In this AMA, we discuss the ongoing #decentralize-ai writing contestand the main topics for the HackerNoon writer community: whether you're analyzing the current tech landscape of decentralized AI and ICP’s role in this field, discussing real-world applications of decentralized AI across various industries, offering tutorials on building decentralized AI models using ICP, or examining how decentralized AI can enhance data security, transparency, and trust—this is your chance to contribute.
See the writing prompts here and share your story to win $1,000!
Meet our guest, Jessie Mongeon!
Jessie is a technical writer at DFINITY and author of two Web3 books, 'The Ultimate Web3 Pocket Guide' and 'Mastering Web3 Documentation'. She has a Master's degree in Information Technology Management and has worked in Web3 since 2021.
In this AMA, Jessie shares everything about:
This Slogging thread by Mónica Freitas, Jessie Mongeon, Jonh, Asher Umerie, Jose Hernandez, and Sheharyar Khan occurred in slogging's official #amas channel and has been edited for readability.
Hello everyone! Excited to connect and answer questions 🙂
Hi Jessie Mongeon! Thank you for joining us!
Can you start by telling us a bit about yourself and your journey through Web3? How did you start?
Yeah, so I got started in tech back in high school, I went to a special technical school for CISCO routing and switching, and followed that through college. While in college I was a Linux System Admin for a company that did contract work for MIT, Harvard, BU, and some of the other Massachusetts universities. After doing that for a few years, I wanted to try and mesh my background in tech with my love for writing and made the transition into being a technical writer.
I took a position at a Web3 decentralized storage start-up where I got my initial experience in both technical writing and Web3 by spending a lot of time researching different projects and how the decentralized storage project I was working for could interact with them. I got involved in a lot of Web3 activities like Twitter Spaces, podcasts, hackathons, and online conference speaking opportunities. I tried to start my own Twitter Spaces series called 'Web3 Whitepaper Club', create a Web3 newsletter service, and get involved in different Women in Web3 communities like Crypto Tech Women.
While I was doing all of this, I saw that there wasn't a core location to learn about all things Web3 -- you had to go to dozens of different websites, courses, podcasts, youtube videos, etc to learn about all of the different concepts, technologies, and projects in Web3. So, to try and solve this gap in the Web3 space, I wrote my first book 'The Ultimate Web3 Pocket Guide' which was intended to be a 'Pocket dictionary' for Web3 concepts and projects, but at 500+ pages, it doesn't fit in any pockets 😂
In May 2023, I joined DFINITY, the foundation behind the Internet Computer Protocol network, and have since been working on the ICP Developer Documentation. I've written the ICP educational course 'ICP Developer Journey' that includes 36 written modules with video tutorials, attended hackathons and conferences, and help contribute to the DFINITY Developer Twitter/X account.
Hi Jessie Mongeon! Awesome to have you here!
How do you think traditional educational institutions can integrate Web3 into their curriculum? And what are the biggest challenges in educating people about Web3 technologies?
Jonh, Thank you! I've seen a few different educational institutions integrate Web3 into their curriculum successfully, but they've dedicated the entire course curriculum to Web3, rather than making it a single 'chapter', since there is typically too much content to try and mention in a super condensed format. From these institutions, typically they pick a single focus, such as blockchain architecture and design, rather than try to cover all of Web3 with tokens, NFTs, DAOs, metaverse, etc.
The biggest challenge with Web3 education is the initial onboarding. There are so many places to get started, so many different avenues to explore, concepts to learn. That was one of the challenges I was trying to solve by writing my book to try and provide a single source for initial onboarding. Ideally, people read the book and they then have a core foundational understanding of all the different Web3 components and concepts, then if they're interested in one specific area, they can learn more from there. Right now, there are a few great online courses or cohorts that onboard people really well, but typically they either have a high entry cost to join, or they are application-only, and not everyone who wants to learn gets accepted into the cohort or course. I think making these more accessible for everyone is another challenge.
Huge congrats on your first book! Will add that up to my list 😊 What was the process like for writing 'The Ultimate Web3 Pocket Guide,' and how did you manage to consolidate information from such diverse sources?
Also, can you tell us more about the ICP Developer Documentation and the ICP Developer Journey course you created?
