Karma bar, supernova button and supernova system

Abstract

A mobile application and related system infrastructure is provided with a Karma bar, a supernova button, an action fund and / or impact feature. A user’s Karma may increase in line with the kindness wit

Description

OVERVIEW

[0001] Supernova™ is the Social Network for a Positive World.

[0002] Today we live a world where 99% of Gen-Z (especially those aged 16-24) use social media. Yet 1 in 2 of these tells us that doing so makes them feel ‘sad, anxious or depressed’ (see https: / / www.businesswire.eom / news / home / 20191024005627 / en / 48-of-Gen-Z-Say-Social-Media-Makes-Them-Feel-Anxious-Sad-or-Depressed-58-Are-%E2%80%9CSeeking-Relief%E2%80%9D-from-Social-Media).

[0003] The reasons are well documented and becoming better understood. Simply put, social media can be a toxic environment. When millions (and billions) of people are together in a toxic environment, the problems spread quickly.

[0004] There is of course obvious, outright antagonism. Name calling and ‘hate’ towards one another. Then there are more insidious forms and characters that the toxicity assumes.

[0005] One example is called ‘negative self-comparison’ or ‘NSC’. This is where one individual compares themselves unfavourably to another and it’s to an extent endemic in society as a whole.

[0006] But it’s made far worse on social media where there is no context, where the comparison maybe being made against a false depiction of perfection, and where there is the potential to see one NSC after another for hours on end.

[0007] Supernova’s unique technology and product design uses R&D to help address this and is the subject of our Patent application.

[0008] The initial results are very positive. From 15+ Million content impressions served so far, Supernova™ users have experienced effectively 0% toxicity. There has not been a single complaint. This is due to the overall technology and product design called the Supernova™ system.

[0009] This is a significant scientific advance that has the potential to change the world for the better, improving the mental health and self-esteem of the world’s 5 billion Social Media users. It’s potential contribution to the future mental health safeguarding of millions of users is world changing on a number of levels.

[0010] The advance benefits the entire technological community. While Supernova™ is looking to build on this remarkable advancement with new and further original technology, for some of which Patents will be applied (for example the ‘Karma Bar’ which is part of the unique experience) this technology- being in the public domain- will potentially form the basis of future social media evolution.

[0011] Supernova™ furthermore helps share the $230 Billion in annual Social Media advertising with charities and good causes.

[0012] Its technical innovation helps address the serious threat to Mental Health posed by other social media. This is documented in research and exposed in cases such as the Frances Haugen ‘whistle-blower’ revelation of 2021 that, for example, 40% of teenagers who feel ‘unattractive’ say these feelings began on Instagram, a service currently used by 34.7 million people in the UK.

[0013] Innovatively, on Supernovas ‘Likes’ earn money for charities. ‘Likes’ are really votes for the cause you support to get a share of the Action Fund that Supernova™ pay into from advertising revenue. The more likes your chosen cause receives, the larger percentage of the Action Fund the partner charities within it will get.

[0014] When a post on Supernova™ makes a user’s day, the Supernova™ button sends lOx a normal ‘Like’ to the chosen cause of the poster. However, the user must first earn ‘Karma Points’ to have the right to give a Supernova™. This is done through using the App in a kind, inclusive way, and not being antagonistic or ‘toxic’ towards others through the users’ behaviour.

[0015] This user behaviour is moderated by real people. Supernova™ is developing its own Natural Language Processing (NLP) Artificial Intelligence (AI) technology which is learning from these so the same level of service can be provided at scale whilst retaining as much human insight as possible.

[0016] Once the user has earned sufficient Karma Points™, the Karma Bar™ at the top of the App fills and the Supernova™ symbol displays on every post. A Supernova™ can then be deployed, and the user will be rewarded with an animated display, showing it is even better to Give than to Receive a Supernova™. Through unique techniques like this, they are creating conditions for their community that encourages compassion and kindness and breeds a culture of More Love, Less Hate in line with their Charter (https: / / go-Supemova.com / support / charter.html).

[0017] Supernova™ is the Social Media site for people who want to make a difference and simply by changing where they post, people begin to change the World.

[0018] Supernova™ is a UK based start-up with global ambitions to scale, building on an MVP product that released in Beta end of2021 with partners including ASICS™, British Red Cross™, UNICEF™, Save the Children™, Oxfam™ and RSPB™.

[0019] The Beta product evolved successfully throughout 2022 with the addition of Philips™ as partners and discussions are advanced to add a 3rd major global brand partner in 2023.

