This Startup Fused Human Brain Cells with Silicon Chips | E2295

Summary

Jason and Lon host Cortical Labs founder Hon Weng Chong, returning to the show three years after his 2022 appearance to demo the CL1 — the world’s first biological computer for sale. He brings one to the studio: a 3U rack-mounted “space-age toaster” that houses lab-grown neurons (derived from stem cells, not extracted from people), with a “lung” for gas exchange and a “kidney” filtration cartridge that gets swapped every 4-6 months. Hon explains why neurons are 5,000x more efficient than GPUs at reinforcement learning, walks through the new Cortical Cloud (corticallabs.com) that’s hosting 120 deployed units in Melbourne with Singapore next, and explains the chiplet-style PDMS microfluidics that segment thousands of neurons so multiple users can share one device. The conversation gets deep into the bioethics — Hon explains the company’s red line: don’t create conscious systems, because conscious systems can suffer. Cortical Labs is raising $30M.

Second guest is Michael Norcia of Pyka, who walks through the company’s lineup of large-scale electric agricultural drones (the Pelican) and the new hybrid turbo-diesel Dropship UAV. Pyka started at Y Combinator with first flight a week before Demo Day, and is now a “Group 4 UAV” company — meaning aircraft over 1,320 pounds, putting them in MQ-9 Reaper territory. Pelican drones operate in Brazil thanks to deregulated agricultural drone rules, doing crop spraying 5-15 km out of line of sight at 2 gallons of diesel per hour vs. 50 for the Air Tractor it replaces. The new Dropship is a dual-use product: commercial logistics and defense, with an architecture Michael claims no other airplane has — parallel hybrid turbo-diesel with electric props for ballistic takeoff/landing then shut off in cruise. Pyka went CAD-to-first-flight on Dropship in 180 days.

Highlights

Neurons Are 5,000x More Efficient Than GPUs at Reinforcement Learning

“The world’s first biological data center. When they compared it against their reinforcement learning systems, the neurons we had were 5,000 times more stable efficient.” — Hon Weng Chong, 0:04

Clip command

yt-dlp --download-sections "*00:04-00:30" "https://www.youtube.com/watch?v=wJH_ymNbv30" --force-keyframes-at-cuts --merge-output-format mp4 -o "neurons-5000x-efficient.mp4"

The Vatican Worried About Conscious Computing

“The Vatican were worried about this. You do not want to create conscious systems because ethically a conscious system has the ability to suffer and we do not want any suffering to come about from any technology.” — Hon Weng Chong, 0:17

Clip command

yt-dlp --download-sections "*00:17-01:13" "https://www.youtube.com/watch?v=wJH_ymNbv30" --force-keyframes-at-cuts --merge-output-format mp4 -o "vatican-conscious-systems.mp4"

The CL1 Has Lungs, a Kidney, and Sugar Water

“I’m just laughing because we’re talking about, you know, rack mounting and how much space it takes up in a server rack, and then you’re like, ‘And here are the lungs, and here’s where it expels waste.’” — Jason Calacanis, 5:14

Clip command

yt-dlp --download-sections "*05:14-06:00" "https://www.youtube.com/watch?v=wJH_ymNbv30" --force-keyframes-at-cuts --merge-output-format mp4 -o "cl1-lungs-kidney.mp4"

The Red Line: No Conscious Systems

“I think internally at the company this is really important to understand this whole discussion of consciousness because I think we’ve drawn a red line for us which you do not cross… You do not want to create conscious systems because ethically a conscious system has the ability to suffer.” — Hon Weng Chong, 22:27

Clip command

yt-dlp --download-sections "*22:14-23:32" "https://www.youtube.com/watch?v=wJH_ymNbv30" --force-keyframes-at-cuts --merge-output-format mp4 -o "cortical-red-line.mp4"

Pelican: 2 Gallons/Hour vs Air Tractor’s 50

“The competition for our aircraft is a vehicle called the Air Tractor. It’s a big, you know, roughly 8,000-pound vehicle. It burns roughly 50 gallons per hour… Our Pelican in the worst case, if you’re charging it off of a diesel generator, is about two gallons per hour.” — Michael Norcia, 46:55

Clip command

yt-dlp --download-sections "*46:55-47:25" "https://www.youtube.com/watch?v=wJH_ymNbv30" --force-keyframes-at-cuts --merge-output-format mp4 -o "pelican-vs-air-tractor.mp4"

”We could deliver bombs but that’s not the most interesting use case”

“Could it carry bombs in the little container? Sorry to be rude, but I’m curious.” (Jason) — “I mean, technically it could. That’s not the most, like, interesting use case for it, I would say.” (Michael) — Michael Norcia, 52:18

Clip command

yt-dlp --download-sections "*52:14-53:14" "https://www.youtube.com/watch?v=wJH_ymNbv30" --force-keyframes-at-cuts --merge-output-format mp4 -o "dropship-bombs-question.mp4"

