Andrew Blum discusses the physical infrastructure of the internet, hyperscaler monopolies, and his new focus on energy grids.
Summary
In this episode, author Andrew Blum explains that the internet is a physical 'network of networks' reliant on tangible infrastructure like fiber optic cables and data centers, rather than an ethereal cloud. He highlights key interconnection points such as 60 Hudson Street in New York and Amsterdam’s policy-driven hub status, while noting that hyperscalers like Google and Meta are vertically integrating by owning this underlying infrastructure. Blum warns that this concentration of power threatens the open nature of the internet and creates accountability gaps, particularly regarding AI. He also introduces his upcoming book, which examines the electric grid, citing the Netherlands as a model for decarbonization and noting that AI demand is driving unprecedented grid construction in the US.
Episode sections
- 0:12 – 3:00
Introduction and the Physicality of the Internet
Andrew Blum discusses the necessity of physical infrastructure for the digital internet, highlighting 60 Hudson Street in New York as a critical interconnection point.
- 3:00 – 11:40
Origins of the Book and the 'Network of Networks'
Blum explains that his book *Tubes* was inspired by a squirrel chewing a cable, leading him to investigate the physical paths of internet traffic.
- 11:40 – 15:04
Key Interconnection Points: 60 Hudson Street
Blum details his visit to 60 Hudson Street in Manhattan, a historic telecom building that serves as a major physical interconnection point for international fiber optic cables.
- 15:04 – 25:23
Technical Layers and Signal Transmission
The speakers discuss the technical journey of data, explaining that voice and video are converted into pulses of light traveling through fiber optic cables.
- 25:23 – 29:14
Amsterdam as a Deliberate Internet Hub
Blum highlights Amsterdam’s unique rise as an internet hub due to deliberate policy, such as allowing multiple providers to lay fiber in public trenches.
- 29:14 – 31:54
Geography of Internet Hubs and Cables
Andrew Blum identifies key internet hubs in the US and Europe that serve as connection points for undersea cables.
- 31:54 – 33:40
Trust, Privacy, and Centralization
The speakers discuss the tension between the internet's foundational need for trust/openness and consumer desires for privacy.
- 33:40 – 35:09
The Physical Reality of the 'Cloud'
Blum clarifies that 'the cloud' consists of physical data centers with hard drives, not ethereal storage, highlighting significant environmental impacts.
- 35:09 – 40:03
Inside a Data Center
Blum describes data centers as loud, dark, and cold, with technicians always present working manually on hardware issues.
- 40:03 – 42:34
Hyperscalers and Infrastructure Monopoly
Blum expresses concern that hyperscalers like Google and Meta have moved from building applications to owning the physical infrastructure.
- 42:34 – 46:49
AI, Regulation, and Accountability
The discussion turns to AI, with Blum noting that AI companies have not yet demonstrated accountability to markets or regulators.
- 46:49 – 51:21
Infrastructure Fragility and Resilience
Blum explains that while the internet is designed to reroute traffic when cables break, multiple simultaneous cuts can severely degrade connectivity.
- 51:21 – 59:54
The Human Element and Future Outlook
Blum reflects on the human element of internet infrastructure and expresses cautious optimism about humanity's ability to manage AI risks.
- 59:54 – 1:03:26
New Book Topic: Energy Transition and Grid Infrastructure
Andrew Blum states his new book focuses on the electric grid and the energy transition from fossil fuels to renewable generation.
Assets discussed
| Asset | Kind | Stance | Context |
|---|---|---|---|
| AlphabetGOOGL | equity | neutral | Um, and and you know, when you talk about you know, Google or Meta, you're talking about layer two or layer three, but then you're also talking about layer one, because now they actually own physical cables across the ocean and certainly they own data centers. |
| Meta PlatformsMETA | equity | neutral | Um, and and you know, when you talk about you know, Google or Meta, you're talking about layer two or layer three, but then you're also talking about layer one, because now they actually own physical cables across the ocean and certainly they own data centers. |
| AmazonAMZN | equity | neutral | So what's interesting to me is that companies like Google and Amazon started by building businesses on top of the internet. |
“Stance” is how the speakers talked about the asset, not a recommendation.
Claims worth verifying
60 Hudson Street in Lower Manhattan is a significant physical interconnection point for internet networks, originally built in 1929 for telegraph lines.
Andrew Blum ·13:22
How to verify: Search for historical records of 60 Hudson Street, New York, and its role in telecom infrastructure; verify construction date via NYC landmark or building records.
KPN is a major Dutch telecom company that is approximately 125 years old.
Andrew Blum ·15:34
How to verify: Check KPN's corporate history or Wikipedia for founding date (1852) and current status as a major Dutch telecom provider.
Amsterdam became a major internet hub due to deliberate policy allowing multiple providers to lay fiber in public trenches.
Andrew Blum ·26:13
How to verify: Research Amsterdam's municipal fiber policies from the late 1990s/early 2000s and the formation of AMSIX.
The internet is fundamentally a 'network of networks' where logical protocols overlay physical circuits.
Andrew Blum ·6:48
How to verify: Review technical documentation on Internet architecture (TCP/IP stack) and the concept of peering and transit between autonomous systems.
Data signals are converted to pulses of light for transmission over fiber optic cables, often remaining in light format for long distances.
Andrew Blum ·21:11
How to verify: Consult technical resources on fiber optic communication, specifically regarding optical transceivers and signal conversion processes.
Nanog (North American Network Operators Group) is a group of network operators who meet quarterly to discuss internet infrastructure.
Andrew Blum ·11:52
How to verify: Visit the NANOG website or archive to verify the organization's name, purpose, and meeting frequency.
In the late 1990s and early 2000s, the top internet hubs were Frankfurt, London, and Amsterdam.
Andrew Blum ·29:01
How to verify: Review historical internet exchange reports or industry analyses from the early 2000s regarding major IXPs and traffic volumes.
Riverside is a podcast recording platform that prioritizes low latency by placing servers near major network hubs like Ashburn, VA.
Andrew Blum ·17:35
How to verify: Check Riverside.fm's technical documentation or blog posts regarding their infrastructure and server locations.
Hyperscalers like Google and Meta have built their own undersea cables and data centers over the last 10 years, dominating both application and physical infrastructure.
Unidentified speaker ·40:41
How to verify: Check SEC filings or press releases from Google (Alphabet) and Meta regarding infrastructure investments, undersea cable consortiums (e.g., Dunant, 2Africa), and data center construction announcements.
Data centers have a distinct smell described as 'burnt toast' due to equipment ions and air conditioning.
Unidentified speaker ·35:44
How to verify: Search for anecdotal reports or technical articles describing the sensory experience of entering data centers; verify with industry blogs or journalist accounts of data center tours.
Multiple cable cuts in the Red Sea and Mediterranean have severely degraded internet connectivity in Africa and the Middle East.
Unidentified speaker ·47:47
How to verify: Consult reports from internet monitoring firms (e.g., Cloudflare, Akamai) or news archives from 2022-2024 regarding cable cuts in the Red Sea and Mediterranean affecting regional connectivity.
The Netherlands implemented a moratorium on data center construction, initially based on land issues and later on power constraints.
Unidentified speaker ·44:57
How to verify: Search for Dutch government announcements or news reports regarding the 'moratorium' on data centers in the Netherlands, specifically looking for reasons related to energy grid capacity.
Network operators in the field were almost entirely men, and the social environment was sexist and derogatory.
Unidentified speaker ·58:19
How to verify: Review Andrew Blum's book 'Tubes' or his subsequent interviews/articles where he reflects on the gender dynamics and workplace culture of the telecom industry during his research period.
AI requires significant physical infrastructure and power, leading to concerns about carbon emissions and grid costs.
Unidentified speaker ·34:30
How to verify: Review industry reports from energy agencies (e.g., IEA) or tech companies regarding the power consumption of AI data centers and their impact on local electricity grids.
The internet's underlying infrastructure is designed to reroute traffic when cables break, but this has limits.
Unidentified speaker ·47:22
How to verify: Consult technical documentation on BGP (Border Gateway Protocol) and internet routing resilience, or reports from internet infrastructure organizations (e.g., RIPE NCC) on network redundancy.
Google and Meta dominate the application space while also owning physical infrastructure, creating a quasi-monopoly.
Unidentified speaker ·40:03
How to verify: Analyze market share data for search (Google) and social media (Meta) alongside their infrastructure ownership disclosures in annual reports.
Andrew Blum's new book is about the electric grid and the energy transition from fossil fuels to renewable and carbon-free generation.
Andrew Blum ·59:57
How to verify: Check Andrew Blum's official website (andrewblum.net) or publisher announcements for details on his upcoming book project.
The Netherlands is decarbonizing primarily through the construction of offshore wind and offshore grid infrastructure.
Andrew Blum ·1:00:10
How to verify: Review energy reports from the Dutch government or organizations like RVO (Dutch Enterprise Support Agency) regarding offshore wind capacity and grid expansion.
During the Biden presidential administration, there was a huge effort to get solar, wind, and batteries on the US grid.
Andrew Blum ·1:00:44
How to verify: Review executive orders, DOE announcements, or IRA (Inflation Reduction Act) implementation reports from the Biden administration (2021-2025).
When Trump became president, efforts to build solar, wind, and batteries on the grid were actively stopped, though not entirely successfully.
Andrew Blum ·1:00:44
How to verify: Review executive orders, policy changes, or regulatory actions taken by the Trump administration (2025-present) regarding renewable energy incentives and grid projects.
There has been a growth in the US electric grid that has not been seen in 50 years, driven by a boom in AI.
Andrew Blum ·1:00:55
How to verify: Check data from the US Energy Information Administration (EIA) or the Electric Power Research Institute (EPRI) on recent grid capacity additions and load growth forecasts.
Andrew Blum's website is Andrewblum.net.
Andrew Blum ·1:02:42
How to verify: Visit the URL andrewblum.net to confirm it is owned by Andrew Blum and contains his biography and book information.
Andrew Blum's former Twitter handle is AJ Blum.
Andrew Blum ·1:02:44
How to verify: Search for the Twitter/X profile @AJBlum and check the bio for association with Andrew Blum.
Do your own research
- Verify the physical location and current operational status of 60 Hudson Street as a major interconnection point.
