AI HACKATHON
London, March 8-10th 2024.
Our first in-person AI event!
We had a blast!
On March 8-10th, 600+ of our community members joined us for our first AI Hackathon!
The event featured some incredible workshops, bounties, and networking opportunities.
Find the full summary of the event here.
A unique event in the heart of London.
-
Delve into GenAI's versatile applications, explore how NLP is revolutionizing communication, and uncover the depths of Deep Learning and Neural Networks in driving AI forward. Our expert-led sessions will illuminate these technologies' transformative impact across various industries as well as the implications of the EU’s new AI policies and initiatives.
-
Connect with top AI leaders, exchange groundbreaking ideas, and explore collaborations that shape the future. Be inspired by visionary talks, engage in dynamic discussions, and build lasting relationships in the AI community. Register so you can be the first to know once our speakers are announced!
-
At an AI hackathon, you could build innovative solutions like custom chatbots, predictive analysis tools, healthcare diagnostics applications, personalized content generation systems, and smart automation technologies. These projects can leverage AI to solve real-world problems, enhance user experiences, and push the boundaries of technology and creativity.
Get inspired.
-
Dive into the world of Generative AI, exploring Large Language Model (LLM) chatbots like ChatGPT, Copilot, Bard, and LLaMA. Additionally, delve into the realm of text-to-image AI art systems, including Stable Diffusion, Midjourney, and DALL-E, to create visually stunning and innovative artworks. Build projects for music composition, image generation, and personalized text writing, tailored to sectors like healthcare, legal, and cybersecurity.
-
Merge Realities for Enhanced Experiences: Combine Augmented Reality (AR) with AI to create immersive and intelligent applications. Whether it's enhancing retail shopping experiences, developing interactive educational tools, or innovating in entertainment, your project can blend the physical and digital worlds in unprecedented ways.
-
Harness AI for revolutionary decentralized apps in Web 3.0. Develop smart contracts that react to market changes, AI-driven DeFi platforms, and NFT marketplaces with sophisticated recommendation algorithms.
-
Pioneer Quantum Computing and AI integration. Create faster problem-solving algorithms and quantum ML models. Aim to address global challenges in healthcare, finance, and logistics.
Schedule.
| 14:00 | 🛎️ Doors Open |
| 15:00 | 🧑🤝🧑 Team Formation Session |
| 16:00 | 🏁 Kick-off Event |
| 16:15 | 🎙️ Intro to ICP & Elna Jordan Last (Demergent Labs), Arun PM (Elna) |
| 16:30 | 🎙️ Open Data Economy x AI with Filecoin & Protocol Labs Network Yan Bo (Filecoin Foundation) |
| 17:00 | 🎙️ Intro to Stability AI Tom Mason (Stability AI) |
| 17:30 | 🎙️ AI for Cyberdefence: Bridging the Reality Gap Vasilios Mavroudis (AI UK - The Alan Turing Institute) |
| 18:00 | 🎙️ Keynote: The State of AI and the Promise of Computational Liberty Harry Grieve (Gensyn) |
| 18:30 | 🎙️ Co-owned, Human-Curated AIs for Gaming Matthew Tang (Virtual Protocol) |
| 19:00 | 🎙️ Accelerate AI Application Development with MongoDB Atlas and Atlas Vector Search: A Primer Harshit Mehta (MongoDB) |