Key Points

CL1 stock sold out (1:45) - All 30 units sold at ~$30K each
Now also Chief Janitor Officer (1:55) - Hon doubling as cleaner for US lab
CL1 form factor (3:11) - 3U rack-mountable; resembles GPU sleds
Hardware tour: neural chamber, gas inputs, USB-C (3:54) - Lungs and kidneys for biological compute
Up to 2M neurons per CL1 (5:40) - Organoid scale; 200K is the sweet spot
Cockroaches are smart (7:21) - Hon’s case for why neuron count isn’t everything
Reinforcement learning sweet spot (10:16) - Where biological computing already beats CPUs/GPUs
Robots can’t accelerate time (11:29) - Physical world constraint for embodied RL
Day1 Data Center partnership (12:10) - 120 units in Melbourne, Singapore next
No supply chain constraint (13:46) - Neurons grown locally, decouples from central vendor
Sugar water diet (15:08) - What feeds the neurons
Algorithms still the bottleneck (16:10) - Not hardware
Doom on a CL1 (26:09) - From Pong (2022) to Doom 3 (2026)
Disordered stimulus as punishment (27:41) - Why “no conscious systems” matters morally
120 deployed, 20-30 active users (29:34) - Biology teething issues
Microfluidic chiplets (30:48) - PDMS-based segmentation for multi-tenant biological compute
Cortical Cloud live demo (33:46) - Live neuron activity from Melbourne lab
Raising $30M (33:33) - Current funding round
Hon’s wish list: hire devs (37:06) - Looking for talented developers
Pyka started with paper airplanes age 4 (38:49) - Lifelong aviation obsession
First flight a week before YC Demo Day (39:54) - 600-pound aircraft built in 1 month
Group 4 UAV definition (41:30) - Drones over 1,320 lbs, MQ-9 Reaper class
Brazil’s deregulated ag drones (44:30) - Two years ago Brazil opened up beyond-line-of-sight ag operations
US line-of-sight restrictions (45:00) - Why Pyka couldn’t operate same way in US
Pelican fuel efficiency (46:55) - 2 gallons/hr vs Air Tractor’s 50
24-hour operations possible (48:42) - Customers already doing 12 hours/day
Dropship dual-use (51:43) - Commercial logistics + defense
180 days CAD to first flight (53:22) - For Dropship; faster than Big Bird
Hybrid turbo-diesel architecture (56:00) - Unique aircraft architecture, parallel hybrid
Build everything in-house (60:00) - Pyka vertically integrated like Apple, Tesla, DJI
US BMS sourcing 2x cost (1:00:41) - Order of magnitude US vs China for owned-design components

Mentions

Companies

Cortical Labs (0:28) - Hon Weng Chong’s biological computing company
Day1 Data Center (12:10) - Partner for biological data center deployment
Neuralink (16:25) - Mentioned for BCI side of the field
Synchron (16:25) - Another BCI company referenced
NVIDIA (24:00) - Reference point for CUDA democratization strategy
Mass General (21:15) - Alzheimer’s/dementia research customer
Johns Hopkins (21:15) - Research customer
Pyka (37:56) - Michael Norcia’s drone company
Y Combinator (39:54) - Pyka went through YC
Zipline (45:06) - Mentioned for beyond-line-of-sight regulation similarities
Joby Aviation (implicit) - Defense tech comparator
DJI (1:03:12) - Hardware/software integration example
Apple (1:03:10) - Hardware/software integration example
Tesla (1:03:11) - Hardware/software integration example
Boeing (1:04:10) - Cautionary tale of outsourcing manufacturing
SpaceX (59:01) - Reference: blew up rockets, figured it out
Air Tractor (46:55) - 8,000-pound competitor, 50 gallons/hr
Deel (9:11) - Sponsor read
Quo (formerly OpenPhone) (19:07) - Sponsor read
LinkedIn Jobs (28:34) - Sponsor read

Products & Technologies

CL1 (2:22) - Cortical Labs’ biological computer
Cortical Cloud (23:47) - Hosted access to biological compute
PDMS microfluidics (30:48) - Polydimethylsiloxane chiplet substrates
CUDA (24:43) - NVIDIA’s developer enablement as model
Doom 3 (26:09) - Played by biological computer
Pong (26:09) - Original 2022 demo
Pelican (47:11) - Pyka’s electric ag drone
Dropship (51:43) - Pyka’s hybrid dual-use UAV
Big Bird (53:31) - Pyka’s original YC plane
MQ-9 Reaper (41:30) - Most common Group 4 UAV

People

Hon Weng Chong (0:04) - Founder/CEO of Cortical Labs
Michael Norcia (37:56) - Founder/CEO of Pyka
Brent (CSO) (18:31) - Cortical Labs CSO engaging bioethicists
Jensen Huang (24:43) - NVIDIA founder; CUDA-free strategy reference
Steve Ballmer (37:44) - “Developers, developers, developers” reference

Surprising Quotes

“The Vatican were worried about this.” — Hon Weng Chong, 0:17

“Actually, so cockroaches are actually pretty smart.” — Hon Weng Chong, 7:21

“It’s really good it’s not conscious, otherwise that would be kind of rude like ‘I was sleeping!’” — Michael Norcia (incorrectly attributed in JSON; context is Jason in Cortical segment), 35:18

“If a VC’s listening to this, hello! Come on! Stop backing SaaS companies that are slapping on an AI wrapper. Back something cool. We’re building drones, y’all.” — Lon Harris, 1:04:37

“If you own the design and manufacturing in the US versus China is order of magnitude 2x difference.” — Michael Norcia, 1:01:16

Transcript

Jason Calacanis: 0:00 If you ever heard terms like MVP, minimum viable product, or lean startup…

Hon Weng Chong: 0:04 The world’s first biological data center. When they compared it against their reinforcement learning systems, the neurons we had were 5,000 times more stable efficient.