- Research the specific municipal policies in Amsterdam that allowed open access to fiber infrastructure in public rights-of-way.
- Analyze the vertical integration strategies of major hyperscalers (Google, Meta) regarding their ownership of undersea cables and data centers.
- Review environmental impact reports or sustainability disclosures from major cloud providers regarding electricity and water consumption.
- Examine regulatory filings or public statements regarding AI company accountability and governance frameworks compared to established platforms.
- Assess the current state of undersea cable redundancy in Europe versus regions like Africa and the Middle East.
- Monitor US electric grid construction data and policy shifts under different administrations, specifically regarding solar, wind, and battery projects.
- Review the Netherlands' energy transition strategy, focusing on offshore wind integration and grid infrastructure development.
Full transcriptexpand
- 0:12
- Unidentified speaker: I think that's kind of one of the keys of understanding how the physical infrastructure of the internet works is that is that these are two s two separate realms, the kind of physical and the digital, um, but they absolutely need each other to operate.
- 0:27
- Unidentified speaker: I could say they they very clearly knew exactly what the most important places on their map were.
- 0:33
- Unidentified speaker: One of the first that I that I visited and and still kind of a favorite is 60 Hudson Street.
- 0:39
- Unidentified speaker: It was pretty shocking to me that there was a building that I'd walked by all the time that I didn't realize was one of the sort of most significant places on the planet where the networks of the internet physically connect to each other.
- 0:57
- Unidentified speaker: Do you think the internet gave ordinary people more power, or do you think the internet actually gave more power to uh these large companies?
- 1:05
- Unidentified speaker: But it's a tough one.
- 1:06
- Unidentified speaker: I mean, I mean it depends.
- 1:07
- Unidentified speaker: You it's just definitely a sort of story of disruption of the kind of institutions of media, but to what at what cost?
- 1:14
- Unidentified speaker: The reason we lost the understanding is because of the success of Google and Facebook primarily.
- 1:20
- Unidentified speaker: Um you could say Apple too, perhaps, but Google and Facebook primarily in taking up so much of the air.
- 1:27
- Unidentified speaker: So much of you kind of have to sort of peek around those giants to see the other things underneath, to really remember that there is um that there is a lot, uh a lot beneath the surface.
- 1:58
- Unidentified speaker: We work online, we talk to friends, on the other side of the world, we watch videos, transfer money, and increasingly we have conversations with AI.
- 2:08
- Unidentified speaker: And most of it happens almost instantly.
- 2:11
- Unidentified speaker: But there's something I hadn't really thought about until recently because where is the internet actually?
- 2:18
- Unidentified speaker: And that's when I stumbled upon this book.
- 2:21
- Unidentified speaker: If I send a message from my home in the Netherlands to someone in New York, where does that message go?
- 2:27
- Unidentified speaker: And who owns those connections?
- 2:29
- Unidentified speaker: We talk about the cloud as it is something that is floating somewhere above us.
- 2:35
- Unidentified speaker: My guest today is Andrew Blum, the author of Tube's A Journey to the Center of the Internet.
- 2:41
- Unidentified speaker: And Andrew actually went looking for the Internet and he visited the places where it exists, where it doesn't exist, met the people who kept it running, and discovered a world most of us never see.
- 2:54
- Unidentified speaker: Andrew, I hope this introduction does you some justice, but uh thanks for uh joining me.
- 3:00
- Unidentified speaker: Thanks for having me.
- 3:00
- Unidentified speaker: It's uh it's it's like the internet powers on almost 20 years later since I started working on this.
- 3:06
- Unidentified speaker: So it's great to great to great to talk about it again.
- 3:09
- Unidentified speaker: The internet it hasn't left you since you did your research and uh So how uh I was quite late to getting your book.
- 3:21
- Unidentified speaker: I found this book actually because I wanted to learn my students more about the internet, so I decided to start reading up a little bit uh more on it.
- 3:29
- Unidentified speaker: So what edition did I buy?
- 3:31
- Unidentified speaker: I think you have you have a US publication hardcover, but like from the end of the hardcover publication, like not a first not like the first print run, but a subsequent print run.
- 3:41
- Unidentified speaker: So uh, you know, you you found a good one.
- 3:42
- Unidentified speaker: Yeah, well, yeah, not at all.
- 3:44
- Unidentified speaker: I it's still pretty old.
- 3:45
- Unidentified speaker: It's still pretty old, still aged, still, still well aged.
- 3:49
- Unidentified speaker: Awesome.
- 3:49
- Unidentified speaker: Uh when I opened the book, it actually uh usually the first 10 pages already tell me whether I'm going to like the book or not.
- 3:59
- Unidentified speaker: Uh and when I started opening the book and started reading, it it started with um a squirrel, which which actually already caught my attention in the beginning.
- 4:09
- Unidentified speaker: It's quite uh quite something.
- 4:11
- Unidentified speaker: You your internet was having some issues at home, and then somebody came over and turned out a squirrel had been chewing on a cable.
- 4:18
- Unidentified speaker: And I mean, most people would just get it fixed and forget about it, but that sort of brought you on a journey.
- 4:26
- Unidentified speaker: Yeah, I mean, I just I I got obsessed with the with the map of it.
- 4:30
- Unidentified speaker: Like I just I I couldn't like it, it just seemed like if there were actually cables between me and everyone else that I was exchanging messages with.
- 4:38
- Unidentified speaker: And this was 2009.
- 4:40
- Unidentified speaker: So it was a different kind of messages, different sort of different different level of pervasiveness.
- 4:45
- Unidentified speaker: You know, we just we weren't we weren't online all the time in the same way.
- 4:48
- Unidentified speaker: And it's like if there all those cables were there, how how could I find them?
- 4:52
- Unidentified speaker: You know, it's the there had to be a continuous path.
- 4:55
- Unidentified speaker: And I was told repeatedly that because the internet breaks data data up into packets, that the you could never find the path.
- 5:03
- Unidentified speaker: And if there were many paths, that meant there was no path at all.
- 5:06
- Unidentified speaker: Um, but of course, that that can't actually physically be true.
- 5:09
- Unidentified speaker: There has to be a path.
- 5:10
- Unidentified speaker: And so I went, I went looking for that path.
- 5:13
- Unidentified speaker: I went looking for the people who could explain to me how it fits together.
- 5:16
- Unidentified speaker: And then I decided to make my own judgments about what the important parts were, you know, what what the sort of meaningful parts were.
- 5:23
- Unidentified speaker: You know, really with the belief that like, you know, places, you know, we we places are places on the map, but places are also how we think about them and how we imagine them and how we talk about their importance and how we protect them or how we how we value them or how we buy them or sell them.
- 5:36
- Unidentified speaker: You know, the places are about all of those things.
- 5:39
- Unidentified speaker: So this uh this issue, this problem existed at home.
- 5:44
- Unidentified speaker: What was the first step after I I can imagine that the technician just fixed it and told you what was wrong.
- 5:50
- Unidentified speaker: So after he left, what was going through your mind of, okay, what next?
- 5:56
- Unidentified speaker: It was really where does that cable go?
- 5:59
- Unidentified speaker: I mean, I had my internet service provider at home, which was my cable company.
- 6:04
- Unidentified speaker: I didn't particularly like them.
- 6:05
- Unidentified speaker: I didn't know much about them.
- 6:06
- Unidentified speaker: They didn't really want to talk to journalists.
- 6:09
- Unidentified speaker: Um, but it was became self-evident that if my cable connected to where whatever their kind of you know neighborhood or or or kind of citywide hub was, then that citywide hub had to connect to everyone else.
- 6:25
- Unidentified speaker: It's hard to, it's hard to kind of take myself back to kind of not that sort of the ignorance of that.
- 6:31
- Unidentified speaker: But it really was this conviction that like, you know, the internet can't, we didn't even use the word cloud then.
- 6:36
- Unidentified speaker: Like the internet just can't be this sort of ethereal thing.
- 6:38
- Unidentified speaker: And it can't just be the telephone system.
- 6:41
- Unidentified speaker: There has to be something more to it.
- 6:42
- Unidentified speaker: There has to be some place where the networks of the internet, I knew I'd always known the internet is a network of networks.
- 6:48
- Unidentified speaker: That's why we call it the internet.
- 6:50
- Unidentified speaker: Like that was a I'd known that since, you know, since the 90s when I first got online.
- 6:54
- Unidentified speaker: But then it's like, well, okay, so where are those places?
- 6:57
- Unidentified speaker: Like who can tell me where these places are where one network of the internet connects to another network of the internet?
- 7:03
- Unidentified speaker: And if those places existed, could I go visit them?
- 7:06
- Unidentified speaker: Could who who who had the keys?
- 7:08
- Unidentified speaker: Who owned them?
- 7:08
- Unidentified speaker: Who could let me in, who could show me around?
- 7:12
- Unidentified speaker: Could you uh explain that concept a little more?
- 7:15
- Unidentified speaker: Because I the network of networks, I I actually didn't know that before reading the book, actually.
- 7:22
- Unidentified speaker: Yeah.
- 7:23
- Unidentified speaker: It's funny, you know, since in more recently we think of the internet in terms of the hyperscalers.
- 7:29
- Unidentified speaker: We think of them in terms of, you know, the the the primary ways we we kind of the sort of the these well-worn the yeah, there's they've kind of the hyperscalers have kind of you know enlarged so much that they now kind of obscure the infrastructure underneath.
- 7:44
- Unidentified speaker: So so much you know is travel is done on Meta, so much is done on Google, um, you know, so it's it's hard to kind of see what what's what's beneath.
- 7:53
- Unidentified speaker: But really, um, you know, the initial notion was that you had um you had logic networks.
- 8:02
- Unidentified speaker: You had the connection, universities primarily connecting with other universities in the in the in the very early days.
- 8:08
- Unidentified speaker: And so you would have a uh sort of a telecom connection, uh a circuit as it was called.
- 8:13
- Unidentified speaker: And so a research university in in Los Angeles, UCLA, for example, would connect to a research university in in Silicon Valley, Stanford.
- 8:21
- Unidentified speaker: And that becomes one of the first circuits, and that's a single network where they are they are sort of uh you know defining a logic network.