| 19:30 | 🍽️ Nibbles |
| All night | 🛠️ Hacking (venue open all night!) |
| All day | 🛠️ Hacking continues! |
| 10:00 | 🎙️ ChatGPT Adoption Methodology Iliya Valchanov (TeamGPT) |
| 10:30 | 🎙️ Explainable AI in Lending Arjun Perumalla (Fuse) |
| 11:00 | 🎙️ From Potential to Practice - Delivering AI Solutions in Life Sciences Jaisal Kapoor (Faculty AI) |
| 11:30 | 🎙️ Pitch Perfect: Crafting Success Through Storytelling Monique Grinstein (The LEGO Group) |
| 12:00 | 🎙️ Generative AI’s Impact in Banking: Opportunities and Challenges David Vinagre Solans (Santander) |
| 12:30 | 🎙️ Beyond ChatGPT: Empowering Industry with AI Agents & Web3 Tools Maria Minaricova (Fetch.AI) |
| 13:00 | 🍽️ Nibbles |
| 14:00 | 🎙️ An AI Ethics Approach to Identifying Algorithmic Bias Rafael Kollyfas (CuAI Ethics), Alva Markelius (CuAI Ethics) |
| 14:30 | 🎙️ Fireside Chat with Didar Gelici Didar Gelici (Just Eat) |
| 15:00 | 🎙️ Ideation, Funding and Growth in Tech Bryony Disbury (VDK Capital) |
| 15:30 | 🎙️ Synthetic Image Recognition: How Gen AI is Transforming the Retail Space Patric Fulop (Neurolabs) |
| 16:00 | 🎙️ A Practical Guide to the Current State of Large Language Models Stas Chistyakov (London Stock Exchange Group) |
| 16:30 | 🎙️ Add Privacy to Machine Learning with zkML Laurence Kirk (Extropy.io) |
| All night | 🛠️ Hacking |
| All morning | 🛠️ Hacking |
| 11:30 | ⏰ Submissions Due |
| 12:00 | 🍽️ Nibbles |
| 13:002 hours | 🧑⚖️ Judging |
| 16:00 | 🏆 Finale and Prizegiving |
| 18:00 | 🛎️ Doors Close |
Partners.
Society Partners.
Universities that joined
our previous events.
Speakers/
Mentors/Judges.
Adnan Salehin
JPMorgan Chase & Co. (Judge, Mentor)
Laurentiu Ionele
JPMorgan Chase & Co. (Judge)
Laurence Kirk
Extropy.io (Speaker)
Ishan Misra
Facebook AI Research (Judge)
Edoardo Pona
King’s College London (Mentor)
Stanislav Sirakov
LAUNCHub Ventures (Judge)
Arjun Perumalla
Fuse (Speaker)
Jimmy Ye
OKX (Mentor)
Bryony Disbury
VDK Capital (Speaker, Judge)
Rafael Kollyfas
CuAI Ethics (Speaker, Mentor)
Gökhan Er
IOSG Ventures (Judge)
Matthew Tang
Virtual Protocol (Speaker)
Ilyas Bakbergenuly
LMAX Group (Mentor)
Eliot Frazier
AI for Good Foundation (Judge)
Jaisal Kapoor
Faculty AI (Speaker)
Alva Markelius
CuAI Ethics (Speaker, Mentor)
Arun PM
Elna (Speaker)
Jordan Cambe
Armalytix (Mentor)
Abdul Salam
Stability AI (Judge, Mentor)
Harry Grieve
Gensyn (Speaker)
Jyoti Y
Blockchain.com (Mentor)
Tom Mason
Stability AI (Speaker)
Vasilios Mavroudis
AI UK - The Alan Turing Institute (Speaker)
Iliya Valchanov
TeamGPT (Speaker, Mentor)
Stas Chistyakov
London Stock Exchange Group (Speaker)
Didar Gelici
Just Eat (Speaker)
Harshit Mehta
MongoDB (Speaker)
Jonathon Carr-Brown
Healthily (Mentor)
Riya Chhikara
London School of Economics (Mentor)
Zachary Scott
LMAX Group (Mentor)
Jenks Guo
Filecoin Foundation (Judge)
Jordan Last
Demergent Labs (Speaker)
Monique Grinstein
The LEGO Group (Speaker)
Alex Cheema
HomeDAO (Mentor)
Ella Hughes
Mercuri (Judge)
Anushka Kulkarni
Thought Machine (Mentor)
Pablo De Remedios
Cocoa (Mentor)
Lilly Reppin
Faculty AI (Judge)
Yan Bo
Filecoin Foundation (Speaker)
Patric Fulop
Neurolabs (Speaker)
David Vinagre Solans
Santander UK (Speaker)
Bounties and Prizes.
Top Prizes: $5,000
Exciting gear like drones, keyboards, earphones and phones!