[0020] It was the most nominated App at the App Awards ’21, including for ‘Best App in the UK.’ Supernova™ was also named a top 25 UK Company Innovating the UK Charity Sector, within 10 weeks of launch. It has been described as the ‘Ethical Alternative to Facebook and Instagram’ and is the number one search result of that phrase among 100 Million + organic results globally on Google™. Throughout 2022 it was listed in the Top 20 Social Networks on Exploding Topics (usually at #6). It was named 8th in the World List of Definitive Social Networks in Jun 2023, above Facebook™, Twitter™, Instagram™ and Pinterest™.

[0021] Its world class founding team is comprised of very successful professionals, investors and highly qualified, award-winning experts from the worlds of advertising, technology, media and entertainment (inc. BAFTA™). BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figures 1 to 3 illustrate a Karma Bar™;

[0023] Figures 4 to 7 illustrate a Supernova™ Button;

[0024] Figures 8 and 9 illustrate an Action Fund; and

[0025] Figures 10 to 12 illustrate an Impact Feature. DESCRIPTION OF THE INVENTION

[0026] Figures 1 to 3 illustrate a Karma Bar™. Thru a complex series of ‘Nodes’ in tire Proprietary Back End Code, the Users ‘Kanna’ increases in line with the kindness with which they use the app until they see it reaches 100%. Then, they can give a Supernova™.

[0027] In Figure 4, user earns 100% Karma - sees a post they love. Decides to give a Supernova™. Hits button.

[0028] In Figure 5 and 6, user sees exhilarating animated display. Gives lOx votes for the users’ cause to receive more of the Action Fund. In this case, user supports Mental Health (see symbol next to handle). Feels good to do good.

[0029] In Figure 7, User sees exhilarating animated display. Gives lOx votes for the users’ cause to receive more of the Action Fund. In this case, user supports Mental Health (see symbol next to handle). Feels good to do good.

[0030] Figures 8 and 9 illustrate an Action Fund. The Action Fund records the amount Supernova™ has set aside from advertising income to donate to charities, in line with the % ‘Likes’ and Supernovas users’ posts have earned for their supported causal areas, where every Supernova™ is worth lOx regular ‘Likes’, which is another reason it’s so great to get and to give one. @artkoop supports Mental Health, as shown above.

[0031] Figures 10 to 12 illustrate an Impact Feature. The Impact feature clearly displays the tangible benefits that are happening as a result of the users’ use of Supernova™ and the money from advertisers (like ASICS™) their content is earning for their chosen causes via the unique Karma->Supemova™-> Support System. The total for each causal area is divided equally between its beneficiary charities. For Mental health, there are two. TECHNOLOGY

[0032] R&D centred on development of the unique BackEnd technical infrastructure for the overall systems architecture as well as on creating a user interface that was easy and simple to use, and enjoyable for people of all ages and backgrounds.

[0033] Their central stack, which is a service-oriented architecture built on Kubemetes, is as follows:

[0034] Nginx

[0035] Postgres,

[0036] Neo4j,

[0037] Neo4j Express,

[0038] Node.js, and

[0039] SocketlO.

[0040] The key issue that their R&D sought to solve was that the world’s addiction to social media is not doing it much good- and is in fact, in many ways, causing it a lot of harm.

[0041] It's a widespread, growing social issue: an immense societal problem-that has contributed to and continues to contribute to problems with mental health and self-esteem, particularly among the many younger and more vulnerable users of social media.

[0042] Instagram™ identified in their own research that ‘40% of teenagers who feel ‘unattractive’ say these feelings began on Instagram’ (source https: / / www .motleyrice .com / ne ws / instagram -mental-health-children#:~:text=Among%20teens%20who%20reported%20suicidal, began %20while%20using %20the%20product). But it's a problem that impacts us all.

[0043] Supernova™ aims to significantly and fundamentally change this by implementing unique new technology that rewards "karma" for kind and inclusive behaviour while penalising negative and toxic behaviour thru loss of karma. Users can see their Karma increase within the app, and once their Karma is complete, they can give a Supernova™. This sends more of the Action Fimd to the posters chosen cause. And users can see the tangible benefits of that donation clearly expressed within the app. Together with the exhilarating animation display when the user gives a Supernova™ this makes the giver and receiver feel good. This mood enhancement is the contrary user experience to the self-doubt and lack of self-worth reported as an experience on other social networks, and brings a major shift for the world of social media (nearing 5BN strong) which already benefits many users and will benefit millions more as Supernova™ further research, develop and scale their unique technology.