- 8:29
- Unidentified speaker: They're they're sort of you have both the a physical telephone circuit, but then also a protocol for how to communicate over it.
- 8:35
- Unidentified speaker: And then there's this question of, okay, well, if we have our circuits and we connect to a third person, but then there's two other people or three other people who also have their circuit, can we then all connect together?
- 8:45
- Unidentified speaker: Can we interconnect?
- 8:47
- Unidentified speaker: Um, and that sort of basic idea that you can have a uh, you know, a sort of a network overlaid upon another network that's connecting to each other becomes the foundation of the internet.
- 8:57
- Unidentified speaker: I mean, the kind of visualization for that that I like is a sort of pile of phishing nets.
- 9:00
- Unidentified speaker: It's like I have my phishing net and it's got all the sort of nodes where then where the the the rope, the cables of the net connect to each other, where then you throw your phishing net on top, and then we, they're they're kind of resting on top of each other, and then they they interconnect in different places.
- 9:13
- Unidentified speaker: And then that once you have the kind of protocol, not just the kind of the logical protocol, not just the software, but also the kind of social construct, the kind of the agreement for how these networks will connect to each other and who's going to pay whom for that connection, then you can begin to have a uh the the growth that begins kind of in the 80s and really takes off in the in the 90s.
- 9:33
- Unidentified speaker: Um, you know, of course, with the dot-com boom in, you know, around around 2000, 2001, things really got going and Amsterdam's role in that was so crucial.
- 9:41
- Unidentified speaker: Uh and then a lot has happened since, certainly.
- 9:46
- Unidentified speaker: You know, today kids grow up with uh TikTok, YouTube, uh Google.
- 9:52
- Unidentified speaker: And I kind of I was kind of growing up in in an in an in a time where the internet was already there, 1998.
- 10:01
- Unidentified speaker: Uh it it wasn't as important and then as big as it is now, obviously, but I kind of grew up with the internet growing.
- 10:09
- Unidentified speaker: But these kids today, I feel like to them the internet is basically just a collection of apps and uh and websites.
- 10:19
- Unidentified speaker: Uh do you think uh we sort of lost the understanding that the internet is actually like thousands of different networks working together?
- 10:28
- Unidentified speaker: Yeah, I mean I wouldn't say that I think we lost the understanding, but I would I would place the kind of the blame or the kind of explanation for how we lost the understanding with the hyperscalers.
- 10:40
- Unidentified speaker: You know, the reason we lost the understanding is because of the success of Google and Facebook primarily.
- 10:46
- Unidentified speaker: Um you could say Apple too, perhaps, um, but Google and Facebook primarily in taking up so much of the air, um, you know, and in really kind of, you know, uh, you know, um sort of sort of can you know being in front of our eyeballs for so much of the time that we spend online.
- 11:02
- Unidentified speaker: And certainly there's an internet still that is not Google and and Meta, but so much of you kind of have to sort of peek around those giants to see the other things underneath, to really remember that there is um that there is a lot, uh a lot beneath the surface there.
- 11:20
- Unidentified speaker: Um that of course now is being sort of scrambled by our our sort of emerging understanding of how the the AI companies now are kind of redefining those, that, that, that, that dynamic.
- 11:31
- Unidentified speaker: Yeah.
- 11:32
- Unidentified speaker: Yeah, it's a good point.
- 11:33
- Unidentified speaker: Take me back to um to the beginning of this book of uh trying to find like the center of the internet.
- 11:40
- Unidentified speaker: What was like the one of the first places that uh you visited that really surprised you when you saw it?
- 11:47
- Unidentified speaker: Yeah, well, the the first thing, I mean, I had to figure out where to go, and then I had to figure out who was gonna let me in.
- 11:52
- Unidentified speaker: And the key for that was a group called Nanog, the North American Network Operators Group.
- 11:58
- Unidentified speaker: Um, and that was the kind of organization of geeks who were the network operators, the that sort of the technicians who, you know, connected the the networks of the internet to each other.
- 12:10
- Unidentified speaker: And at the time, that group was, you know, they're they met quarterly because it was, they kind of always had to be together and be in touch with each other.
- 12:18
- Unidentified speaker: And at the time, the group was just the meetings were a few hundred people, even less.
- 12:22
- Unidentified speaker: Um and what they they they they show up at that meeting, each representing a network.
- 12:29
- Unidentified speaker: So a network could be a telecom company, a network could be Google, you know, a network could be uh a kind of uh somewhere in between, a as they call it, I say a transit provider, a company that sort of um you know uh sells the traffic on the internet's pipes.
- 12:42
- Unidentified speaker: Um and uh once I sort of got to their conference and started talking to this group of people, I could say they they very clearly knew exactly what the most important places on their map were.
- 12:55
- Unidentified speaker: Uh one of the first that I that I visited um and and still kind of a favorite is 60 Hudson Street, which is uh uh an office building, an old telecom building in Lower Manhattan, um, you know, about uh, I don't know, about about a about a mile and three quarters from where I'm sitting right now, um that um that was originally a place where telegraph lines connected when it was built in 1929, and then became a place where telephone lines were connected.
- 13:22
- Unidentified speaker: And then when the current generation of fiber optic cables across the ocean, across the Atlantic, were laid to the end of the in the 90s and the early 2000s, it was kind of self-evident that of course they would go to the place where all the networks already connected, and that was 60 Hudson Street.
- 13:38
- Unidentified speaker: And so from there, once you kind of had those important international connections, then everything kind of grows up from there.
- 13:43
- Unidentified speaker: So I it was pretty shocking to me that there was a building that I'd walked by all the time that I didn't realize was one of the sort of most significant places on the planet where the networks of the internet physically connect to each other.
- 13:59
- Unidentified speaker: And when I say physically connect to each other, I mean with like a yellow cable.
- 14:02
- Unidentified speaker: Like there's a cage of equipment for that belongs to one network, and there's a yellow cable that goes up into the ceiling and goes down into the cage of equipment of another network.
- 14:10
- Unidentified speaker: And that that connection is absolutely physical and it happens, it happens in multiple places for redundancy, but it certainly happens in in that place at 60 Hudson, um, which is kind of has a kind of, you know, it has a sort of a you know art deco vibe.
- 14:23
- Unidentified speaker: Like it's not, it's not, it was high-tech, it was high-tech 100 years ago, but um, but is uh but certainly has a different feel today.
- 14:29
- Unidentified speaker: Aaron Ross Powell More of a museum today.
- 14:32
- Unidentified speaker: Well, uh I mean a museum that's a busy and uh very valuable place, um, you know, certainly.
- 14:38
- Unidentified speaker: So I would love to understand like the basics a little better, because we all use the internet today, but many of us still struggle to like fully understand it or explain it.
- 14:50
- Unidentified speaker: So I'm sitting here in the Netherlands, you're sitting in the United States, uh, yet we're talking almost instantly.
- 14:58
- Unidentified speaker: So what happens between this the space between us?
- 15:01
- Unidentified speaker: Can you walk us through that journey from here to there?
- 15:04
- Unidentified speaker: Yeah, as best I can.
- 15:06
- Unidentified speaker: I mean, it's well, I'll I'll I I mean it's I'll I'll add another piece of complexity to it, but it's um I mean the first thing to recognize is that we're each connected to an internet service provider.
- 15:16
- Unidentified speaker: You know, where we each have a have a company that services our home or office and can and and whose job it is to connect us to the broader internet.
- 15:24
- Unidentified speaker: So where I'm sitting right now, I'm I'm connected to the internet as a spectrum customer, you know, one of the big, you know, pr very big media companies in the US.
- 15:31
- Unidentified speaker: Using KPN, the one you mentioned in the book.
- 15:33
- Unidentified speaker: Yep.
- 15:34
- Unidentified speaker: Yep, yep.
- 15:34
- Unidentified speaker: You know, again, big, you know, made major Dutch telecom, you know, 125 years old, probably, whatever it is.
- 15:39
- Unidentified speaker: Yeah.
- 15:40
- Unidentified speaker: Um and um so that's the first step.
- 15:42
- Unidentified speaker: But then you have to think, okay, so we have our connection.
- 15:45
- Unidentified speaker: They're responsible for the truck, they're responsible for the wires outside the house, they're responsible for the router on my, you know, the cable modem and the router on my desk.
- 15:53
- Unidentified speaker: Um and then from there, they have to connect to everyone else.
- 15:57
- Unidentified speaker: And so I can I can be pretty confident that that that um that spectrum is connecting to one of the international um networks uh at 60 Hudson Street.
- 16:07
- Unidentified speaker: That would be a pretty good guess.
- 16:09
- Unidentified speaker: Um, you know, there there might be others.
- 16:11
- Unidentified speaker: And because we're um we're both on pretty big networks, it might be as simple as that.
- 16:16
- Unidentified speaker: It might be actually that KPN has a um has a has a piece of equipment in Manhattan and Spectrum and KPM connect directly or peer.
- 16:27
- Unidentified speaker: They connect as as peers to each other.
- 16:29
- Unidentified speaker: And so it really could be as simple as that if we were, if we traced our our our path and then it was just from Spectrum to KPM, that one connected to the other.
- 16:38
- Unidentified speaker: Um because KPM or Spectrum uh goes through the trouble of of buying bandwidth across the ocean because it's more cost efficient and and and uh and more you know um more better protect better you know more redundant to um to do that themselves rather than than buy traffic on a third party in some form or another.
- 16:59
- Unidentified speaker: But the complexity of to add to that, of course, is that we're not just talking to each other, but we're uh using the you know Riverside as a platform to record the podcast.
- 17:08
- Unidentified speaker: And so there's a kind of open question of, you know, where, you know, what is Riverside's infrastructure in this?
- 17:14
- Unidentified speaker: Are we both connecting to Riverside's data center?
- 17:17
- Unidentified speaker: And Riverside being a company that's you know very focused on latency, you know, it's really important for their, for their their product that we, there's no delay in our call, um, you know, that all that it's a high bandwidth connection that we can sort of both see.
- 17:30
- Unidentified speaker: They are probably making an effort to put their their servers as close as possible to these network hubs.
- 17:35
- Unidentified speaker: So maybe that means that they're in Ashburn, Virginia, which is a kind of major center of the internet and a huge data center hub.