-
One-year long subscriptions to each member of the winning team (with a maximum of 6 individuals).
🎁 This gift would include:NordVPN - Ensuring secure and private online experiences.
NordPass - Simplifying password management and enhancing security.
Incogni - Safeguarding online privacy by removing personal data from various databases.
-
Check out the bounties our Sponsors are offering!
-
💰 Prizes
1x 2,000 USDC
1x 1,000 USDC
3x 500 USDC
All winners will be invited to become SuperteamUK members for access to exclusive perks
🏋️ Challenge
Superteam UK, a community of builders on Solana, will support all projects & builders during and after the hackathon.
Your challenge for the Encode AI Hackathon is to build an AI solution within one of our selected emerging tech industries, then after this in-person hackathon you’ll have the opportunity to continue development by integrating the Solana blockchain and submitting into the $1M Colosseum hackathon.
If you’re not already aware; Solana is a high-performance, open-source blockchain platform. Solana offers many innovations like no mem-pool, parallel transaction processing, compressed NFTs, etc. You can read more about Solana here.
Selected Industries:
DePin
Gaming
Consumer Apps
Payments
Defi
Tooling / Infra
Focus on the AI part of the complete solution during the course of this hackathon, and then post-hackathon work with SuperteamUK to continue building into the Solana Online Hackathon by Colosseum.
Winners of this Online hackathon will also be accepted into Colosseum’s accelerator program, and have the opportunity to secure $250k in seed funding for their project.
⚖️ Judging criteria
Impact: Does the proposed problem make sense or be helpful to anyone? How large is the effect on the Solana ecosystem?
Solution: How clever is the solution? The rule of thumb is the less code a solution requires, the better it is.
Completeness: How far is the code base from production?
Applications: How does your AI solution apply to blockchain, and utilise the innovative features of Solana
📚 Resources
-
💰 Prizes
Up to US$25,000 in ICP (DFINITY Developer Grant)
The Internet Computer enables truly decentralizing AI by running services on-chain. We are proud of community achievements like having deployed the first ever on-chain LLM and vector database, training Machine Learning algorithms in a canister, implementing practical Zero-Knowledge AI and porting popular ML/AI frameworks and libraries.
We're thrilled to have you join the DeAI on IC community and can't wait to see what you'll accomplish.
Will you maybe port your favorite ML algo to the IC?
How about getting a new LLM to run in a canister?
Perhaps in another language?
Can you get a popular vector db deployed to the IC?
Or will you leverage existing IC components to build ML/AI into your app?
Dream big 🙂
⚖️ Criteria
Your product should be built on the Internet Computer. We judge this bounty based on:
The overall business case of the dapp, i.e. how sound is the business model, how innovative is the product etc.
How well it makes use of ICP technology.
The overall quality of the code.
The design of the user interface and the user experience
Bonus points for providing feedback on the developer experience.
Further acceptance criteria:
Readme that explains the project and provides information on how to deploy and use.
Open source code preferable (MIT or Apache-2.0 license), though optional
Canister ID(s) of deployment on ICP
📚 Resources
For resources, check out the Hacker Pack.
-
💰 Prize:
🥇 First place: $3000
🏅 Two runners-up: $1000 each
The winners will also get a chance to be considered for a Virtual Protocol ecosystem grant (up to $200k).
🏋️ Challenge
We invite you to develop a generative AI solution that offers a compelling new experience or solves an existing problem in the fields of Gaming, Social, or Digital Entertainment. Your submission should leverage Generative AI to either enhance user experience, streamline content creation, foster social interaction, or offer an innovative entertainment experience. In order to spark your creativity, we’ve proposed several potential directions you could explore, but it’s ultimately up to you and your interests!
🧭 Possible Directions
AI Assistant/Companion: Create AI companions that either have practical utility or simply serve to provide you with entertainment/companionship.
AI Livestreamers: Create AI-powered livestreamers that have the ability to react to comments, sing, dance and more.
AI RPG Game (e.g., Dungeons and Dragons): Develop an AI-powered RPG game where AI shapes the narrative, adapts the challenge based on player actions, or even acts as an NPC.
