
27:37
Ejaaz: josh and i aren't short-term kind of thinking types of people we like to put
27:41
Ejaaz: in predictions as well and we have a lot of big and bold predictions which i'm
27:45
Ejaaz: pretty sure 99.9 of you guys won't guess at all uh for 2026 that's coming out
27:51
Ejaaz: next week as well uh we're still going to be shipping episodes uh during the the holiday week
27:56
Josh: So listen you can do your holidays but you still need to tune in you still owe
27:59
Josh: us at least 65 minutes per week to listen to these episodes um.
28:03
Ejaaz: Josh do you know what i am going to ask of people's new year's resolutions
28:08
Josh: Um i'm going to hope it's to be to share limitless podcast with at least a hundred
28:13
Josh: of your closest friends each i.
28:15
Ejaaz: Am going to say no more josh read my mind
28:18
Josh: There you have it uh it means a lot it's been it's been
28:20
Josh: a good year we have some fun and different and interesting content like
28:23
Josh: ijaz was mentioning earlier that's coming in the coming week so
28:26
Josh: you're definitely going to want to tune in for that it's fun it's different ijaz actually
28:29
Josh: wore a tuxedo in one of them so i promise you you're going to want to
28:32
Josh: watch this to see why but it's it's
28:35
Josh: really been fun it's been a great journey and it's all thanks to the people who are so
28:38
Josh: generously leaving comments and sharing with their friends and we read almost
28:42
Josh: almost everyone i try to there's been a lot which is a really good problem to
28:45
Josh: have but trying to keep up with everything um the support really means the world
28:49
Josh: so if you did enjoy this please share it with someone rate the episode five
28:52
Josh: stars and thank you as always for watching and we'll see you guys in the next one.