Lon Harris: 0:13 Has anyone complained that you’re tinkering a bit with the edges of humanity?

Hon Weng Chong: 0:17 The Vatican were worried about this. You do not want to create conscious systems because ethically a conscious system has the ability to suffer and we do not want any suffering to come about from any technology.

Lon Harris: 0:28 Hello and welcome back to TWiST. Now today we’re talking to a company we spoke to in 2022 because I thought they were one of the most interesting startups in the entire world. Called Cortical Labs.

Hon Weng Chong: 1:13 Good, thanks. It’s great to chat again, Alex. And congratulations, I heard you had a new baby.

Lon Harris: 1:21 Yes, that’s why I’ve been extra tired these last four months, but we’re powering through the just the grace of coffee. All right, so Hon, last time we talked, you had just put out your CL1.

Hon Weng Chong: 1:45 We’ve kind of exhausted our entire stock of 30 units that we had capped. So that’s good and you can work out how much that ended up becoming.

Lon Harris: 1:54 Best part of a million.

Hon Weng Chong: 1:55 Yeah, and, uh, we have actually, so I’m in the US right now partly because A, we’re fundraising, but at the same time, I’ve also, you know, CEO stands for Chief Everything Officer, right? I’m also now Chief Janitor Officer.

Lon Harris: 2:22 So tell folks what the CL1 is and maybe because you have one with you, show us a little bit for folks on the video version what it looks like.

Hon Weng Chong: 2:34 Yeah, absolutely. So maybe before I show it, like to give a very brief summary, the CL1 is our attempt to build a computing platform that allows researchers and developers to get going with biological computing.

Hon Weng Chong: 3:00 Bit jerky and this is a CL1. Actually, maybe from the front here you can see what they look like from the front.

Jason Calacanis: 3:07 It looks like a very long space-age toaster.

Hon Weng Chong: 3:11 Yes, so they’re actually very similar to rack-mountable sleds that some of the GPUs come in the same form factor. So this is 3U, so in a server rack cabinet, this will take up about three rows.

Hon Weng Chong: 3:54 But I’m just going to show you what it actually looks like. So we open up the top, this is the neural chamber. So this is where we load up the compute unit. So neurons go into this chip.

Hon Weng Chong: 4:49 And also the really interesting thing is you also have your traditional and non-traditional IO units. So we look at the back here, this is USB-C, USB-A, Ethernet, but you also have gas inputs.

Jason Calacanis: 5:14 I’m just laughing because we’re talking about, you know, rack mounting and how much space it takes up in a server rack, and then you’re like, ‘And here are the lungs, and here’s where it expels waste.’

Jason Calacanis: 5:30 Now, how many neurons can you pack into a CL1, and then in compute-friendly terms, how much power is that that you bring to bear?

Hon Weng Chong: 5:40 Yeah, so in a CL1, you can go all the way up to a million, two million neurons if you so wish. People grow organoids on these things and they have several million neurons.

Hon Weng Chong: 6:00 Quite well, makes it commercially viable, but also we’ve found that you can actually get some learning and training with that.

Lon Harris: 6:08 So, I want to make an analogy here. Because I think that everyone’s very familiar with the idea of parameters in an LLM.

Hon Weng Chong: 6:38 Yeah, I can’t remember exactly the number, but I think it’s like 100 billion neurons that you have in your brain. So, you know, we have that, and then a ton more, like trillions of synapses.

Lon Harris: 6:59 So, that actually doesn’t — tell me why I’m wrong here, Hon, but, cockroach is not famous for their intelligence. Famous for their durability.

Hon Weng Chong: 7:21 Yep. Actually, so cockroaches are actually pretty smart. Along with apologies with bees and flies.

Lon Harris: 7:43 So hard! They’re so quick.

Hon Weng Chong: 7:45 They’re so quick, and they’re so agile, they almost like predict your actions ahead of time.

Jason Calacanis: 8:41 So, when I think about biological computing, I presume we’re talking about a lot of use cases in drug testing and drug discovery, biosimilars, all the stuff that’s known…

Jason Calacanis: 9:00 Silicon and neurons together in a way that creates a computer that is better than today’s GPUs and TPUs and so forth at certain types of calculations that we use in the technology world.

Jason Calacanis: 9:11 [Deel sponsor read]

Hon Weng Chong: 10:16 Yeah, actually there is one that is already like proven to be better than CPUs or GPUs and that’s in reinforcement learning. So this is some very novel work, it’s still getting written up so I don’t really want to share too much.

Hon Weng Chong: 11:29 The caveat with that is you can’t accelerate time if you’re a robot. You’re operating at the same speed like everyone else in the physical world.

