Update from the COO: The Singularity draws closer…

Can you believe it is December already? Hello to all our staunch and loyal community, and greetings to new community members! Welcome to our Q3+ operations update, covering progress from August to the end of October and some recent highlights from November. Key topics are summarized here as we are already in the process of preparing our year-end wrap-up.

Highlights from me include:

• It is a huge honor to welcome on board Khellar Crawford as our new AI Chief of Staff. As our workload and team continue to grow, Khellar is supporting our Chief Science Officer, Matt Ikle, in overseeing the wide and varied range of frontier technology our brilliant AI teams are developing
• Per our commitment at the beginning of 2023, we have reaffirmed and continually ramped up our commitment to the SingularityNET platform development progress throughout the year. In Q3 we are excited to announce the onboarding of an established DevRel expert team, VBRL, to add momentum and capability to our platform development
• A nonstop schedule of events and presentations kept SingularityNET in the media, including our WebSummit Lisbon media events, which reached a staggering audience of 36M. See us here in BTC-ECHO, Infobae, Forbes Mexico, Associated Press, Expresso and SAPO
• Our warm friendship with Cardano continued to grow with the release of AGIX-ADA staking on Cardano, attendance at the excellent RareEvo event in Denver in August, and it was an honor to bring Desdemona Robot and speak at the amazing Cardano Summit in Dubai
• The first ever SingularityNET global AGI Retreat (well, the first since I came on board in April 2021!) was an astounding success with high-energy advanced technology and debate, propelling and adding clarity to our AGI core mission. What a total honor to finally meet so many of our highly esteemed scientists and technologists
• The 2023 leadership retreat, also in Dubai, further crystallized our Knowledge Layer strategy and progressed both the coming together of the leadership team as a powerful delivery force, and several strands of technical directions, including the integration of Hyperon with Blockchain — more to come on this…
• Our partnerships proposition continued to grow and strengthen SingularityNET’s position as the go-to decentralized AI provider
• Continued growth and progress from both our in-house spin-offs and those already spun off, expanding our impact and capabilities across a number of vertical markets
• It has been fantastic to welcome our new Supervisory Council members and begin laying strong foundations for progressive decentralization via our new taskforce, and also via internal restructuring to expand the decision-making layer at the foundation. Decentralize, decentralize, decentralize!
• Finally, we are close to completing our transition of foundation HQ from the Netherlands to Switzerland. This has involved much dedicated and detailed work on our accounting, financial, and legal standards, processes and reports. We are confident that our new home in Switzerland will provide the perfect environment to accelerate our continued decentralization and expansion.

New Executive Team member — Khellar Crawford, AI Chief of Staff

Khellar brings over 23 years of experience as a serial startup founder, investor, and advisor with a unique balance of technology, product design, and management expertise. Having spanned many industries from adtech to Hollywood to fintech, he’s led innovative initiatives for teams from Paramount to Microsoft that have scaled out to reach hundreds of millions of households globally.

In his most recent venture, Otomo, Khellar designed and patented a self-organizing personal money management system. This novel approach enables banks and credit unions to align their profits with the financial well-being of their customers, offering a meaningful alternative to traditional fee-based models. His innovation was recognized globally in 2021 as one of Accion’s Inclusive Fintech 50, showcasing groundbreaking contributions to inclusive financial services. While continuing his leadership role at Otomo, Khellar now brings his deep expertise to enrich the SingularityNET ecosystem.

Khellar’s vision for SingularityNET transcends the boundaries of technology.

Khellar is captivated by the potential of AGI to foster unity, harmony, and humanity, viewing the integration of technology and human interaction as both immense and exhilarating.

He firmly believes in SingularityNET’s approach to introducing this powerful technology responsibly, as an asset for all humanity, that should belong to no one and everyone.

As AI Chief of Staff, Khellar Crawford is not just a leader; he’s a proponent for change, fanning the flames of innovation and fostering a culture of collaborative progress within SingularityNET!

Community Questions:

Q: What is the status of Staking on Cardano?

We have gone live with staking on Cardano, with a list of volunteer users and testers for the first three promotional epochs. We were thrilled to announce the launch of our Cardano staking solution during the Cardano Summit; many in the community and at the Summit volunteered to provide early testing and feedback and received extra rewards in appreciation.

The first exclusive staking-on-Cardano epoch has been completed, with highly positive feedback, and the second epoch is underway, expanded to a larger group of volunteers, again with additional rewards.