Mónica Freitas, the process of writing it was constantly having 100+ tabs open, hundreds of spreadsheets of data, and a lot of coffee 😅 I primarily used the resources for each product/platform directly, so either their whitepaper or developer documentation. I avoided using blogs, youtube videos, or other secondary sources when possible. I focused on one section at a time, so I started with Blockchains and divided it into EVM chains and non-EVM chains. Then I wrote the chapter's introduction discussing how a blockchain works, what is consensus, what are the different methods of consensus, etc, then I looked at CoinGecko and found the top 100 chains, sorted them into EVM and Non-EVM, and went down the list alphabetically. I repeated that process for most of the other sections, but some sections like DAOs or NFT collections, I picked only the ones that were meaningful in some sense, not necessarily the most popular.
The ICP Developer Documentation can be found https://internetcomputer.org/docs/current/home and details all things Internet Computer- from deploying a simple 'Hello, world' smart contract to using the network's advanced cryptographic protocols.
The ICP Developer Journey is essentially answering the question of "how can we walk someone through each of the core concepts displayed in the developer docs, in a linear path that makes sense?" It starts assuming the user has no prior knowledge of ICP, and by the end of the course the user has learned about smart contracts, tokens, the network's DAO, the Bitcoin and Ethereum integrations, and can deploy a full stack application.
Thank you for the breakdown, Jessie Mongeon! Still on this topic:
Jonh, Great questions!
Jessie Mongeon, I can only imagine the number of hours of research :smiling_face_with_tear:
How did you balance technical depth with accessibility to ensure that readers with varying levels of expertise could benefit from your book?
Guess this could be particularly important for our decentralize-ai writing contest participants 😉
Mónica Freitas, honestly, I imagined my mom reading the book (because I knew she would be) but also because she is very non-technical, and would be a great example of an 'everyday user' who has no previous knowledge of crypto, wallets, NFTs, etc, and therefore the target audience for the book to be a true onboarding guide. For more technical users, I included some details in certain sections where they may be more curious, such as consensus mechanisms or decentralized storage methods like sharding, but I kept them high level enough where I wouldn't lose the less technical readers. I also included lots of diagrams and pictures, since personally those always help me understand complex concepts 🤓
Hi Jessie Mongeon, it’s a pleasure to have you with us.
Congrats again on your first book and all your great work so far!
I’m particularly curious about your impressive work with building the ‘ICP developer journey’ course.
Can you share an example of a module or lesson from the ‘ICP Developer Journey’ that you found particularly challenging or rewarding to create?
Building on that question. If you were to start curating the course all over again, with everything you’ve picked up on your journey so far, what would you do differently?
Asher Umerie, the most challenging module to create was the tutorial on vetKeys, our novel protocol for on-chain encryption. This was particularly difficult since the feature is still in early development and uses a rather advanced cryptographic concept, so understanding it fully myself was a challenge, then writing about it in a tutorial manner that properly explained it to the users was a separate challenge.
I'm actually going to be expanding on the Developer Journey series soon, so your second question is something I've been thinking about a lot! Right now, the series only uses basic Markdown files and doesn't have any interactive assets that a course may have, like quizzes or progress tracking. Those are both attributes that I'll be adding in the upcoming revisions to the series 🙂
Welcome, Jessie Mongeon! What are some misconceptions about web3 that people still seem to have?
Jose Hernandez, I think we still battle with misconceptions that Web3 is purely scam coins or JPEGs that people are buying and selling, while in reality the memecoin space actually has a good bit of community and utility and the NFT hype has seemed to die down. I think we need to get more media talking about the tech to help deter these misconceptions, but the tech isn't nearly as 'flashy' or 'click-worthy' as some of the past headlines like BAYC.
Jessie Mongeon, I've noticed that shift in Web3 perception as well, but it seems to be a conversation had in niches, not mainstream yet. Linking the education topic to developers:
Jonh:
Love that, Jessie Mongeon! Based on your work with ICP and other projects, what are the biggest pain points in the current Web3 developer experience, and how can they be addressed?
Mónica Freitas, the biggest pain points I've seen across Web3 are:
ICP addresses both of these pain points: First, ecosystem lock-in is addressed through ICP's Chain Fusion technology, meaning a smart contract on ICP can sign and submit transactions directly to other chains like BTC, ETH, EVM-compatible chains, etc. You can write a single smart contract on ICP that interacts with several chains at once, allowing you to move and pivot across chains much easier.
ICP also uses the reverse-gas model, where developers pay the application's gas fees, so users don't have to connect a wallet and sign transactions for interactions like making a social media post or uploading a video, they can interact with the application like its a Web2 app, which helps foster application onboarding and adoption.