Jason Calacanis: 11:41 So if biological computing is good at reinforcement learning, which as I think everyone listening to this knows is an enormous part of improving AI models today, you could end up pretty far outside of the obvious.

Jason Calacanis: 12:00 What you guys have built. So you’ve built a biological data center, you have 120 units in Melbourne, and you’re going to do one in Singapore next, which I believe can get even bigger?

Hon Weng Chong: 12:10 Yes, exactly. So we’re working with a data center company called Day1. They’ve partnered with us because they liked the technology.

Jason Calacanis: 12:51 Which is basically zero for this conversation.

Hon Weng Chong: 12:54 Correct. So they’re getting more for not much in terms of costs.

Jason Calacanis: 13:19 Oh, so you don’t have to ship in your stem cell-based, I point out we’re not killing people here to steal their brains, stem cell-based neurons, you can just make them, grow them, no?

Hon Weng Chong: 13:31 Raise them? Make them, grow them, same thing, yeah.

Jason Calacanis: 13:35 Okay. I’m sorry, it’s my usual words don’t work as well when we translate them over to here.

Hon Weng Chong: 13:46 A tremendous amount. But it also means that there is no supply chain constraint. It decouples your data center from not just being a place where you just buy from a central vendor.

Jason Calacanis: 14:17 Yeah. Remind me how long the neurons live before they need to be replaced?

Hon Weng Chong: 14:24 Neurons can actually live a very long time if you keep them well-kept. The thing that does require replenishing are the tube sets.

Jason Calacanis: 14:41 If you’re on the audio version, he’s pointing at—

Hon Weng Chong: 14:44 I’m pointing at these cartridges here. Think of them a bit like kidneys.

Hon Weng Chong: 15:00 In another four to six months.

Lon Harris: 15:02 Okay, so really then, it’s, what do you feed the neurons? I was about to say sugar water as a joke, but I think it might actually be a little bit right.

Hon Weng Chong: 15:08 No, it actually is pretty much sugar water.

Lon Harris: 15:10 Okay. So basically, you’ve made a biological system that keeps a small number, compared to our brains, of neurons alive that are connected to these chips.

Lon Harris: 15:23 Here’s my question though. As time goes on, do you think that as chips get smarter, we’re going to need more neurons to interface with them?

Hon Weng Chong: 15:43 Yeah, so it’s always a case of like push and pull, right? So for instance, if let’s say we referred back to GPUs.

Hon Weng Chong: 16:10 And so there’s always this push and pull tension between, are the algorithms there yet? Or is it a hardware limitation? Right now, we see this as an algorithm’s limitation.

Hon Weng Chong: 16:25 So that I think is still being worked on. You know, you have really smart people in Neuralink and Synchron working on that for the BCI side.

Hon Weng Chong: 16:35 We’re kind of working on it as well, but we have an additional challenge, which we have to write information directly into the neurons.

Lon Harris: 16:54 I’m just really excited about what you’re working on, because when we think about the systems we’re using to make artificial intelligence smarter, we’re throwing the equivalent of bodies at it.

Lon Harris: 17:13 But we’re working on systems that are so fundamentally much more power-hungry, less efficient, and less attuned to the problems we’re trying to solve than our brains.

Lon Harris: 17:38 And this is when religion comes into it.

Lon Harris: 17:40 So as background, Hon, I was raised in a very conservative Christian church. And so growing up in the 90s, I heard a lot about stem cells and cloning.

Jason Calacanis: 18:00 But I’m a little worried about how we’re going to tell people that we’re fusing neurons with computers because I think you’ve now taken this from proof of concept when you guys played Pong to early commercial.

Hon Weng Chong: 18:31 Actually, the Vatican were worried about this, but fortunately, Brent, my CSO, has done an excellent job engaging with bioethicists.

Jason Calacanis: 19:07 [Quo sponsor read]

Hon Weng Chong: 20:11 And I think the most important thing that we’re trying to work on right now is to get everybody in this space plus the consciousness space to come together.

Hon Weng Chong: 21:00 The principle of double doctrine is there actually a net good that can be come from this technology versus a net negative.

Hon Weng Chong: 21:15 That is the primary use case for all these labs purchasing these devices. In Mass General, it’s an Alzheimer’s-dementia researcher, at Hopkins, they’re looking at…

Jason Calacanis: 21:30 Oh, that’s nice. These are all by the way, this is why this technology even in its current form is freaking awesome. You just listed three major medical centers in the US that are using this technology.

Jason Calacanis: 21:48 So like the dangers that are hypothetical down the road that we’re touching on are not meant to undercut the current usefulness.

Hon Weng Chong: 22:00 Yeah, thank you. The compute side of it is by all means the smallest but the newest, the one that’s also most fraught with risk.

Hon Weng Chong: 22:14 And I think internally at the company this is really important to understand this whole discussion of consciousness because I think we’ve drawn a red line for us which you do not cross.

Hon Weng Chong: 22:27 You do not want to create conscious systems because ethically a conscious system has the ability to suffer and we do not want any suffering to come about from any technology.

Hon Weng Chong: 22:39 We try to do as much as we can when we have the cloud, we can monitor things, but once these things go out into the real world, we don’t really know what’s going on.