The third epoch will round out our exclusive rounds of staking-testing volunteers, and staking will be open to all by the end of the year. See the blog here for full details: https://blog.singularitynet.io/agix-staking-on-cardano-explained-974881ca4c71

Q: What is the status of the airdrops for launched Phase 2 projects?

Our laser focus for H1 has been on supporting the launch of ecosystem spin-offs during this period of rapid growth. We are making good progress and are excited by the success of the launches throughout 2023, as well as the successes of these young projects in building communities and products.

Behind the scenes, we are working to plan and coordinate the airdrop and claiming process across projects. As these projects are building their success, their next priority is demonstrating their appreciation for the AGI community for their support and extending the benefit of the products they are building. SingularityNET aims to coordinate a maximally efficient and seamless experience for the community. We’ll provide more details on timelines and procedures as they solidify.

Milestones Update — Q3+

SingularityNET Platform/Marketplace Updates:

We are thrilled to share our roadmap, outlining exciting developments that empower developers, streamline processes, and propel AI innovation. In line with SingularityNET’s strategic growth initiatives, we have engaged VBRL, a specialized software house, to assist in several key areas by leveraging their WEB3 expertise. In addition to the ongoing improvement and intensification of platform module updates, we have launched several strategic technical partnership initiatives to expand the capabilities of users and service developers.

Progressed with Daemon, CLI, and SDK Transformation, developing new tools:

Platform Daemon. As part of the work on the daemon, all third-party components and libraries were updated to the latest versions. To make the work of service providers more convenient, a new way of interaction between the daemon and the service using HTTP requests has been implemented in the daemon. Previously, connection to the daemon was limited to grpc, json-rpc, and process (threaded interaction with the running provider service). However, service providers now have options for deploying their services on hosting based on serverless GPU computing (for example, banana-dev). This functionality is currently being tested and debugged. Also, we implemented new functionality for service providers to train models. As part of interacting with the daemon, this allows service providers to activate an additional model training service based on the provided data dataset. The functionality is being tested and improved to meet technical requirements.

CLI and SDK. An in-depth revision of the current Platform architecture and its components was carried out, and a number of possible areas for improving and optimizing internal processes were identified. The stage of updating the libraries and external components of the CLI and SDK has been completed (the update of SDK is still ongoing), and some external modules have been optimized or replaced with the updated ones.

The team is currently working on improving the use of SDK as an independent Platform element so that the use of the CLI and any third-party components is no longer required. This will greatly improve the user experience for service developers.

The transition to a new testnet environment is also in progress. The CLI is currently under active testing and bug fixing.

Supportive modules and tools

We also launched developing a range of boilerplates and templates. These resources are designed to streamline the process of integrating AI services into various applications and systems.

