Let's Talk About Web3
=====================

by kirupa | Interviews with Creative People: https://www.kirupa.com/podcast/index.htm

David Dryjanski (https://www.davidexmachina.com/) gives Web3 a much-needed plain-English treatment. Instead of drowning the topic in slogans, he explains it as a different way of handling trust, ownership, and coordination online, then walks through what that means for wallets, identity, developer tooling, scaling, and the kinds of products that might benefit first.

Watch the interview: https://www.youtube.com/watch?v=txJQRxtJ7lI
A conversation with David Dryjanski | 41m

ABOUT THIS CONVERSATION

This interview is at its best when it stops treating Web3 as a marketing label and starts treating it as a systems question. David frames the web in broad generations: first read-only publishing, then interactive read-write platforms, and now a possible layer where users can also carry ownership and verification with them. That framing does not solve every argument, but it does make the rest of the conversation much easier to follow.

From there, the big idea is trust. Many web products work because a central service keeps the records and decides what is authoritative. Web3 asks when that authority can move closer to protocols instead. For some kinds of identity, assets, and coordination, that is genuinely interesting. Wallets become part login, part signature, and part proof of control, which is powerful but also a lot to ask from ordinary people.

David is also clear about the awkward parts. Key management is hard. Recovery is hard. Convenience often means giving some control back to a platform or custodian. And the developer experience is still uneven compared with mainstream web work. You can feel the gap between what the ideas promise and what everyday users expect from fast, forgiving software.

That is why the most believable parts of the conversation stay focused on nearer-term applications. Storage, portable identity, and some forms of social coordination all seem more plausible than a total rewrite of the internet. The result is not a hype session. It is a useful map of where the ideas are strong, where they are weak, and why the whole space still feels early.

WHAT YOU'LL HEAR ABOUT

- Web3 gets easier to understand when you treat it as a trust model, not a slogan.
- Decentralization matters most when verification is more important than convenience.
- Wallets give users more control while also giving them more responsibility.
- Custodial services speed up onboarding by trading away some sovereignty.
- Storage and identity feel like stronger near-term fits than a full internet reset.

JUMP TO A TOPIC

- 0:00 - Defining Web3: David starts by giving a plain-English definition of Web3 and why the term caught on. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=0s)
- 3:30 - From Web 1 to Web 2: The historical arc from read-only publishing to interactive platforms sets up the next shift. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=210s)
- 7:00 - Why protocol trust matters: They dig into decentralization, verification, and the cases where trusting a protocol changes the model. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=420s)
- 10:30 - Wallets as identity: Wallets enter the picture as a new mix of login, signature, and proof of ownership. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=630s)
- 15:30 - Convenience versus control: Custody, recovery, and user responsibility show why Web3 still has serious usability hurdles. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=930s)
- 20:30 - Smart contracts and the learning path: The interview shifts toward developer concerns, including chains, smart contracts, and how to start learning. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1230s)
- 25:45 - Why it still feels early: Scaling limits and throughput constraints explain why the space still feels unfinished. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1545s)
- 31:00 - Storage and developer experience: Decentralized storage and day-to-day tooling come up as both promising and still rough around the edges. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1860s)
- 36:00 - Where the ideas may land first: The closing minutes focus on the near-term applications that feel most believable today. (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2160s)

TO LEARN MORE

- David's Blog: https://www.davidexmachina.com/
- David's LinkedIn: https://www.linkedin.com/in/davidd8/
- David's X / Twitter: https://www.twitter.com/davidd8

TRANSCRIPT

The automatically generated captions have been organized by speaker, lightly edited for clarity, and broken into paragraphs for readability. Names and wording may still contain errors.

0:00 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=0s)

So, David, thank you for joining me on this chat about Web 3.0, but before we dive in, tell me a little bit about yourself.

0:08 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=8s)

Sure. Thanks for having me, Kirupa. My name is David Dryjanski. I'm an engineering manager at Protocol Labs, which is a Web 3 company, I would say. Before that, I've worked in the software industry for over 10 years, everywhere from Google to Lyft to doing a startup as well. So this is my foray into Web 3. I recently joined Protocol Labs, and I'm excited to talk about Web 3 with you.

0:42 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=42s)

All right. I'm going to start at the very top. You mentioned Web 3. What is Web 3? Is it Web 3.0 or Web 3?