Jason Calacanis: 23:01 But the cloud that you guys have built, the data center in Melbourne, and then Singapore, you know, we’re going to have if it’s 200,000 neurons a piece and we’re going to scale this out…

Jason Calacanis: 23:18 The reason why it doesn’t scare me is it’s all kind of like cut up into little pieces. Each one has X number of neurons, but there is the science fiction question.

Jason Calacanis: 23:32 But the good news is that I think we’re pretty far away from having so many biological computers in the world all plugged in together.

Jason Calacanis: 23:47 Why the cloud? I’m just curious about why that was the right commercial approach to the market.

Hon Weng Chong: 23:51 Ultimately what we really want to do at Cortical Labs is to, hate the word democratize, but it is democratize the technology by reducing the accessibility barrier.

Hon Weng Chong: 24:00 Ultimately, I always go back to NVIDIA. The AI that we have today is actually an accident. It was completely a fluke that it happened.

Lon Harris: 24:33 I like to think we would have had some other serendipitous meeting point along the way.

Hon Weng Chong: 24:43 Exactly. What Jensen did brilliantly at the start was he made CUDA free. He made it so that any GPU, even the crappiest gaming GPU, could also run.

Lon Harris: 25:13 Right, so you need to have a person who’s doing the kidney replacement and the feeding and the waste extraction.

Hon Weng Chong: 25:19 Exactly, exactly.

Lon Harris: 25:29 Got it.

Hon Weng Chong: 25:32 But what if you don’t have any of that, but you have a great idea you want to experiment with? So the idea of the cloud was born.

Lon Harris: 25:58 Could it be called, Kai-Combinator? Or rhyme with that?

Hon Weng Chong: 26:08 Can neither confirm nor deny.

Lon Harris: 26:09 Got it. So last time we talked, Hon, we were talking about the first application of your technology to a video game which was Pong.

Hon Weng Chong: 26:39 Correct, yep.

Lon Harris: 26:40 In Pong, not a lot of variables. In Doom, many more.

Hon Weng Chong: 26:54 Yeah. So pretty much it’s also the same. They’re using auditory stimulus versus…

Hon Weng Chong: 27:00 Disordered stimulus as the reward and punishment signal.

Jason Calacanis: 27:41 I just realized why it’s very important that you don’t create conscious systems because they can feel pain, because if you’re using disordered inputs as a punishment mechanism, it’s a little bit harsh.

Hon Weng Chong: 27:57 Correct, exactly.

Jason Calacanis: 27:59 Yeah, okay. That makes sense. Doom’s great. If you haven’t played Doom 3, do it.

Jason Calacanis: 28:34 [LinkedIn Jobs sponsor read]

Hon Weng Chong: 29:34 So, we have 120 that we’ve deployed. Unfortunately, there has been some teething with the biology, so while we do have 120, only about 20 or 30 users are on the cloud at the moment.

Hon Weng Chong: 30:00 What we see now is the result of decisions made two months ago kind of thing.

Jason Calacanis: 30:24 Yeah.

Lon Harris: 30:27 But, so, when Doom is running on your biological computers, is that one CL1? Is that five?

Hon Weng Chong: 30:39 One. It was just one. So we haven’t come up with the chaining of the systems yet.

Lon Harris: 30:43 Oh okay, so each one’s still a discrete box.

Hon Weng Chong: 30:48 Yeah, exactly. There has been some internal research work about, you know, where we’re doing PDMS microfluidics.

Lon Harris: 31:21 So PDMS is a polydimethylsiloxane, which is also known as a dimethicone. What are we talking about?

Hon Weng Chong: 31:32 It’s a material. It’s like a gel-type of material that you can make microfluidic devices with.

Lon Harris: 31:43 Show me. Because and then define microfluidics for me because again, I nod my head like I understand, but we are outside of my comfort zone.

Hon Weng Chong: 31:57 Okay, so I’m going to show you this. So this is a microfluidic device. So there are probably several thousand neurons in one well.

Lon Harris: 32:11 And what we’re looking at here, if you didn’t know what it was, you would think these were top-down views of water storage tanks.

Hon Weng Chong: 32:26 Octagons.

Lon Harris: 32:27 Each one of these octagons has a bunch of neurons in it, and the microfluidics allows you to feed and then also transfer information between them?

Hon Weng Chong: 32:34 Yeah, the microfluidics are these channels and what they do is they constrain the growth of the axon, the long part of the neurons.

Lon Harris: 32:51 Right, and then you could in theory segregate to have more than one person using the same computer.

Michael Norcia: 33:00 Sharing information, got it. [Note: speaker tag misattributed by transcriber; context is still Cortical discussion]

Hon Weng Chong: 33:01 Right. So this is kind of like a chiplet model that we’ve been experimenting with.

Jason Calacanis: 33:29 So how much are you raising? What’s your target?

Hon Weng Chong: 33:33 Yeah, we’re raising 30, but I also wanted to show you guys this. So this is the Cortical Cloud.

Jason Calacanis: 33:46 Ah. Yeah, if you’re on the audio version, it looks like a standard developer backend, which is a compliment.

Hon Weng Chong: 33:53 Yes. And then you just hit the instance that you’ve acquired. I’m in New York right now, my lab is in Melbourne, and I am live streaming neural activity.