- 17:42
- Unidentified speaker: Um, maybe that means their their server is in in Newark, New Jersey, you know, as close as possible, you know, you know, 10 milliseconds away from 60 Hudson.
- 17:50
- Unidentified speaker: Um maybe because you're uh you're a a European customer Riverside, maybe they're it's being hosted in uh in Amsterdam, in a data center in Amsterdam.
- 17:58
- Unidentified speaker: Uh I mean we could probably figure that, we could probably find the answers to that that question.
- 18:02
- Unidentified speaker: Um but it is important to recognize that at the complexity has grown of the tools that we're using, it's no longer just like connecting a phone call, but but we're kind of layering in all these other things that are so carefully optimized so that they so that we see you know, so that the late there's no latency in our in our conversation, that it's kind of, you know, it's in some ways almost better than it would have been with a phone call um, you know, 30 years ago, you know, with a you know, with a where where we had a uh you know uh an audible delay of almost a second um you know across the ocean.
- 18:34
- Unidentified speaker: Let's say I'm uh I'm 12 years old.
- 18:36
- Unidentified speaker: And uh I'm not, but uh uh let's say I'm 12 years old and you're trying okay, so we know that there's this space between us, there's these data centers, there's these hubs that connect these things together.
- 18:50
- Unidentified speaker: There's a cable going uh through the ocean.
- 18:53
- Unidentified speaker: How does my voice reach your what are those earpods or just uh earphones?
- 19:00
- Unidentified speaker: My old air, my old wired airbuds.
- 19:02
- Unidentified speaker: Yeah, yeah.
- 19:03
- Unidentified speaker: What happens in those in those in those cables?
- 19:06
- Unidentified speaker: Okay.
- 19:06
- Unidentified speaker: So the way that the internet people talk about it is in terms of layers.
- 19:11
- Unidentified speaker: So it's seven layers, conceptual layers.
- 19:14
- Unidentified speaker: And layer zero is dirt, and layer one is glass or copper, and layer two is the kind of encoding, the sort of highest level encoding of the flashes of light through a fiber optic cable.
- 19:28
- Unidentified speaker: And then when you get to layer three, you get to more what we think of as software.
- 19:31
- Unidentified speaker: You know, you get to think of the kind of different programming languages and different different uh encoding mechanisms that do that.
- 19:37
- Unidentified speaker: I I I don't know much about that stuff.
- 19:38
- Unidentified speaker: You know, that's sort of a that's like my focus is always on layer zero, the dirt, where things actually is, and layer one, the kind of physicality of it.
- 19:46
- Unidentified speaker: And occasionally you kind of have to, when you talk about these sort of pile of fishing nets of networks, then you're talking about layer two and sometimes about layer three.
- 19:54
- Unidentified speaker: Um, and and you know, when you talk about you know, Google or Meta, you're talking about layer two or layer three, but then you're also talking about layer one, because now they actually own physical cables across the ocean and certainly they own data centers.
- 20:04
- Unidentified speaker: So to kind of answer your question, it's sort of like I can tell you, I can talk to you about the physical pieces of it, the layer zero, layer, layer one, and a little bit of layer two.
- 20:12
- Unidentified speaker: But when we get up to the kind of um, you know, how is the how does this encoding work?
- 20:16
- Unidentified speaker: You know, what's how do we, you know, what is the kind of theory of information that that sort of compresses, you know, the the output from our cameras and from our microphones and then and then transmits it across that.
- 20:26
- Unidentified speaker: That's more like a layer three or you know, you know, Riverside would be like a layer four, you know, that an application conversation.
- 20:33
- Unidentified speaker: Um so it's kind of interesting.
- 20:34
- Unidentified speaker: It's like a the geography and the physicality of it is kind of you know disintermediated from the from the sort of logic of it, from the software of it.
- 20:44
- Unidentified speaker: And I think that's kind of one of the keys of understanding how the physical infrastructure of the internet works is that is that these are two separate realms, the kind of physical and the digital, um, but they absolutely need each other to operate.
- 20:56
- Unidentified speaker: Hmm.
- 20:57
- Unidentified speaker: So am I correct in saying that uh we're sort of talking over the internet, but our voices are being converted into light and then traveling through those cables?
- 21:10
- Unidentified speaker: Is that sort of what's happening?
- 21:11
- Unidentified speaker: Oh, absolutely.
- 21:12
- Unidentified speaker: Absolutely.
- 21:13
- Unidentified speaker: Yeah, yeah, yeah, yeah.
- 21:14
- Unidentified speaker: And and not only that, but as speeds increase, where once upon a time our it wasn't once upon a time, maybe 10 years ago, 15 years ago, our our our the encoding of our voices was not converted to light until it crossed the ocean.
- 21:31
- Unidentified speaker: I wouldn't be surprised if um I'm actually quite sure that there's a box on the attached to the building I'm in uh that is converting the cop the cut the signals traveling over the copper wire that the modem on my desk is connected into into pulses of light.
- 21:48
- Unidentified speaker: And once it's converted to play, it's probably not converted back again until it hits your desk even.
- 21:52
- Unidentified speaker: Um you might even you probably even have a fiber connection at home.
- 21:55
- Unidentified speaker: So it's kind of light all the way through.
- 21:57
- Unidentified speaker: I mean, there are moments where it's tran it's it's goes back into a circuit board and is converted from light to ele uh to to electrons and then back to light.
- 22:06
- Unidentified speaker: Um, but for a surprisingly long stretch of that journey, um, it is pulses of light.
- 22:11
- Unidentified speaker: I mean, that is the most efficient way to encode the information by far.
- 22:15
- Unidentified speaker: It was worth saying there too, it's like the it's not just light, it's many, many, um, you know, it's it's uh it's it's many frequencies um or colors of light, and and there's so much space there um that you you know that you it's it's kind of you know astonishing how much information can be can be can be uh can be uh can be encoded there.
- 22:36
- Unidentified speaker: So yeah.
- 22:37
- Unidentified speaker: I don't know if it's just me, it must also be you, otherwise you wouldn't have done all this research.
- 22:42
- Unidentified speaker: But sometimes it just amazes me what we come up with as a human species.
- 22:47
- Unidentified speaker: It's just if you think about it, like I don't know if people think too much about this, but I really do.
- 22:53
- Unidentified speaker: It's it's just looking back when we were like primates, we made fire, and now we can make like these zeros and ones and light and turn it into the internet and into visuals and sounds, it just amazes me.
- 23:08
- Unidentified speaker: I mean it's pretty, it's pretty, it's pretty incredible.
- 23:11
- Unidentified speaker: I mean, I you know, it's like it was a leap.
- 23:14
- Unidentified speaker: I mean, I remember the transition from phone to internet and then from internet to to video.
- 23:22
- Unidentified speaker: And and you know, that a phone call was like a specific kind of magic.
- 23:26
- Unidentified speaker: It was like you could be, I remember, you know, being like a college student in Europe and you know, calling home and like, you know, if you if you put in your physical phone card, it would connect you with, you know, the the money would tick down.
- 23:38
- Unidentified speaker: But like, you know, you could your family would be in the other, on the other end of the line.
- 23:42
- Unidentified speaker: And anywhere you were, um, and that was by the time I was around, that was pretty cheap.
- 23:46
- Unidentified speaker: Um, but but it was still a sort of pre-cell phone.
- 23:48
- Unidentified speaker: And then by the time cell phones come in, then you then it's like you any, you can be anywhere and anyone else can be anywhere, and then you have that connection, you know, between you.
- 23:56
- Unidentified speaker: And suddenly, you know, it's not about making a sort of a technological connection between two places, but a technological connection between two people.
- 24:06
- Unidentified speaker: And for me, that was the kind of moment when when we started when we began to forget about the physicality of it.
- 24:13
- Unidentified speaker: Because if you've got a telephone on the wall, you know, that's that's plugged in, then you're you're pretty tuned into the way that um that there is a physical connection between you.
- 24:24
- Unidentified speaker: But when it becomes Wi-Fi and it becomes a kind of uh a sort of communication portal that you bring with you, you know, it never mattered, your email never mattered cared where you were, you know, you always just kind of got it, you you it found you wherever you were.
- 24:37
- Unidentified speaker: Um, I mean, that's kind of a hard thing to remember that that was like a thing, but that was sort of surprising.
- 24:42
- Unidentified speaker: And as soon as you get that, then you begin to lose this connection with the physical connection, um, you know, the physicality of the kind of space in between.
- 24:50
- Unidentified speaker: Um, but that was really a lot of what I was after was trying to understand how um, you know, it's sort of bring that physicality back to kind of remember that like, you know, of course, we're we're sitting in a specific place on Earth, you know, you know, New York and and and then and I don't know if you're if you're in Amsterdam proper or not, but like are pretty special, you know, unusual places, you know, specific places.
- 25:10
- Unidentified speaker: Um, and it's it's tough to, you know, it's yeah, we spend a lot of time forgetting that that matters, um, but because of the sophistication of these digital tools, but it it certainly it certainly still is there.
- 25:23
- Unidentified speaker: Yeah, when we started um emailing each other, or well, I started emailing you.
- 25:28
- Unidentified speaker: You told me that Amsterdam has a special place in your hearts.
- 25:31
- Unidentified speaker: Can you uh elaborate on why?
- 25:34
- Unidentified speaker: Yeah, I mean, Amsterdam was probably one of the places that was most successful at making itself an internet hub.
- 25:42
- Unidentified speaker: Um, you know, in New York, uh New York is special too, you know, if you knew Amsterdam.
- 25:49
- Unidentified speaker: Um, but it, you know, New York was a is a capital of finance, you know, was kind of self-evident that the the networks would find New York, but Amsterdam um was more deliberate in sort of in sort of building and and sort of seeding the infrastructure so that the that the infrastructure so that so that the networks found it.
- 26:08
- Unidentified speaker: You know, it wasn't just like, oh, you know, that we have to build the networks to the to Wall Street.
- 26:13
- Unidentified speaker: In Amsterdam, it was really a deliberate policy of laying as much fiber in the ground as possible, um, allowing all the different anyone who wanted to, whenever there was a trench in the ground, anyone who wanted to could then lay fiber in that trench as well.