AI Game Coach: Invent an AI coach that offers personalized training, strategy optimization, and performance analysis for gamers, elevating their skills in real-time.
Revolutionising game production with Video LLMs: Develop a set of game mechanics that allow for simple interactions (e.g., aiming and shooting) using mouse inputs or screen taps, while utilizing a Video LLM to generate video content in real-time, reflecting the outcomes of player actions.
Open: Let your imagination run wild, unleash your creativity and surprise us!
📤 Submission
Project Details: A detailed description of your AI application, including the use-case, target audience, and potential market impact.
Prototype/Demo: A working prototype or demo video showcasing the AI application's functionality.
Code Repository: Access to the source code in a public repository for evaluation (e.g. Google Colab)
👩⚖️ Judging Criteria
Innovation (20%): Does the solution present a novel approach to gaming, social, or digital entertainment? Does the submission demonstrate creative problem-solving and out-of-the-box thinking? How much of a role does AI play in the solution?
Impact (20%): What is the business case for the solution and the market opportunity? What is the potential for the solution to attract a user base and its scalability in the current market?
Technical & AI Excellence (45%): Is the solution technically challenging? Does it make good use of generative AI capabilities? Were there any creative methods used to enhance model quality (e.g. using unique/synthetic datasets, finetuning, etc.)
Demo (15%): How clearly is the project and its functionality communicated in the submission materials? How effectively does the prototype/demo showcase the AI application’s capabilities?
📚 Resources
For resources, ideas and additional guidelines check out Virtual Protocol’s Hacker Brief.
-
💰 Prize
Up to 50x Healthily lifetime subscriptions
Use the Healthily DOT platform API (and anything else appropriate) to match users with appropriate services based on their symptoms.
💬 Problem Statement
60% of insured people are unclear of which services they should use, resulting in wasteful resource allocation and an increased demand for services. For example, 30% of patients using telemedicine services are doing so for self-care conditions, and a further 30% should have sought a face-to-face appointment instead.
The task is to create a flow which rebalances the asymmetry of knowledge between the patient and healthcare professionals, based on their symptoms and the services available (generalist vs specialist services - physiotherapy, mental health etc instead of a GP).
🔥 Project tips
Please note a solution using LLMs to triage users will not meet safety standards.
This does not rule out the use of LLMs but they should not be used to provide guidance.
This is a behaviour change problem as much as a technology problem.
The users will be anxious, frustrated and dependent on healthcare professionals.
The solution needs to explain the benefits to the user.
📚 Resources
Find the API key sign up in our Hacker Pack
-
💰 Prizes:
1st place: $2,500
2nd place: $1,500
3rd place: $1,000
Challenge:Best project submitted with the potential to turn into a startup!
Filecoin &/or Protocol Labs tech includes: Filecoin, Filecoin VM, InterPlanetary Consensus (IPC), Bacalhau, Coophive, & more (see here)
Build Ideas:
🚀 AI Data Explorers: Quality data determines the quality of many AI projects. Filecoin’s network is great for storing raw data and training AI models. We encourage ideas that combine Filecoin with AI tooling, e.g. a data explorer that helps AI companies to find correct public data on Filecoin for training.
🚀 ML Model Storage & Augmentation: The Filecoin Virtual Machine (FVM) provides an ability to decentralize the storage of model training data and the models themselves. (See more here.) Consider how powerful it could be to create better weights and biases on an LLM, based on an objective evaluation set (such as Human Eval). And go one step further to think about how incentives (e.g. new model architectures, new inference tools) could even enable participants to bootstrap from an open source model and fine-tune it to the point that it starts to outperform known hosted models.
🚀 DataDAOs: DataDAO & DAO tooling solutions on FVM to enable curation, use, storage, analysis, monetization, & governance of valuable datasets using the FVM DataDAO starter kit.
🚀 IPC subnets for AI: Experiment on IPC, leveraging throughput properties IPC subnets and/or tooling to build large scale apps (compute networks and more). Ideas: build subnets for games (board games, turn-based games, or similar), subnets for computation (Compute over Data, programming pipelines, etc), or subnets with interesting FVM runtimes.