Jason Calacanis: 34:06 What is it doing? What are these dots that I’m seeing?

Lon Harris: 34:11 If you’re on the audio version, imagine, you’re looking out the window of a spaceship and you’re seeing stars go by.

Hon Weng Chong: 34:21 Yep. So, the window with stars, this is what we call raster plot. As we go down you can see the channels are like increasing in number.

Michael Norcia: 35:18 It’s really good it’s not conscious, otherwise that would be kind of rude like ‘I was sleeping!’ [Note: speaker tag misattributed - this is during Cortical segment]

Hon Weng Chong: 35:22 I know, yeah, how rude. Now, had a good dream there. We can see like the little spike waveforms.

Hon Weng Chong: 36:00 The matrix and the neurons are going to be put in. You specify the parameters of the game, the reward and punishments, the objective functions.

Lon Harris: 36:24 You know, it’s one of the best parts about talking to founders that I get to do on this is just straight up getting to see the future.

Hon Weng Chong: 36:44 Oh yeah.

Lon Harris: 36:45 I got to let you go. Come back on in six months and tell us how the cloud’s going and how Singapore’s going.

Hon Weng Chong: 37:06 Yes, so check us out at corticallabs.com. The cloud is at cloud.corticallabs.com. I’m on Twitter as well, DR1337. And also, we are looking for talented developers.

Lon Harris: 37:31 Are you going to go to that?

Hon Weng Chong: 37:33 Yeah, probably will have to go for that one.

Lon Harris: 37:36 So I only live like half an hour away from that area.

Hon Weng Chong: 37:41 Absolutely. Yeah, for sure.

Lon Harris: 37:44 All right. So stay tuned and yeah, more developers, channeling the inner Steve Ballmer: Developers, developers, developers! Hon, thank you.

Hon Weng Chong: 37:53 Thank you.

Jason Calacanis: 37:56 We’re talking about drones. No, not drones that sit on the water or go beneath the water or just fly up in the sky carrying a single hand grenade. No, today we’re going to talk about very large drones.

Michael Norcia: 38:24 Awesome. Thank you for having me. Doing well.

Jason Calacanis: 38:27 So I’m really excited about your company because I thought of you guys as the company that makes large drones that are powered by batteries that sprays crops. But as it turns out, you guys have really expanded.

Michael Norcia: 38:49 Yeah, great question. So honestly, the start of the story begins when I was like four years old. I’ve been just incredibly passionate about aviation my entire life.

Michael Norcia: 39:00 I was about 4, rubber band powered airplanes, little electric planes, pretty much obsessively built aircraft throughout my childhood. Ended up studying physics.

Jason Calacanis: 39:42 When was the first flight and how far have you guys gone in terms of commercial penetration of the agricultural sector with your drone?

Michael Norcia: 39:54 Yeah. So first flight was actually like a week before Demo Day. We went through Y Combinator. We spent one month designing a sort of 600-pound aircraft.

Jason Calacanis: 41:18 That’s weird. Talk to me about Group 4. I don’t think that’s a particularly well-known metric outside of the UAS space.

Michael Norcia: 41:30 Yeah, so it’s a more like military classification of drones. It’s basically drones that are bigger than 1,320 pounds. Things like the MQ-9 Reaper, that’s the most common Group 4 UAV.

Jason Calacanis: 41:40 All right. Now, you said you’d only done 1% of the work when you’d gone from no plane or no big drone to having a big drone that can fly.

Jason Calacanis: 42:00 It’s only 1% of the work because from the outside going from no plane to plane sounds like you made a lot of progress.

Michael Norcia: 42:05 Yeah, totally. So the thing with our products are a blend of hardware and software. The last nine years have been this just steady march of maturing both.

Jason Calacanis: 43:05 Put it on YouTube immediately we’re raising the next round.

Michael Norcia: 43:08 Exactly. With customers it’s completely different. Our customers are very upset if the aircraft is down for more than 24 hours.

Lon Harris: 43:53 Yeah, yeah, product reliability — market product reliability fit maybe we could call it.

Michael Norcia: 43:57 Yeah, exactly.

Jason Calacanis: 43:58 Why are we not using more of your ag drones here in the states?

Michael Norcia: 44:30 Yeah, so it’s coming. It’s coming fast. The things that drew us to Latin America were twofold. One is regulatory. Brazil actually deregulated agricultural drones about two years ago.

Jason Calacanis: 44:59 No, that’s not, that’s like, that’s…

Lon Harris: 45:00 It makes sense. Why?

Michael Norcia: 45:02 It has to do with line of sight to the vehicle.

Lon Harris: 45:06 Oh, Zipline told me about this, that if you want to fly outside of line of sight, there’s another entire set of regulations on top of that.

Michael Norcia: 45:12 It does, yes.

Lon Harris: 45:14 Is the air different in Brazil compared to the United States?

Michael Norcia: 45:22 Actually, it’s a totally diff— No, it’s not. Yeah, so in Brazil we’re operating typically anywhere from 5 to 15 kilometers from where we took off.