0:49 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=49s)

Either's okay. So, to be clear, this is new terminology that just became popular because I think it has a knack. If you remember, Web 2.0 was a big deal maybe 10 years ago. Web 3 is the evolution of that, and I think it has a nice ring to it, but 3.0's okay too. I say Web 3, but when you talk about Web 3, you really want to talk about the evolution of the internet and what it means for user behavior.

If you go back to the beginning of the internet, or the World Wide Web, which I would say is Web 1.0, the big thing there was that it was pretty much a static, read-only set of information that you could find. That was a huge innovation. Netscape really became a big thing, but the end user was reading websites and couldn't really do much more than that. Then came the evolution of Web 2.0. Again, to refine it's not like there's one point in time when someone said, 'Hey, now we're in Web 2.0.' It's just when a set of technologies advances to the point where there's a new set of behaviors becoming common.

That's when I think these terms make a difference. So it's not uniformly distributed. It's kind of like a trend. Web 2.0 really brought about what I would call the interactive web, where not only is the content being posted by the host of the website, but users could actually generate content and add to the website. Wikipedia's a great example of Web 2.0, where people could edit and add information. This brought about the social media landscape, where you could post articles, interact with your social graph, and chat with people. Anywhere you see this level of interactivity, where the user can now contribute, that I would deem as Web 2.0.

I think the majority of websites today are still in Web 2.0, which is great, and that's where a lot of really cool app development happens. So then that brings us to Web 3.0. And what is Web 3? In my mind, there are a few different concepts here, but it's the idea of building trust and verification into the data that you're posting. What does that mean? If you look at the content on Web 2, that is still owned and operated by the authority of that website. For example, you can add all the posts you want to Facebook, but if Facebook decides to take it down, they can do that, right?

Web 3 gives individuals and users not just interactivity, but also ownership of the data long-term. There are three main things that I wrote about in a blog post recently, just to make sure I don't forget them. One is decentralization, which is the idea of decentralized players pulling control over and management of that data more to the user. So, for example, the data that you post on any site, you can still manage, operate, and maintain. Then there's this notion of building on top of self-certifying protocols. I can say, 'Hey, I posted this data.'

And you or anyone else in the world could verify that, 'Oh yes, David posted this data.' It's not based on the authority of a middleman or host of the content. It's self-identifying, which is powerful because then you don't need centralized management of a lot of this data that you're reading, and you can verify its authenticity without any interaction with a third party. You can just do it yourself, which opens up possibilities. Then the last piece, which a lot of people attribute to Web 3, is the blockchain.

Blockchain is a way of using some of these technologies to create that level of verification and consensus of future states of what has happened, right? One of the reasons there are a lot of centralized systems is because it's easier, right? You trust Google to give you the right search results, and Google has built up this trust over time because they've provided very valuable and relevant search results. However, blockchain is trying to do that for certain types of data.

You have trust in these protocols, not the end service, and the protocol can guarantee that trust by having consensus mechanisms that allow different parties that don't know anything else about anything to just adhere to the protocol. As long as you trust that protocol, the data being generated there can be verified and validated for whatever purpose it may be. I think that was a little wordy. I can dive into those details.

7:07 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=427s)

No, I think it's helpful. If I had to summarize it using a very short 140-character limit, it would be: Web 1.0 is read-only. Web 2.0 is read and write. And Web 3.0 is read and write with decentralization, where the data you're storing isn't based on the goodwill of where it's being stored to be accurate or represented correctly, but is inherently based on technologies that have no motivation to do one thing or another except verify and validate that the data is true.

7:39 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=459s)

Yeah, that's a good summary of it. I agree on the trusting through these verifiable technological means, these cryptographic primitives that give you that power.

7:54 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=474s)

Yep. The difficulty I always had is this: with Web 1.0, we had examples of how my everyday life would have changed from the websites that I used. And with Web 2.0, you saw the gradual transition as sites added commenting, and then you had Digg, Delicious, and some of these sites that started doing bookmarking. The small things that are not just based on my own activities, but the activities of others, slowly influenced them into the bigger things we see today in terms of all the social networks.

For Web 3.0, the part I also have difficulty with is how my day-to-day activities on some of these sites that I use regularly will change. Or will there be a net-new kind of world created that's going to be different than what I've done before, and it's going to be eye-opening and interesting in that respect?

