About the AW Meetup
This informal gathering for creators and builders in autonomous worlds and blockchain gaming is organized by AW Research.
We warmly welcome all interested in blockchain gaming and autonomous worlds to attend - creators, investors, enthusiasts.
This is a chance to gain first-hand insights into fully onchain gaming, projects expanding the AW narrative, and the builders behind them.
We have invited these incredible builders from the AW/FOCG to join our panel session on exploring Autonomous Worlds.
This interview will be divided into two segments. Segment one will consist of 3 standard questions, while segment two will feature tailored questions prepared by the AW Research community for each guest.
Liron Hayman, Head of Business Development at StarkWare
I would like to express my gratitude to everyone for joining us today. Although we are serving as sponsors, it's truly remarkable to witness the growth of the StarkNet ecosystem. Just a year ago, this level of activity and community engagement was not a reality. We used to be intimately acquainted with every facet of the community's progress, and now it's a thriving ecosystem that constantly captures the attention of the broader web community.
Let me take a moment to share my thoughts on Autonomous Worlds and why we believe StarkNet technology plays a pivotal role in unlocking their potential. To set the stage, let's rewind to 2013 and Vitalik's original blog posts. Vitalik was inspired by a personal experience in an online game (World of Warcraft) where he woke up one morning to find all his in-game assets gone. He realized that if these assets were on a blockchain, such a mishap would have been impossible. People often get engrossed in the price fluctuations of cryptocurrencies, but what we're fundamentally advocating for is custody autonomy.
Vitalik had this vision a decade before Ethereum became a reality. In recent years, some have referred to this phase as "Web 2.5," characterized by on-chain minting of in-game assets while everything else remained off-chain, reliant on game developers to maintain fair gameplay. The natural progression from this point is to bring the entire game onto the blockchain, ensuring transparency and integrity in gameplay.
So, why was this not feasible with existing blockchains, and why is it achievable with technologies like Dojo on StarkNet? The issue with traditional blockchains is that as they gain more users and activity, everyone experiences higher fees, making it unsustainable for large-scale applications. ZK Rollup technology addresses this problem by making every additional transaction more cost-effective for everyone. The more transactions, the better the cost per transaction becomes, creating an ecosystem where millions of daily active users can thrive without crippling fees. This is where StarkNet comes into play, helping to unlock the full potential of these Autonomous Worlds.
One fascinating aspect is that many development teams are discovering StarkNet through the realm of gaming. They initially approach it with the idea of building a game, perhaps utilizing Dojo, without necessarily realizing the powerful blockchain infrastructure underpinning it. It's a beautiful evolution to witness, and I had the pleasure of meeting such teams at a recent conference. StarkNet's future looks promising, and we're excited to see where this journey takes us.
If there's any way I or StarkWare can assist you in this journey, please don't hesitate to reach out. You can find me on Telegram at @satoshi_moon, or simply send me a direct message. We can help you connect with the community, introduce you to the Dojo team, assist with hiring developers, and provide infrastructure support to bring your app or game to life. We're here to support and contribute wherever we can.
Once again, thank you, everybody, for being a part of this exciting journey.
Panel Guest Introduction
Damian: Please introduce yourselves to our audience and share what you've been working on.
Changlu: Certainly, my name is Changlu, and I serve as a core contributor for AGLD DAO. Currently, our focus is on the development of the Loot Chain, but we're also committed to sharing knowledge from the world of Autonomous Worlds and actively working towards increasing adoption.
David Amor: Hey, I'm David Amor, based in the UK, and we have a team here. We're essentially video game veterans who have been working on on-chain games and Autonomous Worlds for the past two years.
Sebastien Guillemot: My name is Sebastian, and I'm one of the co-founders of Paima Studios. We specialize in the world's largest roller framework for on-chain games and Autonomous Worlds. Our support extends to every EVM-based chain, and we also provide support for certain non-EVM-based chains through a novel account abstraction framework.
Yijia: Hey everyone, my name is Yijia. I'm a co-founder of Curio. We're based in San Francisco and Shanghai, and we've been working on on-chain games for almost two years now. We're excited to announce that we'll be launching our first commercial game, a social diplomacy game, at the beginning of next year.
Hello, everyone! I'm Shramee, and I am a Dojo developer. Can I use that as a title? I'm a software engineer at Cartridge, and Dojo is an ECS framework designed for developing Autonomous Worlds and fully on-chain games.