Michael Norcia: 45:30 So the good news is we have data about this. The FAA understandably is sort of apprehensive about what that looks like.

Lon Harris: 45:59 Alright, so I think this is good then to talk about fuel because it’s going to come up when we go over to drop-ship in a minute. You currently make mostly electric drones.

Michael Norcia: 46:25 Yeah. So farms have electricity. Some farms are pivot irrigated, for example.

Michael Norcia: 46:55 What’s really remarkable, so the kind of competition for our aircraft is a vehicle called the Air Tractor. It’s a big, roughly 8,000-pound vehicle. It burns roughly 50 gallons per hour.

Lon Harris: 47:09 Which is not small at today’s prices.

Michael Norcia: 47:11 No, yeah, it’s a lot. Our Pelican in the worst case, if you’re charging it off of a diesel generator, is about two gallons per hour.

Lon Harris: 47:23 That’s a lot better. Even if you’re doing dirty charging, as you might call it, it’s still much more efficient.

Michael Norcia: 48:00 Yeah, that’s a really good question. So it depends highly on how heavily you utilize the vehicle.

Lon Harris: 48:26 How comfortable would that make you as the guy who has to take the phone call if it goes down while it’s working 24 hours a day?

Michael Norcia: 48:42 It’s intense, but I mean we’re already, we have customers doing 12 hours a day right now.

Jason Calacanis: 48:52 In time are we going to see like smaller farmers be able to do ride-share equivalent of rental for these things?

Michael Norcia: 49:04 Yeah, correct. So in the US for example, you would need to have a farm that is almost 20,000 acres in order to fully utilize the aircraft.

Jason Calacanis: 49:31 You know we could make an acronym out of that, we could call it SAS. No one’s ever used that before.

Michael Norcia: 49:34 Yeah, totally. Maybe our valuation would go up.

Jason Calacanis: 49:39 Totally.

Michael Norcia: 49:41 Anyways, the payback period depends on the utilization, two to three years roughly.

Jason Calacanis: 50:37 Oh yeah.

Michael Norcia: 50:38 You’re trying to deposit this fine mist of chemical very evenly over an entire crop including around the boundaries.

Lon Harris: 50:46 And right now, the Strait of Hormuz is still totally blocked off and I presume by the time this comes out it’ll still be blocked, which is killing fertilizer prices.

Michael Norcia: 51:00 Yeah, totally, so input costs are, you know, like chemicals are a very significant portion of growing food.

Jason Calacanis: 51:06 Man. That’s…

Jason Calacanis: 51:08 It’s worrying to me that the solution to that, which is apparently you guys, is not as applied as the need for fertilizer is.

Michael Norcia: 51:52 Yeah, great question. So, just as general background, Dropship is a dual-use product. Within the commercial sector, it’s used for logistics. Within the defense sector, it sort of forks yet again.

Jason Calacanis: 52:14 Could it carry bombs in the little container? Sorry to be rude, but I’m curious.

Michael Norcia: 52:18 I mean, technically it could. That’s not the most, like, interesting use case for it, I would say.

Jason Calacanis: 52:27 Okay, and why not? Because when I think about drones today, we think a lot about drone-based warfare in Ukraine and in and around the Middle East.

Michael Norcia: 52:43 Yeah, so our vehicles are designed to be quite reliable and low operating cost. So if you want to do more kinetic stuff, you don’t build a vehicle like we did.

Lon Harris: 53:04 And probably also electronic emissions and noise and there’s probably other factors that would have gone into making this a bomber if you wanted it to be one.

Michael Norcia: 53:13 Correct. Yeah.

Jason Calacanis: 53:14 So you guys had a test flight recently of Dropship.

Michael Norcia: 53:22 Yeah, so we went from initial CAD renderings to first flight in 180 days, which is pretty awesome.

Jason Calacanis: 53:31 Yeah, dude.

Michael Norcia: 53:32 That wasn’t the fastest. I guess what we call Big Bird, the plane that we built in my parents’ backyard during Y Combinator technically was faster. But Dropship is way more complicated.

Michael Norcia: 54:00 The 200-mile useful range, so that was a sort of key limitation for defense. We built eight of those. We sent three of them to the Air Force.

Lon Harris: 55:13 Ah, okay, got it. Because the first cargo plane opened up in the back like a C-130.

Michael Norcia: 55:21 Yeah, exactly. So the nose opens on both of them, but on the electric one, the whole floor was a battery.

Jason Calacanis: 55:31 The nose! The nose opens up, not the back. Oh, I totally misread that image.

Michael Norcia: 55:34 No, no worries. Yeah, the nose opens on both. But yeah, we had this giant battery in the way, so we couldn’t make the electric one air drop.

Jason Calacanis: 55:52 Which is I feel we’re beating around the bush, the new dropship is hybrid.

Michael Norcia: 56:00 Yeah, yeah, exactly. So it’s a really cool architecture. I don’t think there’s any airplane out there that’s a hybrid turbo diesel architecture.

Hon Weng Chong: 56:22 Oh, so you run all three at the same time?

Michael Norcia: 56:24 Oh yeah, absolutely.

Jason Calacanis: 56:27 Oh, so it just must be going mad like a little bee, zzz, up into the sky.