8:40 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=520s)

Yeah, no, that's a great question. What are the small features that you could start to see that would signify that, 'Hey, maybe this is a Web 3 site,' and that then becomes commonplace? I think the best analogy here, or the best feature that's already live in a lot of these, is the notion of a wallet, or a cryptographic wallet, that a lot of these blockchains use. The way you can see it on a lot of these sites is that it's replacing login. For example, if you go to one of these NFT marketplaces. Where you can buy these non-fungible tokens—we can talk about that too.

That's a very popular thing right now. That's part of Web 3, but not only Web 3. If you want to interact on that website and purchase an NFT or make a transaction of any kind, you don't sign up and put in a username and password anymore. You use what is called a wallet, and that wallet resides on your local machine. So you have these third-party extensions; MetaMask is a very popular wallet extension. You also have wallets for different cryptocurrencies, or whatever that NFT marketplace accepts, based on the different blockchains. That's how you actually log in, and that's your identity on these websites for Web 3.

I think that's probably going to become a bigger feature that's going to extend to all these Web 3 concepts, because if you imagine Google auth for logging into all these third-party websites and SSO, that might all go away. As long as you just have your local wallet and that's the way you sign in, you don't need any other credentials, so to speak.

10:51 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=651s)

Yeah, it's almost like the secondary verification you get with an app that gives you a code. In this case, it's a much larger hash of letters and numbers that is uniquely tied to you. But when I say uniquely tied to you, it also provides some anonymity as well, correct? Because no one will ever have to know who the person is behind it. All they'll have to identify you is this very unique code, in some ways, that is pretty much tied to your machine.

11:20 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=680s)

Totally. I think that's also why Web 3 is exciting. There have been a lot of cryptographic advancements, and with cryptography you can get certain things, like anonymity, to a certain degree. I think identity will be an interesting one, where on some sites you will want to be more anonymous, and on other sites you'll probably see these websites attach a handle, like a username, to these wallets or have some semblance of identity. Imagine having a social network built on top of blockchain. You're probably not going to refer to people by their hashes, which is their wallet identifier. You're probably still going to want some human-readable name associated with it.

So I think the interesting thing about Web 3 here is that there will be a spectrum for different blockchains and what level of identity you want to portray and give, but you're not tied to that for every platform. You can change it, which is interesting.

12:31 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=751s)

Yeah, and the thing that's interesting as well is that when you have a central authority, one of the advantages I often see is that, let's say I lose access to my username or password, I have some way of getting it back. Is there a similar mechanism that exists in this world where there isn't any central authority that can help with this?

12:49 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=769s)

Yeah, so this is one of the challenges of Web 3. When you give more power to the user, there's that famous saying in Spider-Man, I believe: with more power comes more responsibility. Or great responsibility, I think, is the official quote. So I think, by default, in a decentralized system, the user is responsible for their keys. If you lose that private key, that secret that you're supposed to keep in the safest space, you lose access pretty much forever and there's nothing that you can do. That's the basic default. When Bitcoin came out, that was the big thing, that default feature.

The reason I'm not too worried about this is because I think, over time—we're in the early stages of Web 3, what it means, and how sophisticated the software will be—it's going to be a spectrum, much like anonymity is a spectrum. On Web 3, you can also see that there will be services that provide different levels of control and access to your wallets. For example, Coinbase, a very public company in the crypto space, is really in between Web 2 and Web 3. They're onboarding Web 2 people who want to be involved in crypto or buy crypto assets with a normal login, and they'll store your wallet for you.

What that really means is they have complete access and control to that wallet, much like a bank, right? They control your asset. They call that a hot wallet in Web 3. The reason it's hot is because it's always accessible by some party. So if they get hacked, which would be a worst-case scenario, you're trusting Coinbase with your wallet. So that's one end of the spectrum. Then you have Bitcoin on the other end, which is you have your wallet, you protect your secret key, you write it down on a piece of paper and save it in a secure location, and it's on you if you lose it.

But there are also other interesting technologies, like multi-signature wallets, which are pretty cool. That's the idea that you can have multiple parties who have to sign off on any transaction from that wallet. What that gives you is some kind of redundancy or backup. So you could essentially have—let's say the Kirupa wallet. If you trusted me, you could give me one of the secret keys. You could give one to your family, one to yourself, and maybe one to a third party that you trust. Then, as long as two people sign off on the transaction, that could go through.