Norswap: Greetings, everyone! I'm Norswap. I'm a founder at 0xfable, which is a fully on-chain trading card game, a little bit like Magic: The Gathering or Hearthstone, but where all the logic lives on the blockchain. The idea there is to make the game extensible, so that people can experiment with variation to the rules and new game modes.
Panel Questions One
Damian: In your own words, how would you define an Autonomous World? I understand that a fully on-chain Texas Hold'em Poker game may not encompass the full concept of an Autonomous World. So, please provide your own description of what an Autonomous World represents.
Changlu: I believe the most significant aspect of an Autonomous World is that there are no privileged parties with the authority to pull the plug and halt its support. Individuals can choose not to participate, but what's crucial is that the rules within an Autonomous World can be governed by those who have an interest in it. Whether they are signatories or they submit some form of governance proposal, they theoretically have the ability to modify the rules of that Autonomous World, essentially leading to self-governance.
David Amor: I'm not entirely sure if I possess a definitive answer. To begin with, the concept of an Autonomous World isn't precisely well-defined, so everyone might have their own interpretation of what such a world entails. It's somewhat akin to the idea of immortality, where we envision a world that is perpetual and its usability remains constant. In this world, the rules don't undergo frequent changes, and there isn't an external party that can arbitrarily alter them.
Sebastien Guillemot: I completely agree with everything that has been mentioned. Additionally, I'd like to emphasize that one key aspect of creating an Autonomous World is that users should log in for the world itself, rather than solely for the game. If we take "World of Warcraft" as an example, many players log in for the social experience and to explore the game world, not necessarily just for the game mechanics. Therefore, I believe this is a fundamental element of an Autonomous World. We want people to be able to interact with these worlds purely for the novelty and the experience of engaging with the world itself.
Yijia: I don't have what I would consider to be a definitive definition. "Autonomous Worlds" is a vast and evolving concept, and it's up to each of us to contribute to its definition and development. As of now, I don't believe there have been any truly successful Autonomous Worlds. We are all in the process of building the foundations for such worlds.
Shramee Srivastav: Yes, that's a great point, and I wholeheartedly agree with everything that has been discussed so far. I'd like to add a small detail, which is that traditional games can be likened to predefined experiences, while Autonomous Worlds provide more of a blank canvas where you can begin building your own creations. In traditional games, the experience is often fixed, but in Autonomous Worlds, it starts with a more basic framework, and players can collaboratively add more content and features to it. Essentially, players become co-creators within these worlds.
Norswap: I think the key is permissionless extension, as others have said. And maybe I can give you an example to make things more concrete. So imagine you have a Pokemon world. You can move in the world, capture Pokemons, etc. But now imagine that players can add new continents and maybe new Pokemons, maybe change existing Pokemons, ... That would be an autonomous world that can keep evolving even when the creators are not involved. And then there are some rules that some people call digital physics and are fundamental to the world. Maybe it's like: what level the Pokemons are in certain areas, or their abilities, and things like that. And those rules tend to be retained across all the extensions and form the core of the game. But I guess people can also choose to opt away from them if they really want, but maybe the extension won't compose as well.
David Amor- Additional question
I'm curious about the perspectives of others who are involved in building Autonomous Worlds. How significant do you consider extendability to be? Do you believe that an Autonomous World can't truly be considered autonomous if it isn't extended? Is this attribute important in your view, or not?
Yijia - Additional answer
I think it depends on what layer players are extending.Is it the fundamental physics or is it social relations on top. I think it's pretty critical.I think Autonomous World is ultimately about getting power back to players in some significant way that allows you to shape the virtual of the engagement right and that is unlikely to be the digital physics layer because that would be very disruptive.But anything above that I think give it if you can give back to players in a way that enhance their experience.I think that's the beauty of autonomous world.
Sebastien Guillemot - Additional answer
Additionally, I would like to emphasize that while I believe this is incredibly important and a fundamental characteristic, another perspective to consider is that value should be able to be created by the users themselves. In other words, it's about the ability for users to generate content that enhances the value of the world from within the world itself. So, in a sense, it's a different way of looking at sustainability, where the users play an active role in driving the world's value. I consider this to be a key property as well.