🚀 Harness Bacalhau: Bacalhau is a Filecoin ecosystem-native peer-to-peer computation layer enabling compute over data jobs like GPU-enabled AI, ML, analytics, data engineering, data science, de-sci and more.Evaluation Criteria:
- Innovative Idea: An idea that’s technically creative, and solves a real problem innovatively.
- Demonstrated understanding of the technology: Clear perspective on why you chose this technology or its associated ecosystem for your use case.
- Product Prototype: A working prototype demonstrating how the idea comes to life for its users.
- Strong Team: A strong team comes with a great mix of skills, including technical expertise, ability to execute and a strong understanding of Product..
- Business Model: A simple plan (at this stage) with the potential to generate revenue or create value through other means.
- Great Pitch: A clear, concise pitch that summarizes the above.
NB - building on IPFS alone does not qualify.
-
Prize: 1 x $1000
Challenge:
Most innovative use of CoopHive
Build Ideas:
🚀 Decentralized AI apps: CoopHive supports Stable Diffusion XL (SDXL) to generate high quality images from text prompts. Using the SDXL module, build a dApp which allows users to generate high quality images. Your choice of use-case! E.g. marketing images, featured images for blog posts, just to name a few examples.🚀 Advanced:
Containerize your choice of AI model: CoopHive supports user-provided modules, so you can package your choice of AI model as a simple git repo. Pick a model from huggingface, for example a speech to text or text to speech model, Dockerize it (example), get it running on Bacalhau, then wrap the Bacalhau spec in a CoopHive module (examCheck out the CoopHive community on Discordple). Use this to build an even more innovative decentralized AI app, as described above.Resources:
-
💰 Prize
1 year Stability AI Membership and 20,000 Stability AI API creditsStability AI is the world’s leading open source generative AI company. They deliver breakthrough, open-access AI models with minimal resource requirements in imaging, video, 3D, language, code and audio.
🏋️ Challenge
Unleash your creativity and innovation in our open challenge, using Stability AI to build something amazing!We can't wait to see what you envision and bring to life - the possibilities are endless!
📚 Resources
Visit our Hacker Pack to claim your Stability AI API credits!
-
💰 Prize
Limited edition AI Hack t-shirtsExamine machine learning models to understand how they can exhibit biases in their decisions and explore ways to mitigate these biases.
🏋️ Challenge
Identify biases in a Harry Potter-influenced loan approval dataset (to be released on the first day of the hack).Re-evaluate the models from both technical and socio-technical perspectives:
Presenting your approach, findings and computational results
Share a brief reflective report on the socio-technical implications
Implement some mitigation strategies (as will be discussed during our workshop).
(Once the Submission Form is opened, the above should be submitted via a linked document, or in the ReadMe section of your GitHub repo.)
📄 Additional context
Real-life examples of algorithmic biases in Machine Learning models include:
A dataset related to loan approval processes, in which a model might prioritize or penalize applicants based on postcodes, reflecting socio-economic biases.
Another could involve medical treatment recommendations, with potential biases emerging based on the model's interpretation of gender or race.
📚 Resources
The dataset will be released on the first day of the hackathon!
FAQs.
-
No, the event is free but but we will ask you to confirm your spot!
After you have registered, you will receive an email asking you to confirm your spot.
If you are having any issues regarding the deposit, please get in touch with us!
-
Anyone can take part in this hackathon, as long as you’re there in person!
-
No, you can hack solo.
If you’d like to form a team, we will help you connect with other hackers.
-
Yes! The Encode team will be on the ground to direct you to the best person to address your technical query.
Some of our mentors will also have office hours available to sign up for!
-
For updates, announcements, and everything else on the organisational side, please keep an eye on Encode’s Telegram and twitter!
The dedicated telegram chat for this event is here.
-
Yes, you can, but only if your project hasn’t raised significant funding already and significant feature changes have been made during the hackathon weekend.
-
Yes, you/your team is the sole owner of your project.
-
Yes! We actively encourage this.