VERSES AI Q3 Corporate Update Company Announces Launch of Next Generation Intelligent Software Platform
We’d like to begin by thanking all of you who have invested in VERSES, championed and promoted our efforts and remain interested in our mission. We genuinely appreciate your support and commitment to funding small companies doing truly cutting-edge technology work.
This company update provides an overview of VERSES, our background, mission, team, traction and key milestones, and updates on upcoming product releases, major partnerships and go-to-market strategy. In addition, we offer a thorough landscape and competitive analysis of this early-stage but rapidly growing AI Industry and our place in it relative to the competition. Finally, we share our thoughts on the future of AI technology and the implications for businesses, markets, governments, consumers and society at large.
As you finish this report, we hope you will find yourself as thrilled and excited for the next phase of our company as we are. As we exit the cocoon of our early-stage development, spread our wings, and take flight into this new Intelligence Age, we hope you will join us for this next phase of our journey.
2023 has been a pivotal year for both AI and VERSES. We've accelerated our pace, advancing key long-term goals and sharpening our focus. We want to provide an overview of our achievements this year and what lies ahead. More crucially, we aim to clarify how to best understand VERSES as a company and how to gauge our progress, thus empowering you to make informed investment decisions.
Founded by tech veterans, each with more than three decades of experience in the Silicon Valley technology industry, VERSES is committed to mission-driven innovation. While technology has generally benefited society, a lack of foresight into its long-term effects has led to today's challenges. This understanding informs our mission: to leverage cutting-edge technology for the global good.
We've been innovating since the internet's early days. When we saw the limitations of Web 2.0—where a few major companies dominate—we started rethinking our approach. Web 3.0, or the "Spatial Web," will extend into the physical spaces around us. We can't afford to let a few entities control this new era of exponential convergence.
We have been anticipating this convergence for decades and have formulated and executed plans to prepare for what's next. The future we have always envisioned is one where machines, AI, and humans collaboratively reshape the world. That vision inspired us to create VERSES.
Our credibility is anchored in our well-documented expertise, knowledge, and foresight, as evidenced by publications like The Spatial Web book. Committed to scientific rigor, we strive to deliver actionable insights backed by exhaustive research. Looking ahead, we see a world transformed by superintelligence, where machines, AI, and humans coexist and collaborate harmoniously.
Our next steps focus on democratizing intelligence, making it universally accessible and applicable. We experienced the transitions from Web 1.0 to 2.0, and now we are actively participating in creating Web 3.0, the next generation of the web that will combine AI, VR, Internet of Things and Robotics. Our ultimate goal is to leverage this technological convergence for the greater good while conscientiously mitigating its risks. We are the future of computing.
What We’ve Been Up To
VERSES began with a very specific plan that included a set of ambitious goals that we hoped to accomplish over several years. As we prepared for the convergence of exponential technology and a new market to mature, we set out to accomplish the following:
1. Work with the top universities and scientists worldwide to develop the hard science to enable AI to be smarter and safer.
Leading Universities: First off, we’ve published more than 30 scientific papers in collaboration with more than 20 of the leading universities and research centers around the world on the complex science of advanced intelligent software. Including MIT, Oxford, Imperial College London, Max Planck Institute, McGill, Northeastern and the National Institute for Physiological Sciences, Okazaki, Japan, to name a few.
Result: Understanding cognition from first principles grounded in physics and neuroscience and securing a solid scientific foundation for our work. We have proven that our approach to intelligence is superior and that our breakthrough mathematical models (Free Energy Principle) can model and predict how real neurons self-organize (learn) in the brain.
2. Work with the top standards organizations in the world to develop the global standards that enable AI to be smarter and safer.
Global Standards Development Orgs: We launched the Spatial Web Foundation, partnered with the IEEE and brought in the CTO of the Open Geospatial Consortium (OGC) to lead our standardization efforts. Over a hundred global working group members are assisting the standards development of P2874 Spatial Web Standards. This new generation of socio-technical standards and protocols is being developed to scale at the speed of AI evolution. If adopted globally, these standards could enable us to steer AI systems, even those that exceed human-level intelligence. These standards lay the foundations for the efficient integration and adoption of AI technologies while minimizing the risk inherent in AI.
Spatial Web Protocol, Architecture and Governance and are on track for approval in 2024.
We met with IEEE leadership last month. They are currently advising the EU Commission on the AI Act and other data privacy-related legislation, and they asked us to advance our partnership to assist them in these global standardization efforts.
Result: Developing standards for global interoperability to drive mass adoption that also ensures guardrails for AI.
3. Work with top governments around the world to drive awareness and develop case studies that demonstrate how AI can be smarter and safer.
World Governments and Leaders: We set up offices in Europe so that we could contribute to the EU's digital transformation and AI legislation efforts. The European Commission selected a consortium led by VERSES to develop the technological infrastructure for developing AI Governance for Autonomous Drone activities. Flying Forward 2020 used the Spatial Web Standards that we have been developing. This led the EU Commission to select VERSES for dAIEDGE, a network of excellence for “distributed, trustworthy, efficient and scalable AI at the Edge, expanding the European AI lighthouse." The aim is to support and ensure the rapid development and market adoption of distributed Edge AI technologies, including hardware, software, frameworks and tools that are reusable, secure and trustworthy. Edge technology is expected to be used in a wide range of fields, such as the Internet of Things (IoT), intelligent transportation systems, robotics and healthcare.
Results: This gave us direct influence, partnerships with governments, and the ability to accomplish real-world standards tests, demonstrating interoperability and guardrails for AI.