Panel Questions Two
Damian: Could you please share a recent frustrating experience you had while playing a fully on-chain game?
Wow, that's a spicy question. I have one, it's going to be easy because the game actually improved on that aspect. I played this game called "PirateNation," where you take on the role of a pirate battling monsters.
I had two frustrations with it. First, the combat system was a little bit janky. There were various actions, but it was unclear what each of them did. For instance, there was a "fast shot" and a "strong shot" with no clear indications of their effects. However, now they've come up with a new combat system which is very clear and actually fun. So I'd recommend everyone to give that game a try.
The second frustration was that the game revolves around farming items to craft more powerful items, and it wasn't obvious to me which items to farm to get access to the more powerful ones, and how to do that. However, the game is evolving rapidly, and many games tend to start out a bit rough but improve over time, so I'm very bullish on the developers improving that aspect of the game.
Yijia: In our case, the experience was quite challenging. We developed a game called "Treaty" and launched it in January this year. It's a 4X strategy game where players essentially compete in a battle royale-style contest for control of the central tile. We had around 20 to 30 active players who were dedicating several hours to the game every day, and they even created a lot of memes in our Discord community.
However, the problem arose when the game started experiencing significant delays, reaching up to 15 or 20 seconds per in-game transaction. Keep in mind that this game was running on the Ethereum Virtual Machine (EVM) with a single sequencer and EVM node, which is the fastest setup possible within the EVM framework. Additionally, our indexer kept breaking down, happening around 30 times in just one week.
This was a pivotal moment for us because it made us realize that while on-chain games and Autonomous Worlds are exciting concepts, we must find solutions to ensure a smoother and more reliable user experience. At one point, our game's issues became so severe that the game sessions couldn't continue, and it was a real source of frustration for us.
Sebastien Guillemot: The most significant challenge for me is when I have to bridge assets. I live in Japan, and often the assets I want to acquire are not legally available here. So, I find myself in the situation of having to figure out how to send money to someone else, have them buy the token on my behalf, and then return it to me so I can use it in the game. It's a cumbersome process, and it's the main issue that I believe many people are trying to address and find solutions for. Fortunately, it seems that many are heading in the right direction to tackle this problem.
David Amor: We had planned a playtest for our game, which was scheduled for today or tomorrow. However, during this playtest, we encountered a situation where the blockchain setup was similar to yours, and the transaction times were six seconds. When you're dealing with an on-chain game where every move corresponds to an on-chain transaction, a six-second delay can severely impact the game's experience and flow. This served as a reminder to me about the challenges associated with the infrastructure surrounding on-chain games.
The world of on-chain games is still in its infancy, and even with tools like Dojo, which is a promising engine, it's comparable to the early days of development. When you compare it to more mature game development platforms like Unity and Unreal, there's still a long way to go. Often, you find yourself having to build a significant amount of infrastructure or compensate for the existing gaps in the ecosystem.
Changlu: My most significant frustration with on-chain games is that many of them appear fascinating, but when I attempt to play them, I often find that there are no other players available. This results in me waiting in the lobby, hoping for someone else to join, and sometimes I have to create an entire guild just to get enough participants to try out a single game. In my opinion, more game developers should consider incorporating AI players so that users can jump in and test their games even when there aren't enough real players available. It would be helpful to have NPCs or AI opponents to make the gaming experience more accessible for testing purposes.
David Amor - Additional question: But when I compare this situation with traditional games, even in the case of 2.5 games, we still need to actively involve players. So, do you think the issue lies in the infrastructure or is it more about a shortage of talented game designers?
Changlu - Additional answer: Yes, I mean, obviously it's both. You need the technical aspects and implementation, but it's also about retaining people's attention, which is challenging in both crypto and traditional gaming. The only way to keep their attention is through compelling content or continuous engagement strategies. If the only time you can engage them is every Friday at 6 p.m. Eastern Time, it won't be sufficient to hold their attention.
Panel Questions Three
Damian: If you were not allowed to run your own project, which project do you have a strong belief in and would like to participate in for the next two years? Since you weren't sponsors of our event today, and thus not presenting your own game, please choose a project that you genuinely have faith in and would like to encourage our audience to join. Share your thoughts on that.