[0044] The technology also assists human moderators in maintaining a safe and positive environment for social media users, as well as for brands whose funding ensures a higher standard of service, and which, exclusively on Supernova™, is donated to charities and good causes.

[0045] It's a ven- complicated area of human behavioural science that presents broad societal and technical challenges. If these are overcome, millions of people will be able to use social media more safely while preserving their mental health.

[0046] All of this technology is referred to collectively internally as ‘the Supernova System’ and it is unique in global social media.

[0047] All that it achieves must be accomplished in a considerate, sophisticated way that enhances the user experience rather than in any way that comes off as "school teacher-like" or overtly "promotional" and might be perceived as patronising. Failure to accomplish this effectively could confinn users' preference for their current social media platfonn, making the task of improving it even more challenging.

[0048] Additionally, Supernova's R&D is looking for a way to start contributing some of the $230 billion in global social media advertising to charities and worthwhile causes.

[0049] There are no other networks using the technology they have developed, to their knowledge, and this is based on extensive research, including attendance at the UK App Awards among thousands of entries in the UK.

[0050] It’s further evidenced by, and the reason for, Supernova™ being named 8th in the Global ‘Definitive’ List of 250 Social Networks, above major industry names such as Facebook™, Instagram™, Twitter™ and Pinterest™. Supernova™ has also been recognised on countless Industry ‘Top Tables’ of new and innovative social networks for the same reasons of unique product and technology design and development through R&D.

[0051] Supernova is now continuing work in these areas in 2023, especially in the area of artificial intelligence to further ensure the safety of its users, based on Knowledge Graph technology. These areas saw significant progress in 2022.

[0052] Making and launching an MVP / Beta (www.go-supemova.com) to industry acclaim with a global sponsor on board and ready to scale (Global Sports Brand, ASICS™) in '22 was the major accomplishment.

[0053] This, however, turned out to be more difficult and time-consuming than anticipated. The numerous technical challenges to be overcome as well as COVID's continued disruption of progress into 2022. These were dealt with internally due to the team's expertise.

[0054] The key R&D focus will be on constantly raising the bar for Brand and User Safety and Inclusion, as well as on smart technology and AI to support Human Moderation without degrading the human eye. As a result, Supernova anticipates scaling with new Media Partners in place for Q4 of‘23.

[0055] Supernova is a distinctive social network in the expanding "tech for good" sector. It is a part of a "Purpose Led" movement in which brands prioritise their long-term relationships with their customers over their immediate financial interests.

[0056] Given that its brand name is an acronym of the Latin phrase "Sound Mind. Sound Body," ASICS™ is widely regarded as a leader in the field of mental health and collaborates closely with Mind.

[0057] Supernova™ will establish itself as the first Purpose Led social network in the world by using the established channels to immediately put its innovative technology to use and benefit society. DEVELOPMENT METHODOLOGY

[0058] Supernova Networks Ltd. is a small business. All staff have some involvement in ongoing research or development for new product features. Due to their small team size and agile methodology, Supernova Networks Ltd. is able to quickly prototype and iterate in response to user feedback and technological advancements.

[0059] Eligible R&D is considered to start when the company's team starts to evaluate the business case to technically implement a new feature or modification. A crucial step in the development process are discussions about whether or how to implement a new feature or modification and these lay the groundwork for the creation of the necessary technology. Once all technological uncertainties are resolved and clarified and the virtual testing procedure is finished, eligible R&D is complete. ELIGIBLE WORK Introduction

[0060] Our organisation is creating a unique social network that generates hubs of users posting positive content, a significant portion of the ad revenue from which is dedicated to charities.

[0061] To enhance user experience, content recommendation, and user-content classification, we are significantly ramping up our investment and efforts in the development and optimisation of Large Language Models (LLMs) and Graph Technology. Content Moderation and AI

[0062] We are leveraging the capabilities of LLMs to understand the intricacies of language, context, sentiment, and intent, in order to automate the moderation of user-generated content. These models are being trained on a rich dataset that is representative of the interactions expected on our social network. This automated moderation system is designed to maintain our community guidelines, by taking actions such as flagging, hiding, or removing inappropriate or non-positive content.

[0063] Furthermore, to enhance the accuracy of the automated moderation process, a feedback loop is being established to allow the AI models to leam from the moderation actions taken by human moderators. User and Content Classification

[0064] The combination of LLMs and graph technology is proving instrumental in user and content classification. Our AI models are being designed to analyse user-generated text to categorise users based on their interests, sentiment, and behavioural patterns. Graph technology, on the other hand, is used to identify clusters of users based on their interactions with other users or content on the platform.