- 26:30
- Unidentified speaker: And so you, by allowing that to happen, first of all, you don't have to dig up the street as often, but second of all, you end up with a ton more fiber to work with.
- 26:38
- Unidentified speaker: And while in the beginning, and this is kind of in the late, late 90s, early 2000s, that seemed like overkill.
- 26:43
- Unidentified speaker: Pretty quickly, all that fiber was the enabling infrastructure for Amsterdam to become a major point of interconnection for the networks of the internet.
- 26:54
- Unidentified speaker: And add on top of that, the kind of Dutch tradition of being a meeting place or a marketplace and of being, you know, uh accepting of everyone, uh, especially for for trade, for commerce.
- 27:06
- Unidentified speaker: And so you end up with um the uh Amsterdam Internet Exchange, AMSIX, being a kind of really special, really unusual um, you know, hub for the networks of the internet to connect with each other.
- 27:17
- Unidentified speaker: And that wasn't, that wasn't, that didn't kind of happen by accident.
- 27:20
- Unidentified speaker: Like that happened very deliberately by a pretty small group of people pursuing a set of policies and allowing infrastructure to be built that were very much in keeping with the long Dutch tradition of being a meeting place and a harbor, you know, not for ships, you know, not for uh infrastructure of ships, but for the infrastructure of bits.
- 27:41
- Unidentified speaker: And so I just I loved that, you know, that I mean, it's as a New Yorker, there's a lot of close associations, but I loved um in my exploration of the internet, I loved the way in which the Amsterdam and the Dutch more broadly were were sort of open to the op open to the openness of the internet.
- 27:56
- Unidentified speaker: Um and not only that, but organized and deliberate enough to build the infrastructure that would actually allow the kind of uh business, you know, business construct of the of the internet to to happen um in the Netherlands.
- 28:10
- Unidentified speaker: Yeah, pretty fascinating.
- 28:11
- Unidentified speaker: Yeah.
- 28:12
- Unidentified speaker: It's uh apparently it's somewhat in our uh in our DNA.
- 28:16
- Unidentified speaker: And uh Certainly.
- 28:18
- Unidentified speaker: I I I would have never thought of I didn't even know this was a a thing in Amsterdam until uh I read it in your book and I did a little uh more research.
- 28:26
- Unidentified speaker: So uh where did um all your research bring you?
- 28:29
- Unidentified speaker: Can you give me like a a summary of of some of the more known or cool places you've been to around the world?
- 28:37
- Unidentified speaker: Yeah.
- 28:38
- Unidentified speaker: I mean, I you know it's interesting.
- 28:40
- Unidentified speaker: It it was at the time, and this is you know more than 15 years ago now, um, there were fewer places, there were sort of fewer capitals of the internet.
- 28:51
- Unidentified speaker: Um I mean, I suppose there's still a hierarchy of the most important places, um, but so many places have gotten so built up that I think that's it's it's harder to sort of keep track of that.
- 29:01
- Unidentified speaker: Um at the time though, um the top of the list by far was was um Frankfurt, London, Amsterdam.
- 29:10
- Unidentified speaker: Yeah.
- 29:11
- Unidentified speaker: Um those were, you know, those were those were the those were the hubs.
- 29:14
- Unidentified speaker: Um, you know, they're they are the you know financial hubs, they're the telecommunications hubs, and they became the internet hubs.
- 29:20
- Unidentified speaker: And then in the US, a little bit more complicated, um, but it was New York, Northern Virginia, um, for reasons we can talk about, uh, LA and Silicon Valley primarily, um, a little bit of Miami, Dallas, Chicago, um, but those were those those are the most important places.
- 29:36
- Unidentified speaker: And then the other kind of key, key kind of list of key type of places that I that I visited were the ones that that serve the undersea cables across the ocean.
- 29:48
- Unidentified speaker: And there you're talking sometimes about the same places, about big port cities, um, but there's also a kind of um, you know, uh more specific geography that the cables, you know, need that's separate from the ports.
- 29:59
- Unidentified speaker: Um so that uh uh that in one case was the coast of Nova Scotia, um, you know, where kind of closest, closest point to Europe, um, from one of the closest to sort of populated or sort of accessible places in the North America to Europe, um, or outside of Lisbon in Portugal, you know, which has always been a kind of port city for Africa and and and and and um and and the south.
- 30:22
- Unidentified speaker: And the same was true for the undersea cables, where Lisbon became the the um the connection point to a new cable down the west coast of Africa.
- 30:30
- Unidentified speaker: Aaron Powell You you didn't uh visit any Asian countries?
- 30:33
- Unidentified speaker: I didn't make it to I didn't make it to Asia.
- 30:35
- Unidentified speaker: I didn't make it to Asia.
- 30:36
- Unidentified speaker: The they were too too too complicated to bring that in as well.
- 30:40
- Unidentified speaker: I think you mentioned a couple uh places in the book, right?
- 30:43
- Unidentified speaker: Aaron Powell Yeah, I talk about the Asian countries a little bit.
- 30:45
- Unidentified speaker: Um but um you know Singapore is an important hub.
- 30:48
- Unidentified speaker: I didn't visit um for a couple reasons.
- 30:52
- Unidentified speaker: One, the Chinese internet was its kind of own ball of wax, you know, and and and there were um then and still now a lot of challenges in terms of reporting about that as a journalist, you know, going and asking a lot of questions.
- 31:04
- Unidentified speaker: Um and then the other, you know, Tokyo is a major Asian hub.
- 31:09
- Unidentified speaker: Um, but Tokyo, as is often the case, kind of has its own specific stories and its own specific way of doing things.
- 31:15
- Unidentified speaker: Um and so why that while that overlapped, um, you know, with the way it's done, other places, um, it was just a it was a a different story, a different thread of this of the narrative, where Frankfurt and Amsterdam um operated very similarly.
- 31:29
- Unidentified speaker: The people who ran them were very close with each other, close competitors, um, close colleagues.
- 31:34
- Unidentified speaker: Uh, and then and then London was a sort of um you know opposition to the two of them in some ways, all competing to be the most hospitable place on the internet for networks to connect to each other, to charge a toll in a certain way when that happens, um, but then also to recognize that their growth is is mutually beneficial.
- 31:54
- Unidentified speaker: As one grows, the others grow?
- 31:56
- Unidentified speaker: Please correct me if I'm wrong, but the internet only works, and I believe you t you you say this in the book as well, because it is sort of based on trust.
- 32:06
- Unidentified speaker: You are where you say you are, and the networks connect, and that way it sort of works.
- 32:11
- Unidentified speaker: Does that mean that if somebody really wants to find out where you are, like where something is, there's always a way?
- 32:19
- Unidentified speaker: Or because the openness is like openness and trust is so important for the internet, but at the same time, everybody wants sort of privacy and is scared to get found online and stuff.
- 32:31
- Unidentified speaker: Do you get what I'm my question is, sort of?
- 32:34
- Unidentified speaker: Yeah, yeah, absolutely.
- 32:35
- Unidentified speaker: I mean, it's interesting.
- 32:36
- Unidentified speaker: One of the, you know, that that sort of openness and trust on the internet is kind of at the highest level where the networks have to be found.
- 32:46
- Unidentified speaker: Like if you're, you know, you can't, you know, we can't communicate with each other if my network doesn't know where your network is.
- 32:52
- Unidentified speaker: So there has to be a kind of, you know, but then you're talking about going up the, going up the stack, going to different applications.
- 32:58
- Unidentified speaker: And there you have, you know, in a sort of consumer way, layers of privacy added into that.
- 33:04
- Unidentified speaker: Layers of privacy and occasionally, you know, layers where, you know, moments where it's less about um, you know, it's it's it's less about privacy and more about um uh more about sort of a verification of who you are.
- 33:16
- Unidentified speaker: Um so yeah, it's I mean it's a complicated tension, but but but it's I think the important thing is to separate some of the applications, not just some of the applications, but the applications that have become really dominant, you know, particularly social media in different ways, um, which is very centralized.
- 33:31
- Unidentified speaker: You know, it's a small number of dominant social media companies, um, with the original underlying sort of philosophy of the internet itself.
- 33:40
- Unidentified speaker: Um and they're really two totally separate things.
- 33:43
- Unidentified speaker: So the cloud.
- 33:44
- Unidentified speaker: Yeah.
- 33:44
- Unidentified speaker: Andrew, please tell me, is is everything we save online, is that stored in a in like a real cloud?
- 33:50
- Unidentified speaker: Yeah.
- 33:50
- Unidentified speaker: Well, a real cloud, a real data a real a real hard drive, a real data center.
- 33:55
- Unidentified speaker: Like in the sky?
- 33:56
- Unidentified speaker: As clouds in the sky?
- 33:58
- Unidentified speaker: No, not uh no, in a big warehouse with just with with hard drives and you know that take electricity and have to be cooled and um have to be physically connected to the rest of the internet.
- 34:09
- Unidentified speaker: And of course those Come on.
- 34:11
- Unidentified speaker: Really?
- 34:11
- Unidentified speaker: Is that how it works?
- 34:14
- Unidentified speaker: Yes.
- 34:14
- Unidentified speaker: So no, no, no, uh so it's a it's a um, you know, that but the cloud, I mean the cloud was originally in kind of marketing terms, just recognize that these things are always where they are.
- 34:24
- Unidentified speaker: Um you know, these things are we that we hate that they're accessible everywhere.
- 34:28
- Unidentified speaker: Um, but that obscures the physicality of them.
- 34:30
- Unidentified speaker: And of course, now in the AI conversation, uh it's gotten a lot more uh you know, it's gotten a lot more intense in the recognition that these data centers that power these tools take a lot of power, occasionally use a lot of water.
- 34:45
- Unidentified speaker: You know, the power they need has consequences for for carbon emissions, um, has consequences for people's costs on the grid, um, you know, costs of their electricity.
- 34:54
- Unidentified speaker: So it's, you know, now we're we're I think we're getting back in touch with the fact that the the the cloud isn't just you know ethereal, but has major consequences and sort of and sort of uh you know has uh you know has major impacts in all sorts of ways.
- 35:09
- Unidentified speaker: Can you paint uh I think you've been into uh several data centers?