But if you lose your key, you can go to your trusted friends and family, and maybe even regenerate another one for you and have that accepted. So there are some backups and different parts on the spectrum where you can get away from this. It's not only on the user, and it's not just, you know, lose it and you're out of luck.

16:06 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=966s)

Got it. So essentially, you have choice. Today you don't have much choice in terms of whether you want to have your credentials associated with a central authority like Coinbase or whether you want to go completely decentralized. You pretty much choose the one you can. Whereas in Web3.0, depending on your risk and comfort level, you can choose exactly what you want and still be able to participate in the larger ecosystem of what Web3.0 supports, which today is very tricky.

16:31 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=991s)

Exactly. That's a great way to think about it. Another way I like to think about it is, back to the social media example, you have all your friends on Facebook. If one day Facebook decides to just shut down, you lose all that data, right? You have essentially very little control over what Facebook does on a day-to-day basis, where, depending on your level of preference and involvement and interest, you can really control, to the degree that you want, the data, the sites that you use, and how you use that data and how they use that data as well, which is pretty exciting.

17:15 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1035s)

Yeah. So getting back to more of the developer angle of it, when I think of Web1.0, I think of some FTP server somewhere. You're manually uploading static files to it, and that's pretty much your entire world. Then we come to the Web2.0 world where you need more dynamic read and write. You can imagine the traditional LAMP stack—Linux, Apache, MySQL, and PHP, or whatever—and you're getting content into it, and of course all the stuff that we have right here. So how will I, as a developer, participate in this Web3.0 world? Because that to me also is an area where I have no clue. What do I start with?

What tools do I use? Is VS Code the one, and is that what I use for this? What do I do here?

17:57 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1077s)

Yeah, no, this is one of the exciting things about Web 3 and also one of the major hurdles of Web 3, is that there's so much going on that it's hard to even know where to start. I feel the same pain that you feel, and that new people coming into the space feel, which is, where do I even go? Part of the reason is that every blockchain is different, right? There's a whole other set of developer tools there, a whole new set of primitives and APIs, and what does this blockchain do even?

But I would say, if I could boil it down, it's a lot more server-side development at the moment. I think this, again, is because of the maturity of the space. A lot of the functionality of blockchain and Web 3 technology happens on the server-to-server side. Building that consensus mechanism, doing transactions, that's all server-side. Part of it is because on-chain transactions are very expensive at the moment, because the scale of the systems is still relatively low compared to Web 2. So server-side development is probably the starting point for most people if you want to get into this space now. Now, that doesn't mean there's no front-end development opportunities.

In fact, I think that's one of the things that's really missing in Web 3. We need better UIs. MetaMask is one great example of putting a wallet on a Chrome extension. That's Web 2 technology, right? And they just call APIs for these back-end servers. I think there'll be more of that. But a lot of the blockchain technology is really on that side. All of those server-side technologies are pretty standard, whether it's Rust or Go. Golang are the two really popular ones, even Python for some. The area where it's a little bit in between server and front end is what I would call the smart contracts area.

There are new languages around smart contracts that Ethereum made really popular, like Solidity. So if you are interested in learning how to write a smart contract, Solidity is probably the go-to. There's a few newer ones that are a little bit easier to use. Again, different blockchains have different languages for their smart contracts. But for someone who wants to get started learning, I would start with the bigger blockchains that just have more users and a bigger community. There's going to be more resources there. I was going to add one more thing, but I forgot about it. It'll come back to me later.

20:59 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1259s)

No, you know, for me the way I learn is by drawing parallels to stuff I already know. So it almost seems like, at the end of the day, all this stuff we're building with 3.0 is still built on top of TCP/IP and more fundamental concepts that have been around for many decades. But just like in the late '80s and early '90s, the internet was very expensive to use because there was a lot of capacity and skill needed. I think there's a similar world right now. I guess the way to think of it is not like we're extending what we have right now to do these sorts of things.