I think a roundabout way to answer this question is by mentioning infrastructure. As many people have mentioned, it's rather lacking in some aspects. If I wasn't building a game, I would be building infrastructure, especially at the chain level: custom chains and engine, improving throughput, making rollups easier to deploy, ... I know Yijia and Curio are working on the Keystone engine, there is also Argus Labs working on their own blockchain game engine. The game frameworks like MUD and Dojo are also very cool, though I believe the chain level might present a more fundamental problem for the ecosystem.
I actually try to work on anything that I find really interesting. So, here are a couple of things to keep an eye out for. Recently, StarkWare opened sourced their official prover, It's called Stone prover, So you can get stoned, checking out its source code.
There's also another cool project by Andrew Milson. Which is also a Stark prover, and it's hardware accelerated and also something I want to explore and work with. I would encourage you all to look into these projects as well.
Yijia: To be honest, I'm not entirely sure. I find the concept of the "share sequence" itself quite intriguing, so I'll definitely look into it more. However, I don't have much knowledge about it at the moment.
Sebastien Guillemot: Paima is actually composed of three different projects. The first one is the Paima Engine, which is an open-source framework for building the core infrastructure of autonomous worlds. Additionally, we have two partner projects that address specific aspects. ZekoLabs is our ZK layer for Paima, essentially our custom ZK rollup framework. The other project is Shinkai Network, which serves as our peer-to-peer AI layer for autonomous worlds, providing on-chain AI agents for games. If I were not working on Paima, my main priorities would be joining my other co-founders in working on Shinkai Network because it's an incredibly interesting project.
David Amor: I think you managed to share your own thing haha.There's an on-chain project that I find quite fascinating. It's called "Word3," and they are doing something entirely different from what everyone else is trying to achieve with big games. They focus on creating really polished and enjoyable casual games. I'm impressed with their approach and the quality of their games. That's one project that caught my attention.
Changlu: If I had to join an on-chain project, it would definitely be GGquest. The founder and their team are doing really awesome work in the space, so a shout out to them. If I were considering an off-chain project, I think what Nvidia is doing with autonomous agents that look like real people is fascinating. They recently did a tech demo of it, and it's just mind-blowing. The amount of hardware they have and their commitment to pushing AI boundaries is likely to revolutionize gaming in general. So, I would probably join a project like that.
Damian: There's a concerned community member who has raised questions about the motivation behind the loot chain. They feel that AGLD has potentially betrayed the Loot community by choosing to allocate resources to building another chain instead of enhancing the utility of AGLD with Loot in existing games. Given the abundance of chains already available, what are your thoughts on this matter?
Changlu: I believe that having our own chain makes it easier to leverage partnerships and media opportunities, consolidating all our relationships and value-add into one place. This is something we offer to all the games that collaborate with us.
Moreover, it's a step in the ongoing development of our ecosystem. We've recently integrated with Layer 0, which will enable us to add interoperability to both AGLD and loot NFTs, allowing them to become ONFTs and OFTs, and facilitating interactions with other platforms and assets.
Damian: You are a true OG in this space. What are your thoughts on entrepreneurs constantly reinventing the wheel? Do you believe there are too many individuals working on infrastructure projects and not enough focus on creating high-quality games?
Norswap: Yes. Both are important, but I believe if you are not building a game yourself, it's difficult to know exactly what game builders want. After I left my previous job, I wasn't sure what to do and I explored the infrastructure side. There was so much competition there, so many projects doing more or less the same thing. Too much infrastructure and not enough apps as well. So I figured I'd do an app, then do infrastructure on the side, tailored to my needs, and hopefully that can be useful to other people too. Also, people who run infrastructure businesses tend to try to capture market share and create moats, and that's something that we don't want as builders. Each time I talk to infrastructure ompanies, the first thing I ask them is how I can migrate away from them if things don't go well I ideally want all the infrastructure companies to act as service providers that adhere to standardized interface and actively work to create new standards. RPC providers would be a perfect example of that model, it's the same interface for all of them and you can just switch providers based on factors like service quality and cost. I wish infrastructure providers open-sourced all their components. They're my complement as a game builder, and so I want them to be commoditized.
David Amor- Additional answer: I want to point out that it seems a bit risky for infrastructure to get ahead of the games and try to predict what the games will need. It feels like at the moment, infrastructure is ahead of the games and is trying to dictate what the game developers should do. Personally, I'd like to see more effort put into creating game content rather than just infrastructure, but finding a balance is important.