- 35:14
- Unidentified speaker: Aaron Powell Yeah, lots of data centers.
- 35:15
- Unidentified speaker: Yeah.
- 35:16
- Unidentified speaker: Can you describe what it's like to you you you open the door, I imagine there's a door.
- 35:21
- Unidentified speaker: Uh what it what is it like in there?
- 35:23
- Unidentified speaker: Is it hot?
- 35:24
- Unidentified speaker: Is it there many people?
- 35:26
- Unidentified speaker: Can you paint us a picture?
- 35:27
- Unidentified speaker: Aaron Powell Yeah.
- 35:28
- Unidentified speaker: I mean the data centers um they're often really loud.
- 35:32
- Unidentified speaker: Um they're always very loud uh because of the fans and the equipment.
- 35:36
- Unidentified speaker: Um they sometimes are dark because there's aren't that many people in them, so they keep the lights off.
- 35:42
- Unidentified speaker: Um they have a really particular smell.
- 35:44
- Unidentified speaker: It's almost like a like a toast, like a burnt toast smell.
- 35:47
- Unidentified speaker: That's sort of a common.
- 35:48
- Unidentified speaker: Yeah, yeah.
- 35:49
- Unidentified speaker: It's it's it's very distinct.
- 35:50
- Unidentified speaker: It's kind of like a combination of the of the all the equipment and the kind of like ions from the equipment or and all the air conditioning.
- 35:57
- Unidentified speaker: Um, but you can kind of you there's a really specific, distinct smell to a data center.
- 36:01
- Unidentified speaker: Occasionally, like walking outside a telephone building, you get a whiff of it.
- 36:04
- Unidentified speaker: Um but as soon as you walk in, it's always the same smell.
- 36:07
- Unidentified speaker: Um, and then um, you know, and then there's often some uh not very happy people uh kind of stooped over you know their equipment trying to fix some problem.
- 36:17
- Unidentified speaker: You know, if there are people in there, either something has to be set up or something's gone wrong.
- 36:22
- Unidentified speaker: Uh and um, you know, they these are the kind of, you know, they it still has to be done by hand a lot of the time.
- 36:28
- Unidentified speaker: And so you you never visit a data center that's empty.
- 36:30
- Unidentified speaker: There's always a couple, couple cages um with people, with technicians kind of hard at work trying to solve some problem at all hours of the day, uh, often sort of looking really frustrated, sometimes wearing like ear protection because it is so loud.
- 36:43
- Unidentified speaker: Uh, always wearing uh, you know, a fleece because um because it's freezing cold in the air for the equipment.
- 36:54
- Unidentified speaker: Hi, this is Ross, the host of Smells Like Humans.
- 36:58
- Unidentified speaker: Each week we talk about the curious things that people do.
- 37:02
- Unidentified speaker: Fans say it's like hanging out with your funny friends.
- 37:05
- Unidentified speaker: If you like light-hearted conversation, personal stories, and a hint of psychology, chances are we are discussing something relevant to you, like relationships, dreams, phobias, weddings, work, cults, a hundred topics and counting.
- 37:21
- Unidentified speaker: Sometimes silly, sometimes serious, but always fun.
- 37:25
- Unidentified speaker: Please join us because no matter who you are, we all smell like humans.
- 37:30
- Unidentified speaker: Please click the link in the show notes.
- 37:39
- Unidentified speaker: Wow, okay.
- 37:40
- Unidentified speaker: So there's that's pretty it's a good picture.
- 37:44
- Unidentified speaker: Um there's this incredible thing about uh uh the infrastructure we made that made this possible.
- 37:52
- Unidentified speaker: Like the the fact that we just can uh do this now on our podcast, people from all over the world can listen to this.
- 37:59
- Unidentified speaker: 30, 40 years ago it would have been a totally different story.
- 38:04
- Unidentified speaker: When you think back to the early days of the internet, you might remember that.
- 38:09
- Unidentified speaker: Is there what did people believe um back then how it would change the world?
- 38:15
- Unidentified speaker: Because now AI is like uh a big thing, and one of the big fears is AGI, where we we get replaced, right?
- 38:25
- Unidentified speaker: I can imagine something similar back in the days with the internet coming up.
- 38:30
- Unidentified speaker: Do you remember what what people believed back then?
- 38:32
- Unidentified speaker: Oh, that's a good question.
- 38:34
- Unidentified speaker: It's funny.
- 38:34
- Unidentified speaker: I haven't thought about I haven't thought about what how what we're doing matches the original terms.
- 38:39
- Unidentified speaker: You know, sorry, the guy I haven't thought about how the conversation that happening now about AI and AGI lines up with that original sense of what the internet could do.
- 38:50
- Unidentified speaker: I think that it was much it was much more optimistic than the AI conversation, certainly.
- 38:55
- Unidentified speaker: There was really a sense that um, you know, we talked a lot about the the digital divide, meaning that if people and parts, if people don't have access to the internet, then they will fall behind economically, which which was true.
- 39:07
- Unidentified speaker: But that there wasn't there wasn't really a whole set a lot of a sense of the downside or of any downside of connecting to the internet.
- 39:15
- Unidentified speaker: It was there was a recognition that if you could connect to the internet, then you would have access to all sorts of positive information.
- 39:21
- Unidentified speaker: You could learn anything, you could, you could, you could look up anything, you could, you could search for anything, uh, you could connect with anyone.
- 39:28
- Unidentified speaker: So you would have an infinite global market, no matter, you know, whatever you were doing.
- 39:32
- Unidentified speaker: I mean, all those things are true, but one of the things that has happened is that infinite global market has been, you know, primarily kind of gobbled up by a relatively small number of companies.
- 39:41
- Unidentified speaker: I mean, I guess that's the the sort of nature of capitalism, that that's that that is that is the goal.
- 39:47
- Unidentified speaker: But rather than the sense of of a kind of you know really diverse, you know, competitive, but sort of, you know, really lively combination of voices, we now have a kind of, you know, we have a a sort of we we live under the control of a smaller number of very large global corporations.
- 40:03
- Unidentified speaker: And so I think when you to think about like what the optimistic view of that was, you know, it was really about a kind of recognition that connecting people, connecting all kinds of people and expanding markets was always going to be positive.
- 40:16
- Unidentified speaker: Um, you know, I it's I mean we can talk about sort of what how that's changed or what's what's what's scared people about it.
- 40:22
- Unidentified speaker: But for me, and this is a kind of internet infrastructure point, one of the more dramatic things that I didn't expect was the extent to which the hyperscalers, as we call them now, you know, primarily, you know, Google, Google and Meta, but would not only kind of dominate the application space, but would actually build their own physical infrastructure.
- 40:41
- Unidentified speaker: So they they have over the last 10 years actually built their own cables across the ocean, certainly built their own data centers, built their own networks, so that they're kind of you know eating the infrastructure from above.
- 40:51
- Unidentified speaker: Um, you know, it's not it's not just that like they're riding on top of some telecom infrastructure, but they are actually kind of taking over that infrastructure themselves, um, build rebuilding it, um, owning all of it, sort of top to bottom.
- 41:03
- Unidentified speaker: Um, and that that was a surprise.
- 41:05
- Unidentified speaker: And and that's that's been really consequential.
- 41:07
- Unidentified speaker: And I think it's one of the reasons that the AI race is so intense right now is because you know it threatens, uh, it threatens the internet that they've built with with some some new innovation.
- 41:19
- Unidentified speaker: What remains to be seen is because AI requires so much physical infrastructure, so much power to run, um it's it's it's a real open question of how that how that that that's then re-rebuilt, how we then sort of build that physical infrastructure to serve uh these new tools, and then what happens to the to the sort of quasi-monopolies that exist already.
- 41:43
- Unidentified speaker: You've mentioned the hyperscalers a few times now.
- 41:45
- Unidentified speaker: So what's interesting to me is that companies like Google and Amazon started by building businesses on top of the internet.
- 41:53
- Unidentified speaker: Uh and like you mentioned, they're now building much of the internet's infrastructures themselves.
- 41:59
- Unidentified speaker: Is that something you worry about?
- 42:02
- Unidentified speaker: Uh I mean, I worry about it, yes.
- 42:05
- Unidentified speaker: Uh, it is something I worry about.
- 42:07
- Unidentified speaker: Um, I worry, you know, they I think we we have we're perhaps well, I won't say lucky, but we we have the natural check on these companies of being public companies that have a responsibility to their shareholders to to to a certain extent disclose what they're doing, to be subject to a court of public opinion, to be subject to litigation, which you know in Europe happens again and again and again.
- 42:34
- Unidentified speaker: You know, they have to be responsible actors um in order to kind of operate in so many places.
- 42:40
- Unidentified speaker: And so I think there's in a certain to a certain extent, those forces, both those market forces as and those regulatory forces, have have kind of kept them in their lane um in terms of privacy, in terms of uh harms to children, uh, you know, things like that.
- 42:56
- Unidentified speaker: Um but um, you know, but still with major impacts um that we're kind of still coming to terms with.
- 43:02
- Unidentified speaker: And what's so dramatic with the current conversation about about the about AI is those companies are not yet have not yet shown themselves to be responsible public actors.
- 43:15
- Unidentified speaker: They're not accountable yet um to uh to the market, and they're not accountable yet to regulation, although that is coming pretty quickly, but no one quite knows how how to put it in as fast as it needs to go in.
- 43:28
- Unidentified speaker: And so they and then they don't quite know what they have, which we're in a strange position to be in.
- 43:33
- Unidentified speaker: Like they don't quite even know what what they've created and what it can do, um, which is very different from Instagram, for example.
- 43:40
- Unidentified speaker: You know, it's like it's like you know, Instagram is has, you know, might might uh you know be bad for children or or or uh or the newspaper business or or um you know our or or our attention spans, but we kind of, you know, but that sort of the possibilities for what it does to society are a little bit you know, a little bit easier to grasp than the kinds of questions we're asking about about AI.
- 44:01
- Unidentified speaker: Aaron Ross Powell If you were writing tubes today, um not about tubes, but about AI, what are some places you would go then?
- 44:10
- Unidentified speaker: It's a great question.