We're still expecting similar kinds of capabilities and applications and so on, but the under-the-hood implementations are completely different. So we're still at the level where we haven't built the layers of abstraction yet, where someone can just bring up some random tool and start poking around on whatever technology or thing they want to build. There's a lot of work still going on in building, starting from the nuts and bolts and adding the fundamental concepts that will eventually enable some of these layers of abstraction. It's almost like back when we had HTTP, Gopher, and FTP. You had all these competing protocols, and eventually one of them won out.

This might be a case where, because of the decentralized nature of the whole thing, there may not be one that wins out. They'll all have their own use cases and support. And it's up to you to pick which one you want to build on, and each of them has its own level of technology sophistication in use, just like it was back in the day when you had various stacks of technology that you had to learn to be successful.

22:24 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1344s)

Yeah, no, I totally agree. One of the other things that you mentioned that reminded me of another thing is TCP/IP or the HTTP protocols. I think there's also a lot of peer-to-peer communication stacks that are going to evolve on the back end. So obviously the web will still use HTTP and TCP/IP, but a lot of these back-end stacks are using peer-to-peer libraries where, again, the view of a blockchain is that there's no server-client relationship on the back end, right? Everyone's a node and everyone participates in the blockchain as a peer. So a lot of that intra-blockchain communication is through peer-to-peer networks more so than your standard HTTP stack.

23:13 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1393s)

Yeah. So you mentioned that everyone is a node. One of the things I read is that pretty much a node contains a copy of pretty much everything on that blockchain. Is that correct?

23:25 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1405s)

Yeah. So this is a good question. Again, it depends on the blockchain, but the idea is that if you are a quote-unquote full node, you do have a copy of the entire chain. Now, for many blockchains there are versions of nodes that are called light nodes that only have maybe, let's call it, the last week of transactions of the chain so that it's manageable. But this is also an interesting area that I think has a lot of disagreement, or discussion, as to what is the best way to set up a blockchain in terms of security and in terms of resiliency of a network.

There's this balance of how big, or what the minimum requirement of a node is. So, for example, if you wanted to be a Bitcoin node, like a miner—right, I guess I don't have to do air quotes; they're called miners—but if you want to be a Bitcoin miner node, which would be a full node on the Bitcoin blockchain, you essentially need a lot of compute, right? So much so that your average laptop would not be able to keep up in terms of the hash rate and even really participate, or it would be needless computation. You really have to have a very professional setup.

So some people argue that those requirements are so high that they don't allow many different participants to participate in the blockchain. Whereas others argue, well, if it's too little—if you could have a node on your smartphone and everyone could be on that—then that's easy to kind of game the system theoretically. If anyone could do it, then maybe you can create more fraud and create these consensus tricks, or attacks, on the consensus protocol itself. So there is a balance, and if you get into the weeds, you can read about a lot of these different debates that a lot of...

25:36 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1536s)

Yeah, 'cause the point I always—I guess I just can't wrap my head around it—is that my laptop, or my machines, have at most 256 gigabytes of storage. That's pretty much it. And my upload isn't as fast as my download. So I can only imagine if we replace our central servers, our cloud computing hardware, and all these things with local machines spread across large areas of the world, one thing I question is what the lag time is between me being able to do something and it actually being verified as truly accurate. Then there's how much bandwidth and storage it's really going to need.

Because if blockchain does become the next evolution of the web, and we know that approximately there are tens of millions of terabytes of data generated across all these things—I'm sure YouTube, for example, is generating a lot of that in probably a few seconds—how does that scale? How does it work when people have different kinds of connections and verification is a big part of it? All of that, I can't wrap my head around.

26:34 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1594s)

Yeah, no, it's definitely a complex issue, and again it's based on the different blockchain that you're talking about. But I think there's a few things I would point out just to make sure people are aware of. You're totally right that currently the way Bitcoin operates—I'm using Bitcoin as an example because it's the most well-known, and it also suffers from the largest scale problem—is that people say, 'Hey, Bitcoin's going to replace currencies.' But actually it can only process so many transactions a second. It would have to 10x before it could really start to consider itself a currency in terms of volume.

So scaling is a major issue for a lot of blockchains, number one. So that is true. And number two, not every transaction, or I should say the data of every transaction, is on the blockchain. So, for example, a YouTube video would probably never be on the blockchain because it's just a gig that everyone now has to store on their hard drive. No, that doesn't make any sense. But you would have a link to that YouTube video, potentially, on the blockchain. I think we're at the point where it's like, 'Hey, none of this scales.' How is it ever going to scale to the masses?