Shramee Srivastav- Additional answer: I agree with David, but I'd like to add that while it's okay for everyone to try to reinvent the wheel, the problem is that very few of these wheels are actually functional. Many projects are missing crucial components. For example, they focus on Layer 3 solutions but often neglect the security aspect, leading to trust issues, especially with trusted setups. There's a gap in this area that needs attention and development.
David Amor- Additional answer: My point isn't just about the technology for on-chain games; it's about the lack of product-market fit for any on-chain games right now. People aren't playing them as much as they should be. What I'd like to see is games that find their audience and then build the infrastructure accordingly to support those games. We're not at a stage where enough games are reaching players and retaining them. We need games that people come back to regularly.
Chang Liu - Additional answer: I'd like to add that I've noticed infrastructure is ahead of games, and some game developers are building without paying enough attention to their NFT strategy. There hasn't been a successful Web 3 game that people played without holding NFTs first. Think about Axie Infinity; people play it because they hold the NFTs. Games need to have an asset and community fit first. If there's no IP fit with the brand, concept, marketing, and community, people won't play the game.
Sebastian Guillemot- Additional answer: I agree with Chang Liu. Currently, many game engines are building games themselves. For example, Paima, Curio, and others are both building games and engines. We've talked to hundreds of NFT projects interested in building games for the NFT community, and many of them have requirements that no engine currently fulfills. We're working toward meeting those requirements, but it's not that the infrastructure is far ahead; it's more about building what the market needs.
Yijia - Additional answer: I have a slightly different perspective. While it's true that many popular games have launched NFTs before the game, I don't believe the problem is with NFTs themselves. Cryptogames shouldn't rely solely on tokens and NFTs to attract players. The real challenge is in game design and addressing the players' needs. We can't tell the same story as Web 2.5 by promising players that their assets will appreciate. We need to find unique value propositions and create games that genuinely engage players, with NFTs playing a supporting role.
Changlu- Additional answer:I agree that utility shouldn't be solely based on speculation. It's more about having a strong IP fit. If there's a good match with a popular IP like Pokemon or maybe something else in the future, and the attention and engagement flow towards the assets you hold, then you become a stakeholder or an active community member for the game, which can drive wider adoption.
Norswap- Additional answer: I have a few points to make. Firstly, there's an example of a game where people can craft, similar to Minecraft on the blockchain. When you add NFTs to such a game, it can create interesting use cases. And speaking of extensibility, the founders of Primodium, which is another great game, were previously working on integrating blockchain with Minecraft. This allowed for a sense of sensibility, where you could have digital physics within a Minecraft-like world that operates within certain rules. So, it's fascinating how the community and infrastructure interact in these scenarios. There are many aspects to consider, such as icon abstraction providers, smart contract providers, and social interactions. Personally, I'm less concerned about game engines because I believe they should be open source, and perhaps they can provide services on top of that. However, it's important to maintain an open-source foundation.
To Shramee Srivastav:
Damian: I think this question isn't genuine but rather a complaint from the developer community. They've expressed concerns that your team often doesn't deliver on promised timelines and that issues frequently lead to delays or roadblocks. Do you believe that Dojo has issues with centralization and exclusivity, making it challenging for other engineers to contribute and improve the Dojo engine for game developers?
Shramme Guillemot: Yeah, I agree there have been some delays. Sometimes we are working on some tasks that depend on another tech. For instance recently we were trying to do optimistic execution on the client side, aiming to later prove this execution to the on-chain world. But I encountered an issue where compilation part of the process required 64 bit usize, but in our runtime we only had 32. These kinds of challenges arise and cause delays. But we are getting there.
To David Amor:
Damian: David, you've been a highly successful game producer and entrepreneur in the gaming industry since the 1990s. Compared to the games you've previously worked on, which had millions of daily active users, how do you plan to handle the negativity if your own game has only around 50 users? Do you feel like time is working against you?
David Amor: Haha, "time is against me" is just a playful expression. Okay, let me explain further. I've been involved in game development for quite a while, and I've witnessed how new ideas and technologies have transformed the gaming industry. For instance, the advent of mobile phones and games integrated with social platforms like Facebook opened up new possibilities for multiplayer gaming.