- 44:10
- Unidentified speaker: I mean, you know, in some ways tubes needs a sequel because the the construction of data centers, especially in the US, I mean, and and we can talk about in the moratorium in the in in the Netherlands, but but the construction of data centers in the US is is taking up all the air.
- 44:25
- Unidentified speaker: Um, you know, it's an incredible uh economic story, um, you know, with real risks of interdependencies that could lead to a real systemic economic collapse.
- 44:37
- Unidentified speaker: Um and uh it's a real social story where there are a lot of communities that are fighting over um whether or not they want a data center in their town.
- 44:47
- Unidentified speaker: Um and, you know, I think there's gonna there's a sort of realignment of what the benefits are to communities um and what the harms are to communities when a big data center comes in.
- 44:57
- Unidentified speaker: I mean, I'd be curious, you know, I know Amsterdam, I know the Netherlands had this conversation and has based essentially a moratorium um on data center construction, which uh at one point I understand was based on land and then now is sort of more urgently based on power.
- 45:12
- Unidentified speaker: Um, and and those are, you know, that's that's a more extreme response.
- 45:16
- Unidentified speaker: And there's some communities in the US that are responding that way.
- 45:18
- Unidentified speaker: Um, but the economic forces are very strong and the political forces are very strong as well in their eagerness to get these the giant data centers constructed that we need, quote unquote, need uh to power AI.
- 45:34
- Unidentified speaker: You said that um the internet sort of gave more power to uh information, more freedom.
- 45:43
- Unidentified speaker: Um do you think the internet gave if you had to give an answer, do you give do you think the internet gave ordinary people more power or do you think the internet actually gave more power to uh these large companies?
- 46:02
- Unidentified speaker: Well that's a tough one.
- 46:03
- Unidentified speaker: I mean, I uh I mean I mean it depends.
- 46:07
- Unidentified speaker: Do you well like do you like certainly influencers or YouTubers, you know, it's a kind of it's just definitely a sort of story of disruption of the kind of institutions of of media.
- 46:18
- Unidentified speaker: Um but you know, but to what at what cost?
- 46:22
- Unidentified speaker: I I don't know.
- 46:23
- Unidentified speaker: It's it's so it's so hard to say because in in some ways they the platforms, the corporations, the hyperscalers control so much, but in other ways they do open up the possibility of of anyone to say anything, and there is a kind of ground up, you know, uh you know, coordination and communication and sort of you know social movements that come out of that.
- 46:43
- Unidentified speaker: So I you know it's a I it's very hard to balance.
- 46:45
- Unidentified speaker: It's certainly a bunch of bunch of bunch on both sides of the the ledger.
- 46:49
- Unidentified speaker: Aaron Ross Powell Am I correct in saying that the internet is a lot more fragile than people think?
- 46:54
- Unidentified speaker: Like there's these undersea cables, and uh these cables have to come ashore somewhere.
- 47:02
- Unidentified speaker: Um looking at the Baltic Sea, for example, with Russia, there's all these incidents involving undersea cables, and even if it's not sabotage or whatever, anchors or fishing equipment can also do accidental damage.
- 47:15
- Unidentified speaker: W what what happens when uh when these cables break?
- 47:22
- Unidentified speaker: Uh they well traffic is primarily rerouted through other cables.
- 47:30
- Unidentified speaker: Like the the system is designed to re to to heal itself, but but there's a limit to that, of course.
- 47:36
- Unidentified speaker: You know, if more than one cable breaks and each cable is occup is uh is operating near capacity, then you know it it's not enough to have a single backup.
- 47:45
- Unidentified speaker: You need multiple backups and other places to go.
- 47:47
- Unidentified speaker: And we've seen that in recent years with different cable cuts, particularly in the Middle East and in the Red Sea, um, in the Red Sea and in and through the through the um the Suez and to Africa where you have multiple cable cuts that have a real impact on on the connectivity in the in those regions.
- 48:04
- Unidentified speaker: Um that's less likely in Europe where there's so much connectivity and there's so many cables.
- 48:08
- Unidentified speaker: Um, but there are still plenty of places where um the connect where the connections are thinner or more fragile.
- 48:15
- Unidentified speaker: And we've seen that recently, I think was it last year or the year before, with the cable cuts in the Mediterranean, where the internet in um in parts of Africa to the Middle East were was severely degraded.
- 48:25
- Unidentified speaker: Um and uh and that that that you know they're they actually couldn't function in the same way.
- 48:31
- Unidentified speaker: But that's uh that's a familiar story, right?
- 48:33
- Unidentified speaker: I mean, you know, infrastructure is more brittle in some places than in others.
- 48:38
- Unidentified speaker: You know, the the you know, the trains, you know, you know, the Netherlands is a great example of a of a more of a more resilient infrastructure in lots of ways.
- 48:46
- Unidentified speaker: Um, you know, just a lot of can a lot of effort and and money and engineering going into uh into sort of making infrastructure of all kinds stronger, you know, trains, planes, you know, shipping, whatever it is, roads, um, where in parts of the world there isn't as much money or attention or engineering paid to it.
- 49:04
- Unidentified speaker: And that's true, that's true for all kinds of infrastructure, and it's certainly true for the internet.
- 49:08
- Unidentified speaker: Yeah, definitely.
- 49:10
- Unidentified speaker: Your book also made me think about something beyond technology.
- 49:13
- Unidentified speaker: Um you talk about the internet as if it belongs to everyone, um, but it depends on land, buildings, companies, cables, yada yada yada.
- 49:25
- Unidentified speaker: Um and all those are very real sources of power.
- 49:30
- Unidentified speaker: Has the internet made us more independent, or have we just become dependent on things we can't really see?
- 49:39
- Unidentified speaker: Oh, I don't know.
- 49:40
- Unidentified speaker: That's a tough question.
- 49:41
- Unidentified speaker: I mean, I think I mean, I think we're we're certainly dependent on the internet.
- 49:46
- Unidentified speaker: Um, but I think that there is still a a sort of fundamental kind of competition that happens on the internet, you know, that that creates more more resiliency or more security.
- 50:00
- Unidentified speaker: Um, you know, there there are, yes, the systems are really entangled with each other.
- 50:04
- Unidentified speaker: And yes, occasionally they do seem fragile, like we, you know, the moments when uh some piece of software running the airline industry breaks and then suddenly planes all over Europe or all over the world are grounded, you know, there you see repercussions like that.
- 50:17
- Unidentified speaker: Um, but then there are, you know, but the economic consequences of that, you know, the the failure of one company becomes the opportunity of another company.
- 50:25
- Unidentified speaker: And it really is, it really is interesting to me that it's that all of these things are dependent, all of the kind of workings of this infrastructure, you know, while they are often seeded by governments, um, they are really operated by private companies and private companies competing with each other to offer the the best and most robust services.
- 50:45
- Unidentified speaker: Um we're dependent on all of these interconnected services and networks.
- 50:51
- Unidentified speaker: Um, but we also do, you know, there is a sort of um freedom in the choice to switch from one to the other, um, you know, from one from one competitor to another competitor often.
- 51:02
- Unidentified speaker: You know, if if Riverside stops working, I think you do have an extra choice.
- 51:05
- Unidentified speaker: Though I guess Riverside is a pretty pretty pretty dominant, pretty dominant player in uh podcasting as an example.
- 51:12
- Unidentified speaker: Yeah.
- 51:12
- Unidentified speaker: Yeah, I don't know.
- 51:13
- Unidentified speaker: Uh no, uh my my world is stopping when uh Riverside breaks.
- 51:17
- Unidentified speaker: Right, right, exactly.
- 51:18
- Unidentified speaker: Exactly.
- 51:21
- Unidentified speaker: You know, after spending so much time exploring the internet, are you more optimistic or more worried about the future of the internet, the the you the infrastructure, the physical infrastructure, perhaps include AI in in there as well?
- 51:37
- Unidentified speaker: Yeah.
- 51:38
- Unidentified speaker: I don't know.
- 51:38
- Unidentified speaker: I mean, I I mean you said before it's pretty amazing that we've humans have built this technology.
- 51:44
- Unidentified speaker: I mean, my my hope is that humans are pretty good at building things and pretty good at our instincts for self-preservation.
- 51:51
- Unidentified speaker: And so my hope is that um, you know, with these new technologies, you know, our our our you know, they them either the technologies themselves or our ability to harness them creates um you know creates a uh a resiliency, you know, creates an ability to kind of you know live through, you know, survive whatever comes um faster than the ability to to destroy ourselves.
- 52:13
- Unidentified speaker: So, you know, so we uh you know that that has that has been the case in the um in the 80 years, you know, since um you know since of atomic weapons.
- 52:24
- Unidentified speaker: Um you know the hope is that that's also the case where you know, yes, we have a technology to to potentially destroy ourselves, but also we have the ingenuity to to prevent that from happening as well.
- 52:38
- Unidentified speaker: Um I mean I'd like to think that optimistic that optimistic view is more is more the case.
- 52:44
- Unidentified speaker: But um but I don't I don't discount the um I don't discount the potential of AI to really cause major disruptions in the way the way we do all sorts of things.
- 52:55
- Unidentified speaker: Um you know it's it's just seems like there's so many it's there's it it can do so much, and we we just we it's gonna be really interesting to see how the things it can do changes what what humans do and what purpose we we put it towards.
- 53:10
- Unidentified speaker: Yeah, I've always been quite optimistic about technology.
- 53:14
- Unidentified speaker: I I I I do believe that throughout history, I really like looking at history, it's a good gives us good guidelines, that technological progress has made our lives better.
- 53:27
- Unidentified speaker: And obviously it always brings risks, and there are always people who la who lose out along the way.
- 53:32
- Unidentified speaker: But when you look at the bigger picture, I think uh we have to be positive, and I think AI will prove to be no different here.
- 53:41
- Unidentified speaker: But um, you know, technology doesn't automatically benefit everyone because uh it has for example, the internet has created enormous value, but the value and power it brought became pretty concentrated in a couple people and uh companies.
- 53:57
- Unidentified speaker: So that's no, absolutely I agree.
- 54:00
- Unidentified speaker: Yeah, no, and that's and that's what that's been I mean, it's one of my one of for me, the dominant story when we talk about the tubes, when we talk about the physical infrastructure, is that kind of replacement of a you know of a of many hundreds or thousands of networks competing with each other with really just a few networks, you know, dominating things and everybody else kind of occupying a lesser space.