I think that's going to require a lot of investment, and it's already starting to happen in a lot of these networks, like Ethereum, for example, moving to proof of stake and a hierarchical consensus mechanism to reduce the load on the edge nodes and have more consensus layers that can do so. You could think of that hierarchy as aggregations of transactions at smaller levels. So, for example, maybe everyone in your city has one node—not one centralized one, but one bigger node for your city—that does all the transactions.

That node then compresses all that information and does something more advanced, like, for example, zero-knowledge-proof rollups, where even from that compressed information, they can compress it probably another 10 to 100x and send that to the global node network that can say, 'Okay, these are all the transactions verified by this aggregator node.' And that's all the information we put up there. It's verifiable, but it's verifiable over this aggregation rather than every single bit of data on the...

29:34 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1774s)

Yeah, it's almost like every sector on a hard disk is now distributed everywhere, and there's an index file that many hard drives have that kind of shows what sector has what part of a file in it. It's also distributed. So it's going to be a lot of just keeping track of things and then collecting them back together when you need it. But I guess it also goes back to how frequently you need access to all that information as well.

It's kind of like the multiple storage tiers you might have at cloud providers, where they have more real-time and then they have the glacial, slow archival one, which takes several days to get back to you on some of these things. I can see that being something that happens for real. So getting back to the YouTube example, that's an interesting one because you mentioned that there'll never be a time when probably a multi-gigabyte YouTube video would be on the blockchain, and that sounds very reasonable to me. Which means that there'll still be a role for a central authority in some cases.

Whoever actually stores the video itself might need to be using more traditional computing, like a cloud-based storage system or a centralized file server that is managed by a handful of companies. So do you see that true success in Web 3.0 will probably be a hybrid, where there'll be a place for private, centralized authorities to coexist with some of the world that we see with the decentralized players?

Or will it really be a case where we eventually get to the point where everything I can do today with all the various infrastructure we have for serving content and just doing things on the web will be replicated in the same fashion, with the same performance and scale and price, in Web 3.0?

31:02 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1862s)

It's a great question. I can tell you what I want to happen. Who knows if it will. It would be more of the latter, where I do think there will be decentralized versions of many of these systems. Maybe some of them are just implausible, and I don't know what those would be necessarily. We can talk about the video use case in a second. But the other piece of it goes back to the spectrum, right? One of the beauties of this decentralized notion is, let's say there was a decentralized streaming system. That doesn't stop YouTube from participating in that decentralized streaming system.

I think it's about building that trust in a decentralized fashion so that if they participate in these decentralized networks, they will still be able to serve users. The difference is the user now has more control over which stream to go to, why they go to that stream, and how they make sure they have access to that content. Will this play out in the next 2 years? No. I think it's going to take a long time to see where this actually plays out.

I think the important thing to know is that the things that will happen on-chain are going to be a lot more nuanced and more about verification rather than real-time application. Real-time will still happen, and you'll just be more like checkpoints of, 'Okay, all this happened. Both parties were happy with the transaction or the streaming capabilities, and this user will now pay for that content.' Then it's less of a real-time notion of a blockchain. So you have to be careful about what happens on-chain and what happens off-chain, if that makes sense.

33:08 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=1988s)

Yeah. No, I think the spectrum analogy is the right way to think about it, where if you really need it, you can pull it off. If you don't need it, you have other conveniences that you can use. So one last question I have has to do with cost in general. I played with a little bit of Solana development just to get a feel for what exactly this does, and there are obviously cases where you can develop on the blockchain live, or you can do it more locally, where you can test these things. Even the most trivial action that I did had a small cost associated with it.

Today I can do builds on my local machine and it's free. If I deploy it's still free, and there's always a free tier that all these providers give me where I don't have to pay for anything. It almost seems like with the blockchain there is no concept of that. At least what I found is that everything you do has a small cost associated with it, and then there's a larger cost for larger activities as well. Is that correct?

33:58 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2038s)

I think that is the case with production blockchains. Many of these blockchains use the concept of gas, which was really spearheaded by Ethereum. One way to think about Ethereum is as one giant computer where everyone runs these instructions. If you tell my node, “Run this, look through your list of 10000 photos, and find this one photo,” what happens if your instruction contains an infinite loop? Imagine every node running those computations all the time. That's expensive. The notion of gas was introduced to say, “Wait a second. This computation is really expensive.” The node can also stop if the expense crosses a certain threshold.