1. Commenced development of Text User Interface (TUI):
   We are progressing with developing a Text User Interface (TUI) written in PyTermTK. This interface is designed to simplify the onboarding process for AI developers, particularly those with limited experience in system administration. The TUI aims to make the Platform more accessible and user-friendly.
2. Emphasized Technical Documentation improvement and Platform AI Assistant development:
   Recognizing the importance of clear and comprehensive documentation, we are working on updating the full package of Platform codebase, documentation, and guidelines to provide users with the best user experience. We started by upgrading Publisher Portal guidelines and enhancing the Developer Portal’s technical documentation. Our approach involves hands-on collaboration with AI Developers during onboarding to identify and address common pain points effectively. Improved and expanded documentation will also be used as a dynamic knowledge base for the interactive AI assistant on the platform, which is currently in an active development phase, which in our opinion will bring a disruptive user experience of interaction with the Platform.
3. Progressed with Publisher Portal & Onboarding Process Improvements:
   We continue to refine and improve the Publisher Portal for the convenience of service providers and onboarders. At the moment, interaction with the test network has been upgraded, which provides important functionality for testing services during the upload process to the platform. Thus, the testnet was replaced, and the logic of interaction with the daemon at the test network level was debugged. An active phase of eliminating errors in the front-end part of Publisher and testing has also begun in order to eliminate cases of incorrect behavior of the portal. Some difficulties in interacting with the portal have also been eliminated for a better user experience, documentation and guidelines are being revised, taking into account support experience and feedback from service developers.
4. Reconsidered Service Development and Deployment Automation:
   We are implementing service wrapping automation and event-driven solutions for easy AI model deployment on the SingularityNET Platform. This means we can significantly simplify the service wrapping process, reduce costs, and scale AI models more efficiently. Previously, service providers had to maintain their own or third-party cloud servers to run AI services. With the use of event-driven service architecture, instances will only be used when accessed by users. We can now scale GPU models from zero to the limits set by the provider and back again. Onboarders will now be able to diversify their model hosting and communication practices to not only host AI models with serverless GPU vendors but also leverage modern practices for operator-based models in Kubernetes and other AI cluster management tools. At the moment, the solution has been implemented to ten AI services on the Platform for testing, and a detailed guideline for reimplementing this approach is being prepared for release, which we believe will dramatically reduce the time and effort for the full cycle of implementing AI models starting from wrapping the service and deploying the model, to onboarding on the platform.
5. Intensified Technical Collaborations, Integrations and Partnerships:
   One of the most important initiatives is the development of the SingularityNET Partnerchain for Cardano. This initiative regards a Substrate layer integration, which is crucial for enhancing the blockchain capabilities of the Platform, ensuring greater security, decentralization, and scalability. We have also started a deep technical collaboration with ICP (Internet Computer Protocol) aimed at fostering a robust partnership between SingularityNET and Dfinity, leveraging the strengths and capabilities of both entities to enhance the platform’s functionality, user experience, and overall ecosystem, aligning with our mission to democratize AI technology.
6. Unveiled Internet of Knowledge concept via efficient Knowledge Graph nodes and LLMs integration on the Platform:
   One of the existing most cutting-edge technologies in the field of AI with numerous applications is neural autoregressive self-attentional networks, also known as large language models (LLMs), which can still be profoundly improved or transformed into the next-generation AI technology via
   – Neural-symbolic hybridization
   – Knowledge-augmentation implemented as LLM conditioning on the knowledge graph content
   – Chaining of LLM inference steps controlled by or altered with symbolic knowledge-based reasoning.
7. In a decentralized paradigm, knowledge nodes and language models might be represented as isolated services, resulting in a diverse and distributed knowledge-driven AI network. In Q3, we explored different technical approaches and conducted experiments with different technical designs of knowledge and memory layers, integrating ontologies and graph databases, mixtures of vector-based and keyword-based retrieval, and LLM inference chaining frameworks.
8. The team is also developing a particular approach and implementing a framework to achieve such hybridization via integrating LLMs into a neural-symbolic cognitive architecture. This architecture includes:
   – Ametagraph-based knowledge storage
   – Our cognitive programming language (MeTTa),
   – Programs in which the content of the storage can represent declarative and procedural knowledge, queries to this knowledge, including programs themselves (implying full introspection), reasoning and learning rules and heuristics, as well as able to perform subsymbolic operations, in particular, by processing information with the use of neural network modules.

Progress toward AGI — OpenCog Hyperon meetup

In September we hosted a global technical OpenCog Hyperon/ AGI meetup in Dubai. This meetup was attended by our AGI team, partners, and external experts from around the world.

The meetup was a great success, with lively discussions and presentations on a wide range of topics related to Hyperon and AGI. We covered topics such as the core Hyperon cognitive architecture, the novel MeTTa AGI language, the integration of LLMs into Hyperon, rho calculus and hypervectors, and the potential applications of Hyperon in the real world.

One of the key takeaways from the meetup is that the field of AI is making rapid progress. Many exciting new ideas and breakthroughs were presented at the meetup, and it is clear that AGI is on the verge of becoming a reality.

Another highlight from the OpenCog meetup was the confirmation and announcement that Hyperon development is on track for Alpha release in March/ April 2024.

When we previewed the first Hyperon development roadmap in 2020, we laid out an ambitious timeline for progressing AGI R&D. In 2023, we updated our development roadmap to present an even more ambitious plan for scaling Hyperon. We are proud to say that we have met all of our commitments to date, and we have high confidence we will continue to do so.

We are committed to developing Hyperon in an open, responsible, and ethical way, and we look forward to working with external and open-source developers to ensure that human-level AGI systems are loving, compassionate, and helpful to humans.

Deep Funding

After completing DFR3 with 43 projects awarded (compared to 12 and 17 in previous rounds), it is clear that Deep Funding is growing fast, and we need to grow our organization along with it!

Organizational developments

We are at a crossroads where we can expand the organization in 2 ways:

1. Centralized, by growing our internal team with additional roles
2. Decentralized, ‘scaling out’ our operational tasks to the community.

While option 1 is the easier option, it will be hard to redirect our approach to option 2 once we have grown a larger internal team.

Therefore, Deep Funding is aiming to build up its internal organization as a distributed, community-driven organization from the start!