Now, with the emergence of autonomous worlds and blockchain-based games, I see the potential for a whole new genre of games. I'm genuinely excited about the concept, and I believe that even starting with just 100 concurrent players this year is a significant step. It may seem small, but it's the beginning of what I hope will be substantial growth. I see ample opportunities in creating and growing this new type of game.
So, in essence, I don't feel like time is working against me. Instead, I see the potential for growth and innovation in this evolving gaming landscape.
Damian: Considering the various game engines available, including MUD, Dojo, and Paima, and based on the answers provided earlier, what are your thoughts on whether there are too many engines in the market, and how do you perceive the differences between MUD, Dojo, and Paima?
Sebastian Guillemot: MUD primarily focuses on providing a Solidity framework for writing smart contracts compatible with games. DOJO, on the other hand, places its primary emphasis on ZK (Zero-Knowledge) technology within a gaming environment. In contrast, Paima specializes in software rollup, empowering on-chain games that use LINK as part of their framework. These approaches are distinctly different, with Paima being notably flexible. Paima's flexibility arises from not being bound by the limitations of Solidity or ZK circuits. This flexibility allows us to create immersive, extensible game worlds, as demonstrated by the games we're developing ourselves.
As a company, we have achieved profitability and garnered a user base of around 1,000 players for our games. Many game developers are adopting our platform to create their own games. In certain segments, Paima is the preferred choice, especially for deploying on-chain games in networks like ArbNova, where it offers cost-effective and user-friendly solutions. Additionally, for non-EVM (Ethereum Virtual Machine) environments such as Cardano, Paima stands as a prominent option. Therefore, we see Paima as a leading choice for various developers and projects.
Damian: In the scenario where Op Stack chooses to endorse MUD as its gaming engine, how would Keystone and Curio respond to this development? Considering the competition among various 'rollup as a service' projects, what are Keystone and Curio's strategies for standing out and thriving in this landscape?"
yijia: I believe the core question here revolves around the role of infrastructure in the realm of on-chain games. Currently, we don't perceive much competition in the field. The share of the pie in terms of successful game launches, compared to the overall potential, is less than 0.1%. Therefore, our primary objective is to expand this limited pie by attracting new players through innovative experiences that distinguish themselves from web 2 and web 2.5 offerings.
Our approach is grounded in developing and enhancing our infrastructure openly. We provide support for anyone who wishes to build on top of it. Our open-source model reflects our belief that Keystone, our infrastructure project, is not the only viable option in the space. We focus on continuous improvement while also developing our games and gaining insights into player behavior.
As the space matures and more games are introduced, player adoption will be a key factor. Market share and endorsements from various entities will become significant. However, it's essential to recognize that in the current landscape, endorsements matter less when there is a scarcity of on-chain game players. The priority lies in attracting and retaining users.
Regarding protocols like Optimism sponsoring, their success ultimately depends on traction and usage. At this point, we lack a substantial player base in the on-chain gaming space. Consequently, the impact of endorsements is minimal. We are continually evolving our infrastructure stack based on the evolving needs of our games, and this is a dynamic process. Similar considerations apply to Dojo and other projects in the space. Overall, I'm not overly concerned about endorsements in the current stage.
David Amor - Additional question: And I just got do you consider yourself an infra company or a content company?
Yijia- Additional answer: Well, that's a perspective from a content company regarding infrastructure. In my view, if you're a company operating in the on-chain gaming space today and you're not investing in infrastructure, it likely means you're not innovating. That's my take on it. If you're an on-chain gaming company and you're not actively involved in building infrastructure, it indicates that you're not exploring how to leverage blockchain technology differently from others.
It's worth noting that when we talk about engines, we're essentially referring to server-side engines, as opposed to client-side engines like Unity and Unreal. The infrastructure space is still in its infancy. If you're not doing something distinct or unique, you might find yourself merely emulating web 2.5 games, essentially doing what others have done before. This may not be a particularly promising path for long-term success.
From our perspective, building infrastructure is a means to address specific needs that arise from the games and projects we're involved in. For instance, we've developed infrastructure to support "Treaties," which we believe is quite unique in the space. It allows players to use smart contracts in real-time as social contracts, such as non-aggression pacts. Our aim is to ensure faster performance compared to existing engines, addressing specific bottlenecks we encounter.
So, to clarify, infrastructure serves specific purposes tied to real-world needs, rather than some abstract goal of achieving an autonomous world, which remains a loosely defined concept.