- 54:21
- Unidentified speaker: And you're right to point out that um that that that sort of dominance of those networks then translates into the into the concentration of power in a small number of people.
- 54:30
- Unidentified speaker: Yeah.
- 54:32
- Unidentified speaker: Yeah, maybe the interesting question isn't whether technology makes us more capable, because it clearly does, but whether being more capable necessarily makes our lives better.
- 54:44
- Unidentified speaker: But um yeah.
- 54:47
- Unidentified speaker: Let's uh work towards an ending, Andrew.
- 54:50
- Unidentified speaker: Um what do you th wish more people understood about the internet?
- 54:58
- Unidentified speaker: I think I wish more people understood that um that it's still there underneath and that there is a way to engage with it independently.
- 55:11
- Unidentified speaker: There's a way to engage with it not on YouTube and not on Instagram, but like there's still a kind of independent internet that operates beneath all of that.
- 55:20
- Unidentified speaker: That's it's possible to kind of still still uncover and still still exist on.
- 55:24
- Unidentified speaker: Um there might be costs to that and it might might be sort of more challenging.
- 55:29
- Unidentified speaker: Um but um but I think there's a lot of benefit in sort of maintaining the vitality of that that sort of you know independent, your small scale um, you know, kind of in internet infrastructure that uh happens underneath, whether in your choice of email provider, you know, for example.
- 55:46
- Unidentified speaker: I mean, like, you know, Gmail is so dominant, but it's like you can find a local email provider, you know, and as just as one example, um, or or even just as a in in a website, you know, you can still put up a website in the internet.
- 56:00
- Unidentified speaker: Like it's still, you know, that's that's still possible.
- 56:02
- Unidentified speaker: It's it's you know, it's dependent on all of the the big systems that that sort of help help people find it, but it you can still just kind of do it.
- 56:08
- Unidentified speaker: And so I think I wish, yeah, so I wish people had recognized that that that internet's still underneath and it's not only about Google and Meta and Amazon, but um, but there's still that that old thing, that old piece, those old pieces are still there, which I think is pretty, pretty fascinating.
- 56:22
- Unidentified speaker: Yeah, absolutely.
- 56:23
- Unidentified speaker: Yeah.
- 56:23
- Unidentified speaker: Looking back at um your journey, it's been uh it's been uh some time for you uh ago.
- 56:29
- Unidentified speaker: Looking back to your writing, to your research, to the center of the internet, what was um the most valuable lesson, or perhaps multiple smaller lessons you took away from writing the book and researching um the internet along the way?
- 56:45
- Unidentified speaker: That's a good question.
- 56:46
- Unidentified speaker: I think I mean I think I always come back to this.
- 56:50
- Unidentified speaker: Um I mean, it's it's it's run by people.
- 56:53
- Unidentified speaker: Like it's you know, you can you can think it's these giant companies, and you can think it's it's just you know, you know, you know, blinking lights or or you know, big data centers um or just like AI.
- 57:06
- Unidentified speaker: Um, but it's still there's still people who are who are building it and taking care of it and expanding it and and trying to control it.
- 57:13
- Unidentified speaker: Um and so I think that, I mean, I I kind of maintained that lesson, you know, certainly from my visit to the internet, from the, you know, it wasn't, you know, the buildings were important, but the kind of people who who take care of them were even more important and more interesting.
- 57:26
- Unidentified speaker: So it's like even when we have these these giant infrastructures, these giant technological constructions, it's still it's still run by people.
- 57:34
- Unidentified speaker: Like people are still the center of it.
- 57:35
- Unidentified speaker: It's one of the things that makes AI so, so, so challenging and so confusing is that it's sort of it's this there's still people involved, there's still people building it, but there's this other kind of whatever weird intelligence that we don't really know what to make of that isn't a person, um, but sort of acts like a person, and we don't really know how to understand that or how to interact with it or how to how to control it.
- 57:59
- Unidentified speaker: Looking back, reflecting on your process, um, what you just talked about, perhaps your research and the journey, is this something you would have done differently now?
- 58:08
- Unidentified speaker: Yeah.
- 58:08
- Unidentified speaker: I mean, there's there is one very specific thing, uh, which um which is that the that the the at the time the network operators were almost entirely men.
- 58:19
- Unidentified speaker: Uh there were very few women working in the space, um, and the men didn't always behave well.
- 58:25
- Unidentified speaker: And it was before the conversation about um about that about sexual harassment, um, about me too, as we call it here, um, it was before that conversation was more vibrant.
- 58:37
- Unidentified speaker: Um, and so I I kind of wrote about it a little bit, um, but I wish that I was clearer in sort of saying um that that it was that the that the the social environment of these people who I admire a lot for their construction, but the social environment um was um was uh was sexist, uh was often derogatory, um, you know, was uh was sort of unacceptable in its in its in the in the dominance of the men who who who who who ran it at the time.
- 59:10
- Unidentified speaker: Is that something you would have included in the book Looking Back or?
- 59:13
- Unidentified speaker: I think I I think I included it a little bit, but I was still speaking to those men.
- 59:20
- Unidentified speaker: Like I still, they were still my audience.
- 59:22
- Unidentified speaker: Like I still kind of wanted to make sure they approved of what I was saying, where in retrospect, um uh I could have been clearer about sort of uh sort of describing um what um where they uh where they were wrong.
- 59:38
- Unidentified speaker: Hmm.
- 59:38
- Unidentified speaker: Yeah.
- 59:39
- Unidentified speaker: Last one.
- 59:40
- Unidentified speaker: I would love to know uh what's next for you, Andrew.
- 59:43
- Unidentified speaker: I believe you're working um on a new book.
- 59:45
- Unidentified speaker: I saw some something on your website.
- 59:48
- Unidentified speaker: Um maybe you can tell me a bit more about that and uh Yeah.
- 59:51
- Unidentified speaker: When can we expect the expect the AI book?
- 59:54
- Unidentified speaker: Well, that's the problem, I think.
- 59:57
- Unidentified speaker: Um the new b the new book uh is about the electric grid and about the energy transition, that the shift from fossil fuels to to to renewable and carbon-free generation, um, and the way in which the grid adapts to that.
- 1:00:10
- Unidentified speaker: Um Netherlands are an amazing story with that, with the offshore wind and the offshore grid and um and the effort to decarbonize primarily through the construction of that incredible infrastructure.
- 1:00:20
- Unidentified speaker: Um the challenge um in the US at the moment is that uh we have had incredible whiplash in in the way that we talk about the electric grid and in the way that the electric grid is being built, where during the the Biden presidential administration, there was a huge effort to get as much solar and wind and batteries uh on the grid as possible.
- 1:00:44
- Unidentified speaker: Uh when Trump became president, all of that was not just slowed down, but actively stopped, um, not entirely successfully.
- 1:00:51
- Unidentified speaker: It continues certainly, um, but not as fast.
- 1:00:55
- Unidentified speaker: And then just as we have this incredible whiplash with a shift in policy, then we also had this crazy boom of AI where we have a growth in electro grid that we haven't seen in 50 years.
- 1:01:08
- Unidentified speaker: And so when I started this book, neither of those two things were happening.
- 1:01:11
- Unidentified speaker: Uh and um now I'm faced with trying to figure out what story I'm telling given this sort of whiplash of of changes.
- 1:01:20
- Unidentified speaker: So when I'll figure that out, uh hopefully hopefully this year, maybe next year.
- 1:01:25
- Unidentified speaker: You've got time.
- 1:01:27
- Unidentified speaker: All uh all good things take time.
- 1:01:29
- Unidentified speaker: They appreciate that.
- 1:01:31
- Unidentified speaker: The more we talked about the internet and the more I read uh in into your book, um the further I went, the more I realized, like you said, how human it still is.
- 1:01:42
- Unidentified speaker: We imagine this enormous automated system that somehow runs itself, but behind it are engineers, technicians, people repairing cables in the middle of the ocean, uh people deciding how these networks connect.
- 1:01:56
- Unidentified speaker: It's almost strange that something so important to modern life still depends on so many individual decisions and tiny things, and it's uh it's a fascinating book.
- 1:02:10
- Unidentified speaker: I would recommend anyone that is interested in knowing a little more about the internet and how it works and to get Andrew's book, Tubes, A Journey to the Center of the Internet.
- 1:02:20
- Unidentified speaker: And on that note, Andrew, thank you so much for uh for joining me.
- 1:02:25
- Unidentified speaker: Uh I think I'll be looking at the internet uh differently from uh for next time I open my laptop.
- 1:02:33
- Unidentified speaker: For anyone else listening, is there a place people can connect with you, find more uh on your research and your books?
- 1:02:40
- Unidentified speaker: Yeah, I mean I'm you I'm pretty easy to find online.
- 1:02:42
- Unidentified speaker: Uh Andrewblum.net is my website.
- 1:02:44
- Unidentified speaker: Um AJ Blum is my is my Twitter, um, my ex, I should say.
- 1:02:49
- Unidentified speaker: Um I spend a lot of time there.
- 1:02:50
- Unidentified speaker: I don't say a whole lot anymore.
- 1:02:51
- Unidentified speaker: Uh, LinkedIn, you can find me on as well.
- 1:02:53
- Unidentified speaker: Um I always welcome, I I I I love getting email from readers and um and people with questions.
- 1:03:00
- Unidentified speaker: Um uh my email address is is is on my website.
- 1:03:03
- Unidentified speaker: Um and um yeah no, I appreciate the conversation.
- 1:03:05
- Unidentified speaker: Appreciate your your your appreciation of the book.
- 1:03:07
- Unidentified speaker: It's it's amazing to me how vibrant the topic is um so many years later, even you know, even more important than ever, even.
- 1:03:13
- Unidentified speaker: I'll make sure to put all the links in our show notes and description.
- 1:03:17
- Unidentified speaker: Thank you, Andrew, and I'm looking forward to your next book, and I'm sure we'll get you on the pot again then.
- 1:03:22
- Unidentified speaker: I appreciate that.
- 1:03:23
- Unidentified speaker: Thanks so much for having me.
- 1:03:24
- Unidentified speaker: All right, thank you so much.