AI Services Roadmap
SingularityNET lets anyone create, share, and monetize AI services at scale. It's a full-stack AI solution powered by a decentralized protocol. While anyone can launch and monetize AI services using the SingularityNET platform, we have an internal team of 50+ researchers that will be seeding the network with an initial batch of services. The roadmap below outlines the development progress of these services.
The SingularityNET Network Analytics will comprise of statistical, neural net, information-theoretic, probabilistic, and logical AI tools to discover patterns and properties in complex networks.
We will integrate a number of standard analytical tools and the latest algorithms, packaging these into services that can be used with complex networks from any domain. We will analyze socioeconomic networks related to politics as a case study and proof of concept.
Social Media Analytics
For analyzing social media content, networks, and relationships, we will develop SingularityNET Social Media Analytics.
As an initial application, we will provide tools for PR and brand management teams. These solutions will allow them to understand social media chatter and trends in a more refined way, leveraging our AI technology to understand who the promoters and detractors are in social media, and how this has changed and is changing over time. Agencies and teams will be able to identify and target the most influential people and understand the competitive landscape better using our SingularityNET Social Media Analytics tools.
Biomedical Data Analysis
The SingularityNET Biomedical Data Analytics tools will be developed in partnership with Hong Kong-based Mozi AI Health.
A key focus is the creation of AI services carrying out genomic, proteomic, and clinical analysis using tools from the OpenCog framework, including the MOSES probabilistic program learning algorithm and the PLN probabilistic logic engine. These solutions also integrate a number of biomedical reference databases, and can therefore not only offer classification and diagnostic models, but also useful insights and explanations into the key genes and processes influencing AI decisions.
SingularityNET GHOST-Powered Chatbots allow users to create conversational Agents capable of verbal and non-verbal interactions. By using our custom developed programming language and authoring tools, users can build their own conversational agents like Sophia.
GHOST is the dialogue management tool used for Sophia conversation authoring, and it supports not only verbal inputs and outputs, but also video and auditory perception and animations as actions. We are polishing GHOST, making it generic and easy to use, and creating a set of authoring tools to support developers who want to create avatar-based chatbots, as well as assistant-type Agents.
Our SingularityNET Social Robotics tools and services are used to control Hanson Robotics' Sophia Robot. Developers will be able to integrate our technology into their own robotics projects in order to create social and emotional robots like Sophia.
In a collaboration between the SingularityNET Foundation and Hanson Robotics, the OpenCog AGI toolkit and the SingularityNET framework are being used to create the next level of intelligence for the world-famous Sophia robot, and also for toy-scale Hanson robots. SingularityNET will be used to provide the basis for a “robot mind cloud” enabling multiple robots to share knowledge and gain a common understanding of the world. Many of the AI tools and software services developed in the course of this work will be provided via SingularityNET for use by other robotics developers. This will transform SingularityNET into the most powerful platform available for social and emotional robotics, including applications such as home service robots, robot sales clerks, entertainers, medical assistant and elderly care robots, and much more.
DNN Model Evaluation & Tuning
The SingularityNET DNN Model Evaluation and Tuning will give developers simple services for evaluating deep neural nets on standard image, audio, video, and natural language problems. We will also allow clients to generate their own models by tuning existing models with their own data.
DNN Model Training & Neuroevolution
By using our SingularityNET DNN Model Training and Neuroevolution, we will allow clients to train neural networks from scratch, using both standard NN techniques and evolutionary computing.
Most machine learning and AI service offerings are limited to evaluation or narrow tuning of existing networks, which limits their scope. We want to allow clients to train networks from scratch, a computationally expensive process that is key to applying deep neural nets to domains where no good pre-training starting model is available. We will integrate and refine the latest advances in evolutionary computing and apply them to automatically design and tune deep neural nets, leveraging the decades of collective experience SingularityNET's researchers have with evolutionary computing techniques.
Broad Third-Party AI Technology Integration
We will integrate a breadth of existing open source AI tools and frameworks with our platform.
To support the general and robust use of SingularityNET as an AI platform, the SingularityNET team is working to provide a variety of well known and powerful AI algorithms within SingularityNET, including evolutionary computing, natural language processing, reinforcement learning, theorem proving, constraint satisfaction, and logic systems, in addition to tools integrated for domain-specific purposes (such as social media and network analytics).