Hey Jessie Mongeon! Thank you so much for taking the time to talk with all of us. How different is the process of writing a technical document for web3 than it is for web2? Did you ever find yourself in a situation where you thought there was too much or too little information for you to work with as compared to web2?
Sheharyar Khan, Web2 and Web3 technical writing aren't so different and the writing process itself is quite similar - outlining a document, identifying an audience, etc, it's more that the sources for each are drastically different. Web2 tech dates back to the 90s for a lot of concepts, so some data is quite old, but still the best source of that information.
In Web3, everything is still pretty new, and in a lot of situations myself, I am working directly with the engineers who are creating the new tech as my primary source of information. Sometimes that is enough to create comprehensive documentation for the developer, but in other situations, the reader has specific questions or runs into a unique error message, and they can't just Google that information and find tons of Stack Overflow answers like they can for a Web2-based question, so then there is some user friction that might occur that for Web3 docs. You need to anticipate these friction points ahead of them whenever possible and try to plan for them in the initial document.
There's also a lot of noise in Web3 for general education, and it can be hard to wade through what is quality information and what is clickbait or already outdated. The space moves and develops so quickly, and it seems like few resources are committed to consistently staying up to date on every fact they include in their content.
Your insights on ICP are super interesting, Jessie Mongeon. It's awesome to see how ICP is tackling big issues like ecosystem lock-in and user interaction. The Chain Fusion technology and reverse-gas model sound like game-changers for both developers and users. To dive deeper into those topics, on ICP chain fusion:
And Reverse-Gas Model:
Mónica Freitas:
Surprised by ICP's innovation in this field. What advice would you give to Web2 developers who are hesitant about transitioning to Web3 due to perceived complexities or common misconceptions, and how do you address these barriers? Jessie Mongeon
Jonh, I would ask Web2 developers what things in Web2 they wish were better or different, and likely there is a Web3 solution that introduces an improved solution, whether it be better security, technology, data ownership, etc. I would recommend they research those Web3 solutions, and then it may inspire them to try and use those solutions or build something that provides that solution using Web3 tech.
I think addressing these barriers by trying to instill excitement and inspiration is the best way to naturally onboard both developers and Web2 developers.
Great to hear, Jessie Mongeon I’ll be looking out for those updates.
On the topic of the web2-web3 divide: for developers coming from a traditional software background, what are the biggest adjustments they need to make when transitioning to Web3 development?
Asher Umerie, the biggest adjustment is usually the role that tokens play in the developer experience on Web3, whether they use a chain like Ethereum with a 'traditional' gas model, or if they use ICP and our unique reverse gas model. Web2 developers are used to just putting in a credit card number for paid services like AWS or Digital Ocean, whereas on Web3 they have to learn how to obtain a wallet, secure that wallet, and then load it with tokens. And, when they pay gas fees, it may not be as consistent or predictable as paying a monthly AWS invoice. I think that is always the biggest difference that traditional devs have to get used to.
Interesting, Jessie Mongeon! There may be however some skill gaps when migrating from web2 to web3. What kind of tech challenges might Web2 developers hit when switching to Web3, and how can they get past them?
Jonh, since the tech behind Web3 is so different fundamentally from Web2, the challenges they might hit are almost endless depending on their specific Web2 experience and background. Some of the most common ones I think is understanding the concept of deploying their smart contract on-chain versus simply pushing the code to a live server. The general idea shares a lot of similarities, but how the tech itself works is very different. To get past them, there are several different intro to Web3 courses and blog tutorials that do a good job of outlining the steps, but honestly IMO the best way to overcome them is through developer trial and error, using the tools yourself and seeing which workflows have which results.
When you point out chatgpt I'm starting to get the picture. How does the tooling for Web3 development compare to Web2? Is it more complex or pretty similar once you get the hang of it? Jessie Mongeon
Mónica Freitas Tooling will vary based on what you're trying to build and on what platform/ecosystem. There are some platforms dedicated to making tools feel as Web2-as possible, while others embrace the unique Web3 qualities in their tooling and try to provide tons of docs and onboarding resources to properly educate devs on their unique tooling.
That makes sense, Jessie Mongeon! I'd love to discuss the #decentralize-ai Writing Contest a bit more. What kind of submissions are you hoping for? Are there any specific topics you're particularly interested in reading?