[0065] In addition, by employing vector databases to store high-dimensional vectors that represent the 'semantic fingerprints' of users and content, we are significantly improving the relevance and accuracy of our content recommendation system. For new content, the LLM generates a vector which is used to query the vector database to find similar content or users, thereby enriching the user experience with personalized content. Content Recommendation

[0066] Graph technology is facilitating the building of robust recommendation systems to help users discover relevant content and like-minded individuals within die network. We are treating user profiles and content as nodes in a graph database, and mapping connections between users, their interactions, shared interests, and behavioural patterns. Advanced graph algorithms such as PageRank, shortest path, and community detection are being utilised to recommend personalised content.

[0067] To enhance the efficiency and scalability of our recommendation system, we are layering the vector database as a 'lookup' layer before querying the graph database. This narrows the scope of the graph query to a subset of nodes that are most relevant to the user or context. Content Classification and Optimisation of Tokenisation

[0068] Our LLMs are being trained to understand the context of a post and classify it under predefined categories. The sentiment of the content is gauged using advanced Natural Language Processing (NLP) techniques like sentiment analysis.

[0069] Optimising the tokenisation process is crucial to the efficiency of our AI models. To handle out-of-vocabulary words, we are implementing subword tokenisation techniques, such as Byte-Pair Encoding (BPE), which splits words into smaller units based on the frequency of occurrence. We are also managing longer sequences of text through methods like sliding window or truncation, balancing the sequence length and computational resources. User Support

[0070] A user support system powered by LLMs is being developed. This system is designed to handle user queries, provide guidance on navigating the social network, and assist with common issues. For more complex issues, the system is being trained to escalate the query to human support, ensuring that our users always receive the best possible assistance. AI Ethics and Explainability

[0071] We are maintaining a strong focus on the ethical implications of our AI implementations. An AI Ethics framework is being established to ensure responsible use. 1. Component 1: Community / User Safety la) Project Description

[0072] With the recent advancements in Large Language Models technology, we're developing a comprehensive system to enhance user safety on our social network. Our focus is on a deep learning system that identifies and curbs the proliferation of low sentiment content, supplemented by a user report mechanism for improved moderation. lb) Scientific or Technological Advancement

[0073] Leveraging recent strides in large language models technology, our project introduces unique deep learning algorithms to foster a more positive and respectful conversation environment on our platform. 1c) Baseline in Science or Technology

[0074] Despite the current advancements in large language models, many social media platforms struggle to effectively manage low sentiment content. Our project aims to address these challenges with innovative solutions.

[0075] The current baseline in everyday life would be that 50% of Gen-Z currently feel sad, anxious or depressed when they use social media (Source: Forbes). And that 40% of teenagers who feel unattractive say these feelings began on Instagram (Source: Instagram). Id) Scientific or Technological Uncertainties

[0076] We face uncertainties in developing an effective system that accurately moderates low sentiment content and handles user reports, especially in the context of the rapid advancements in large language models technology. Ie) Overcoming Uncertainties

[0077] Through continuous development and refinement of our deep learning system and user report mechanism, we're addressing these uncertainties, drawing on advancements in large language models technology. If) Boundaries of R&D

[0078] Our project is ongoing, with current efforts directed towards enhancing the deep learning system and user report mechanism based on feedback and performance data. The continuous advancements in large language models technology significantly impact our objectives for this year. 2, Component 2: User Experience Optimization 2a) Project Description

[0079] Leveraging the tremendous advances in large language models technology, we're developing a context-aware advertising system that specifically caters to the needs of charities. The system employs a custom reinforcement learning model to adjust to user interactions and preferences, ensuring the ads are in harmony with the mission-driven objectives of the charities benefiting from donations they receive as a consequence of the ads. 2b) Scientific or Technological Advancement

[0080] Our project, informed by the latest developments in large language models technology, represents a technological leap in online advertising on social media platforms. By using reinforcement learning to tailor the advertising system to the specific requirements of charities, we're creating a novel synergy between advertising and the pursuit of mission-based objectives. 2c) Baseline in Science or Technology

[0081] Despite the advancements in large language models technology, current ad systems on social media platforms often neglect to align the advertising content with the beneficiary's mission. Our project aims to overcome this by developing a context-aware advertising system that specifically caters to charities. 2d) Scientific or Technological Uncertainties