That's how they can calibrate the network a little, but the larger idea is general compute. These networks have associated costs. Solana should ideally have lower costs because it's intended to be a higher-throughput network than Ethereum. Every participating node may be asked to perform the same work, which is why a poorly written instruction can become so costly. For developers, there are still test nets and developer nets. Those have a cost, but they use fake currency, so to speak. You might have a test version of the Solana token or Ethereum token, with a fake account that's just a test account like you'd use in any testing environment for a Web 2 product.

The challenge is that you're still running this test net globally as a distributed system, so the development experience is more difficult. That goes back to the tools not yet being fully sophisticated. But you're right: let's say you've tested everything and you're ready to go. Your smart contract works, and your app works. Now you want to launch it to people. How do you do that? Cost may become a barrier to entry, and the question is how that cost comes down over time. That's interesting to think about. Filecoin, which is what Protocol Labs works on, has a program where developers receive X amount of storage free because Filecoin provides decentralized storage.

I think we'll see more grants like that. Think of AWS credits. When you sign up for Amazon Web Service, you can receive credits to try running your website on Amazon. I don't know how many they provide, but you might receive $10000 in Amazon Web credits. I think similar programs will exist in the decentralized world, with different blockchains offering developers that kind of support.

37:07 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2227s)

Nice. Yeah, and I think it's an exciting world for developers. The bar is high, but then again, the bar is always high for new and emerging technologies. I think those who learn some of these things will have a leg up when it comes time for broader adoption by everyone.

37:25 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2245s)

Yeah. And again, I think it is early, and if you are interested in learning, I think you can learn. It's a matter of focusing on what parts you want to hone in on, because there are so many different technologies and blockchains. You can actually pick and choose as a developer and say, 'Oh, I really care about what this blockchain is trying to do.' Obviously, you need the time and space to think about that and invest your efforts wisely. But once you do it, you can actually do it. The other really cool thing that is very different from Web 2 is the default of being open.

Every blockchain we've mentioned here is open source. You can read the code and understand it if you want, and really know what it's up to. If you don't like it, you can fork it. You can just say, 'I'm going to make my own version of Solana, right?' That's a really powerful tool that developers have to really shape these technologies in the future. It's all out there. It's all in the open. And you can take the best of all these different blockchains and build your own if you really wanted to.

38:36 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2316s)

Perfect. Last question. What's one killer app that either exists or you wish you could see existing that can only be done on Web 3.0? That can only be done on Web 3.0? Correct.

38:49 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2329s)

Well, one of the reasons I'm excited about Filecoin is that I think decentralized storage will be a big app. What you think about there is, 'Oh, that sounds kind of basic. I have a hard drive on my laptop. Why do I need it? Google works just fine.' But for me, the exciting part is the ownership of your data. I think Filecoin will be at the precipice of giving users control of their data. So imagine Google Photos, right? I think they recently changed their terms so that now you have to pay for any data after some amount of gigs, whatever that threshold is.

Imagine if you could store all of that decentralized for a lot cheaper, because you get these scaling costs that are much more attractive in the decentralized space. So imagine instead of backing up your photos to Google, you could back them up to Filecoin. I think that would be a killer app, lowering that cost for people to just store more and more of their data. The other one, which is less related to the work I do currently, is social media. I really think a lot of the social data should be a blockchain-level network. I think that would give more power to the user, like, 'Hey, this is my social network.

I choose who I want to talk to. I choose the information and the algorithm for what content to read.' That doesn't have to be decided by Facebook. I think that could be really powerful in the future.

40:35 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2435s)

Cool. Well, David...

40:38 - David Dryjanski (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2438s)

No problem. Thanks for having me.

40:38 - Kirupa (https://www.youtube.com/watch?v=txJQRxtJ7lI&t=2438s)

All right, everybody. I hope you enjoyed my chat with David about Web 3.0 and its relevance to a lot of us as developers. Before we call it a day, do take a moment to check out the links below in the description for more information about David and Web 3.0 and other things that might help you further on your journey. And with that, I'll see you all next time.


Browse all Interviews with Creative People: https://www.kirupa.com/podcast/index.htm