Michael Norcia: 56:33 Yeah, it sounds really cool. It’s this combination of turbo diesel and then the electric.

Michael Norcia: 57:00 So yeah, the diesel engine is about, it’s like just over 30 kilowatts peak.

Michael Norcia: 57:09 And then the electric motors in the front are each about 25 kilowatts.

Jason Calacanis: 57:19 Yeah.

Michael Norcia: 57:21 So we use that for ballistic takeoff and landing performance, use it to climb to altitude, and then once at altitude, actually shut down the entire electric propulsion system.

Jason Calacanis: 57:41 You know, I was really excited about Dropship, not because of its dual-use capabilities, but more because I was thinking about domestic commercial applications.

Michael Norcia: 58:13 Yeah, 100%. The reason that we kind of pivoted away from mass manufacturing Pelican Cargo, the electric one, wasn’t because there was a lack of interest from customers.

Michael Norcia: 58:34 Are we shooting ourselves in the foot? Yes, absolutely. These hardware-software blended companies are hard.

Michael Norcia: 59:01 SpaceX is worth how much it is because they blew up rockets, figured out how to stop doing that, and now the idea of a self-launching, recovering rocket is just like taken for granted.

Jason Calacanis: 59:35 Zipline, yeah.

Michael Norcia: 59:37 Yep. And then there’s a bunch of other companies that have gone to Ukraine.

Lon Harris: 59:51 I want to make sure we get to supply chains, components, and kind of sovereignty. Now, if you were just making agricultural drones for Brazil, I wouldn’t really care, but we are talking about the military.

Michael Norcia: 1:00:00 So, I think it kind of because of when we started the company there wasn’t like a bunch of stuff we could just buy off the shelf to build these drones out of. So we actually vertically integrated.

Michael Norcia: 1:00:16 Yeah, good question. So…

Jason Calacanis: 1:00:38 What’s the cost differential there?

Michael Norcia: 1:00:41 It’s about 2x.

Jason Calacanis: 1:00:44 That’s, is that a high price component compared to the overall cost of the device?

Michael Norcia: 1:00:52 No, not I mean so for drop ship, no. There’s two BMS’s in the entire vehicle. For Pelican, it’s more.

Jason Calacanis: 1:00:56 Yeah, yeah, yeah. So more complicated. 2x didn’t sound as bad as I was expecting.

Michael Norcia: 1:01:03 Well, so if you own the design, it’s not that bad. If we don’t source from China, we could source from another low cost region.

Jason Calacanis: 1:01:13 Oh, no way. Oh. Huh. All right, well that’s better than I thought.

Michael Norcia: 1:01:16 Yeah, I mean I think if you own the design and manufacturing in the US versus China is order of magnitude 2x difference.

Jason Calacanis: 1:01:31 Yeah, yeah, quite a lot. Okay, that makes perfect sense to me. But you mentioned how when you started the company, things weren’t available on the shelf.

Michael Norcia: 1:01:46 Yeah, yeah, for sure. So I think we have an extremely strong technical team. It has slowed us down.

Jason Calacanis: 1:02:29 Or…

Michael Norcia: 1:02:32 We’re working on the last couple of resolutions, everybody. Give us two, three weeks.

Michael Norcia: 1:03:00 It’s tough to say. I think the biggest thing for us though is, you know if you think of the most successful hardware companies out there are a blend of hardware software.

Lon Harris: 1:03:10 Apple.

Hon Weng Chong: 1:03:11 Tesla.

Michael Norcia: 1:03:12 Apple, great example. Yeah, Tesla. DJI, perfect example.

Lon Harris: 1:03:15 Sure. Absolutely.

Michael Norcia: 1:03:16 What those companies are able to do is make something incredibly complicated seem just really simple to the end user.

Lon Harris: 1:03:37 What are all these little camera things back here? I don’t know what a single one of those does.

Jason Calacanis: 1:03:43 Me neither.

Lon Harris: 1:03:44 And you know what? My pictures of my kids look fantastic, so thank you Apple.

Michael Norcia: 1:03:48 Totally. Yeah, the way you do that is if you have to own everything. You have to own the hardware and the software.

Lon Harris: 1:04:10 Are we just talking around Boeing’s decision to stop making things in-house and just supply everything externally?

Michael Norcia: 1:04:31 Yeah, good question. I would say yes and no. We’ve been able to raise the money that we need. If we had more money, we would move faster.

Lon Harris: 1:04:37 If you’re a VC listening to this, hello! Come on! Stop backing SaaS companies that are slapping on an AI wrapper. Back something cool. We’re building drones, y’all.

Lon Harris: 1:04:48 Michael, a treat. Where can people find the company?

Michael Norcia: 1:04:51 Yeah, great. So you can find us on our website flypyka.com.

Lon Harris: 1:04:57 P-Y-K-A, not P-I-K-A. Yeah.

Michael Norcia: 1:05:00 Correct, P-Y-K-A. Very active on LinkedIn, very active on Instagram. Most of it’s in Portuguese, I’ll warn you.

Lon Harris: 1:05:11 Fantastic. All right. Thank you very much. We’ll see you soon.

Michael Norcia: 1:05:14 Great. Thank you.