We are creating a number of circles in charge of specific responsibilities.
Luckily, we don’t have to start from scratch, and we can count on a small but fantastically engaged community to support us in this.

• Review Circle: We had already lined up a great team of community members to help us with eligibility reviews during DFR2 and DFR3. This team is now picking up speed by getting more tasks and responsibilities, including contracts, milestone reviews, and analysis tasks.
• Focus group: We also started right before DFR3 with the Deep Funding Focus group, representing the community and mediating between the community and the internal team. They are also picking up steam, setting up new (consent-based) consulting and decision processes, and organizing Bi-weekly listening sessions.
• Marketing Circle: We are in the process of creating a Marketing Circle that will be in charge of social media coverage, regular newsletters, X spaces, support of events, and more. A lot of these tasks are currently not done or not done well enough. We hope that the marketing circle will help us grow our program and the community even further.
• Awarded Teams representatives circle: This is still very early stage, but some awarded team members have already stepped forward with the ambition to help set up this circle and define their scope of work and responsibilities. More about this later!

Deep Funding Platform Development

During the last week of November, we deployed our new custom platform on Deepfunding.ai. While some changes are already visible (such as the renewed awarded team’s review pages and a new login feature), the most significant change will only be visible when we start a new round: the proposal creation flow. Our team put a lot of effort into this, so we can reduce the percentage of proposals that are not eligible to close to zero and have a better experience for both the proposers and the community that is reviewing and assessing the proposals.

There is MUCH more to tell about Deep Funding, but since this is a SingularityNET update and not a Deep Funding update, we will keep it at this. If you like to know more, please visit DeepFunding.ai, especially our blog section.

Ecosystem Updates (Incubation-Phase Pods)

Mindplex

In the third quarter of the year, the team made significant strides in expanding Mindplex Magazine and refining its user experience. The ‘Mindbytes Mirror’ news section has been introduced, providing our audience with the latest news in the realms of AI and crypto.

The successful launch of Mindplex Social’s customized Mastodon instance, with unique features like a blockchain-based reputation system and token-driven boosting, demonstrated the Project’s commitment to decentralization.

The Mindplex community continues to grow, with a notable increase in both magazine accounts, podcast followers, and website traffic.

Mindplex Mobile apps for Android and iOS are in the final stages of development. Mindplex Magazine was officially introduced as a blockchain-based media platform at the Rare Evo event in Denver. The event showcased the potential for a soulbound reputation token like MPXR to revolutionize prevailing and often biased content distribution systems by introducing dependable trust.

Additionally, this quarter saw the successful launch of the community podcast during this quarter, releasing a total of four episodes. Notably, one of these episodes featured Charles Hoskinson, the founder of Cardano, providing the audience with insightful discussions and valuable perspectives from a prominent figure in the blockchain space.

Jam Galaxy

Jam Galaxy has made significant progress in the third quarter, focusing on enhancing real-time music collaboration technology. After a thorough comparative analysis and hands-on testing, a superior low-latency technology has been selected to ensure seamless jamming experiences.

The development team has expanded, bringing new expertise to expedite the creation of the desktop app. Substantial progress has been made on the user interface, with detailed mockups for both the web and desktop platforms nearing completion. A collaboration with a creative agency has resulted in a fresh logo and comprehensive branding guidelines, aligning the visual identity and innovative vision that are yet to be integrated in Q4 2023. Production documentation is underway, paving the way for efficient development workflows. The integration of a new Song Splitter model is in progress, with an initial user interface already designed. A new marketing lead has joined to spearhead our promotional activities, and the team are actively exploring blockchain use cases to further empower artists and fans. These efforts mark a pivotal phase in Jam Galaxy’s mission to revolutionize the digital music landscape.

Musaic

In Q3, Musaic made decisive advancements toward MVP development — the AI Synthesizer. Data processing is nearing completion, with a system for meta-tagging in progress. A communication infrastructure between the servers has been selected and the AI Synthesizer. The team expanded with a main front-end developer and UI designer, leading to a finalized UI mockup. Work has begun on user account and credit systems. The team has also identified and tested a large language model to translate prompts into metadata, streamlining the sound generation process. These steps mark significant progress in our mission to revolutionize AI-powered music production.

Rejuve Biotech

The Rejuve Biotech team has secured a presentation slot at the Biotech Showcase during JP Morgan Healthcare week in San Francisco, the largest healthcare conference in the world. There is great investor interest in the project as well, highlighting the growing awareness of how AI and AGI will transform the longevity research space. Development on t