Mónica Freitas, I'm hoping to see different, unique use cases for decentralized AI. Right now, we're seeing a lot of talk on social media about image generation or image classification. I'm hoping that we see submissions that talk about use cases other than these, such as using DeAI for community data aggregation and analysis, supply chain, workflow automation, etc. I also want to see submissions about different tooling that developers are using for AI, so I can learn what we need to include in the Dev Docs for ICP developers.
Jessie Mongeon, what inspired you to launch a writing contest focused on decentralized AI? In your view, why is now the right time to explore this theme?
Jonh, we're doing a huge marketing push around AI because we have some super exciting AI-focused milestones on the ICP roadmap for the rest of this year. The founder of DFINITY and ICP has done a few really exciting AI demo videos on Twitter that showcase the current capabilities of AI on ICP, and those capabilities will only expand 10x once we launch some of the tech that's on our roadmap. So, right now is a great time to get people thinking about it alongside ICP.
Jessie Mongeon, thanks for your responses! A quick q for you -- what role do you think edge computing is going to play in decentralized AI (if any!)?
Sheharyar Khan, if the edge computing nodes have GPUs, then I think it will play a huge role in enabling large LLMs to run quickly and efficiently. If they're traditional CPU nodes, then I don't think it'll play a huge role. I think the main takeaway is the hardware and prioritizing GPU resources for AI-based workflows.
That's huge for our writers - they get to play a part in ICP Dev Docs!
How can participants ensure their submissions stand out by addressing novel use cases or showcasing unique tools for decentralized AI? Jessie Mongeon
Also, what advice would you give to participants to help them succeed in this contest? Are there any common pitfalls that participants should avoid when writing their submissions?
Mónica Freitas, I think submissions will stand out when they are clearly intentional with what they're trying to showcase, and the reader can get the sense that the writer is either excited about or genuinely wants to use the tech, instead of someone writing a tutorial just to fulfill the submission requirements. I think intention will make some submissions stand out more than others.
I would suggest that participants try to craft a comprehensive submission, and pick one narrow focus and really develop that fully. Sometimes when the prompt is vague, writers try to cover as many possible verticals as possible in the piece without fully composing one versus another. I'd recommend that writers pick one focus, then try to narrow it even further, then write to their heart's content about that specific idea or tool. Also, if you're writing something, say a tutorial submission about how you used a tool to achieve a goal, and you don't get it working - write about that! In contests like this, it doesn't have to be a perfect, production-ready solution. Learning where you ran into problems and got stuck is equally valuable for people who want to learn and explore the same sort of ideas.
Hi, Jessie Mongeon, regarding the #decentralize-ai contest. What would a high-quality submission look like to you? And on the flip side, what would you call a low-quality submission?
Asher Umerie, a high-quality solution would include a clear use case or idea for DeAI, a genuine intention and interest in the tech or idea, and comprehensive content that supports their use case or idea.
A low-quality solution would be something written by ChatGPT or something that may not be written with genuine interest in DeAI and instead is just written to meet the submission guidelines.
Great tips, Jessie Mongeon! Last question from me: considering the importance of narrowing down focus and showcasing genuine interest, are there any specific aspects of decentralized AI that you think are underexplored or particularly ripe for innovation? This might help participants identify unique angles or topics that could make their submissions stand out.
Mónica Freitas, I think a lot of people are focusing on AI for content generation, like text or images, but I'd love to see some submissions on how AI can help automate workflows and improve daily tasks. I've seen a few products where AI is being used as a 'second opinion' for things like medical analysis, product suggestions, and editing content. I think using it as a secondary source rather than a primary source of info/content is a great use case to explore.
Thank you for all your insights! Before we go, what would be the main takeaway you would like people to get from this AMA and your contest?
Mónica Freitas, I'd like the main takeaway from this AMA to be that there is still a lot of education that needs to be done around Web3, and one of the ways we can accomplish that is by participating in contests such as the decentralized AI writing contest. Creating resources that talk about how developers and users can benefit from Web3 tech in all sorts of different applications and use cases, can help continue to onboard folks into the ecosystem 🙂
That’s a wrap for this AMA!
Thank you Jessie Mongeon for your time, thoughtful answers and attention to detail. We’re excited to follow your journey from here on out and see what you do next!
To all the writers on HackerNoon, make sure to leverage the valuable insights shared during this conversation when crafting your entries for the #decentralize-ai writing contest.
We can’t wait to read what you come up with!