[0082] We face uncertainties in developing an advertising system that effectively aligns with the mission-driven objectives of charities while minimizing disruption to user experience, especially in light of the advancements in large language models technology. 2e) Overcoming Uncertainties

[0083] We're addressing uncertainties through the ongoing development and refinement of our context-aware advertising system, leveraging the advancements in large language models technology to better serve the needs of charities. 2f) Boundaries of R&D

[0084] Our project is ongoing, with current efforts focused on refining the reinforcement learning model based on user interactions and preferences. The tremendous advances in large language models technology significantly inform our objectives and strategies for this year, particularly our commitment to serving the specific needs of charities through our advertising system. 3, Component 3: Moderation Application Enhancement 3a) Project Description

[0085] In light of the advancements in AI and large language models technology, we're improving our moderation application. The updates are informed by the learnings and advancements made in Component 1, particularly the deep learning system that manages low sentiment content and the user report mechanism. 3b) Scientific or Technological Advancement

[0086] Our project expands on the advancements in large language models technology to bring about a significant improvement in the moderation capabilities of our application, ensuring a safer and more respectful conversation environment on our platform. 3c) Baseline in Science or Technology

[0087] The current baseline involves manual and semi-automated moderation systems that struggle to keep pace with the increasing volume of content. By leveraging advancements in AI and large language models technology, our project aims to substantially enhance the effectiveness and efficiency of content moderation. 3d) Scientific or Technological Uncertainties

[0088] Uncertainties lie in the integration of our deep learning system and user report mechanism into the moderation application, and the scaling of these technologies to handle the increasing volume of content, especially in the context of rapid advancements in large language models technology. 3e) Overcoming Uncertainties

[0089] We're addressing these uncertainties through the continuous development and refinement of our moderation application, leveraging the advancements in large language models technology and our learnings from Component 1. 3f) Boundaries of R&D

[0090] Our project is ongoing, with the current focus on enhancing the moderation capabilities of our application based on feedback and performance data. The continuous advancements in large language models technology significantly inform our objectives and strategies for this year, particularly in the enhancement of our moderation application. Uncertainties Faced with R&D work Methods used to overcome Developing an effective recommendation engine to suggest relevant content and connections to users. This involves understanding user behavior and preferences. Designed a novel hybrid recommendation system using collaborative filtering and deep learning. Customized activation functions in the deep learning model to better capture user behavior and preferences. Building a robust system for detecting and handling misinformation and fake profiles, to maintain the integrity of the social network. Built a bespoke Deep Learning system for image and text recognition, incorporating unique activation functions to improve accuracy. Further enhanced by integrating a user report mechanism for improved detection. Ensuring user data privacy while still providing a personalized experience. Tins includes complying with data regulations. Developed a novel approach combining homomorphic encryption and differential privacy to ensure user data privacy. This approach allows for personalized experiences using anonymized data, in compliance with data regulations. Scaling the social network to support increasing user base and interaction volume without compromising performance. Utilized Docker Swarm for container orchestration and employed distributed databases, specifically Apache Cassandra, to ensure scalability and high-performance operation of the social network. Creating a sentiment analysis system to gauge user reactions and feelings towards posts and updates, to better tailor the user experience. Engineered a custom sentiment analysis system using advanced Natural Language Processing techniques. The system uses modified Long Short-Tenn Memory (LSTM) networks for improved sentiment detection. Implementing an effective advertising system that doesn't disrupt user experience but still provides value to advertisers. Built a context-aware advertising system using a custom reinforcement learning model. The model adjusts to user interactions and preferences, minimizing disruption to the user experience. 5

[0091] Although the present invention has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the scope of the present invention is limited only by the accompanying claims. Additionally, although a feature may appear to be described in connection with particular embodiments, one skilled in the art would recognize that various features of the described embodiments may be combined in 10 accordance with the invention. In the claims, the term ‘comprising’ does not exclude the presence of other elements or steps.

[0092] Furthermore, the order of features in the claims does not imply any specific order in which the features must be performed and in particular the order of individual steps in a method claim does not imply that the steps must be performed in this order. Rather, the steps may be performed in any suitable order. In addition, singular references do not exclude a plurality. Thus, references 5 to ‘a’, ‘an’, ‘first’, ‘second’, etc. do not preclude a plurality. In the claims, the term ‘comprising’ or “including” does not exclude the presence of other elements.

Claims

1. A mobile application with a Karma Bar, a Supernova Button, an Action Fund and / or an Impact Feature as described in the specification and / or figures.

2. A system / infrastructure for providing an application as described in the specification and / or figures.

Images