4. Work with the top research and consulting firms worldwide to drive awareness in Fortune 500’s that AI can be smarter and safer.
Global Research, Legal and Business Consulting Firms: We caught the attention of Deloitte and they wrote the report “The Spatial Web: What Business Leaders Need to Know about the Next Era of Computing.” Deloitte’s Vice Chairman Jay Samit joined our board. Gartner Research added our category of AI called “First Principles AI” to their annual AI report, closely monitored by the Global 1000. In addition, we collaborated with Dentons, the world’s largest law firm, to release the landmark 80-page report “The Future of Global AI Governance.” which calls for governments to adopt the socio-technical standards to ensure the safe governance of AI.
Currently, this report is being promoted in the US Senate, Congress and AI Caucus, and the Georgia State Legislature. And we are presenting it with Dentons at the Autonomous Vehicle Technology conference this week. In addition, we will be doing a series of webinars with Dentons in Q4 for their clients and worldwide to share our vision of a future of AI Governance secured by Spatial Web standards.
Results: We commanded the attention of the biggest influencers of software purchases for the Fortune 500. We have driven awareness and demand for our commercial solutions, which has led to a huge halo effect on corporations and policymakers alike.
We are considered the leaders by the leaders. Deloitte, IEEE, Dentons, Gartner and others are promoting our approach to the world. With the benefit of foresight and having anticipated this moment, we are now taking full advantage of a landscape in need of our solutions.
5. Work with the top companies in the world to test the market demand for AI that is smarter and safer.
Fortune 500 Corporations: As a result of the book The Spatial Web achieving #1 International Bestseller status on Amazon (the best marketing for our work), we started receiving inbound meeting requests from some of the world's largest corporations. Many of them had read the Deloitte report that described the Spatial Web as the future of computing and the next generation of an AI-powered, IoT-connected, immersively experienced Web 3.0. And they were all essentially asking the same question: What are you building over there and when can we test it? This led to a half dozen pilots with some of the largest companies in the world.
We are happy to announce that we have successfully completed these pilots and are now in the process of selecting and converting these pilot customers to Beta Partners for the launch of our main platform next month.
Additionally, we announced a reseller contract with the world's largest warehouse management company, which many Fortune 500 companies rely on. Just last week, we also announced a contract with a top 10 Fortune Pharmacy Retailer in the US.
Result: This confirmed that our product vision was relevant to the most significant problems the biggest companies in the world are trying to solve. It gave us sandboxes to learn and test our products and was integral to improving our product, targeting the right (more significant) problem set, and maturing our product market fit.
6. Work with and hire the smartest, diverse and experienced people across the world to develop and sell AI that is smarter and safer.
Hire Top Talent: We have grown the Company over the last year to 80+ incredible people. Our research team includes twenty-five PhDs and we’ve hired top engineering and sales talent from companies like Microsoft, Nvidia, Amazon, IBM, Honeywell, Unity, Raytheon, DOD, Accenture and more. For a small team, we have a strategic diversity of deep experience across critical fields that are extremely relevant to our product development across Neuroscience, AI, Robotics, Enterprise Software, Data Architecture, 3D Simulation and more. We are working hard toward our goal of building the leadership core that will enable VERSES to scale.
Result: We raised $23M in June to fund ongoing research and product development. Our fantastic team accelerated the maturity of our product by several years over the last 10 months, moving it from a primarily manual to a predominantly automated pipeline for ingesting and transforming data into AI agents. We have released several milestone papers published in the most prestigious science journals. We developed the research and secured our findings by filing critical patents to ensure that our technological breakthroughs around intelligent databases and intelligent agent-generating software are protected. These patent filings cover two key areas: one for automating the creation of intelligent software agents that can work with various systems like robots, drones, sensors and actuators, and another for predictive searching in databases to provide users with insightful search results based on comparisons and relationships. Furthermore, our science has now been proven. Soon, we will share the benchmarks that prove that we are producing superior results (more on this shortly) than any other AI company, with a fraction of the people at a fraction of the cost relative to the competition.
As stated above. This was by design. This was the plan.
Let’s review. What did we set out to do, and what have we done?
We have proven complex science, developed international standards for AI, influenced global governments, drove market awareness, gained credibility, built a groundbreaking product, and tested and learned from direct deployments with Fortune 500 customers and Governments.
Every one of the above accomplishments is fundamentally value-generating. But there is a more significant outcome we have been working towards. A singular strategic goal for the company. That, after many years of careful planning and preparation, has finally been realized this year.
Be at the right place at the right time and be ready with the solution that everyone will want and need.
That was the plan. And I believe we have successfully achieved that. We’ve positioned ourselves beautifully. And we were able to do that because we’ve been in the industry for decades. We knew what was coming. We predicted it. We wrote a book about it and told everyone we could to read it. Look at what’s coming! We prepared and built the product, the demand, the awareness, the influence, and the company—all to be ready for this moment.
We believe we are the best-positioned organization to become the market leader in the race to Artificial General Intelligence (AGI), which is often referred to as the 'holy grail' of AI. Our conviction stems from a combination of crucial factors: our scientific expertise, a highly skilled and visionary team, verifiable proofs, robust partnerships, credibility and alliances. We've also secured strong intellectual property rights, giving us a unique edge in innovation. Our initial market traction indicates a promising path ahead, setting us apart in a landscape filled with copycats.
VERSES is not just participating in the AI revolution; we're on track to disrupt the most disruptive technology in history. The next 12 months promise to be transformational for our company and the broader AI ecosystem. Our multi-pronged strategy is poised to accelerate our journey toward AGI, bringing along transformative changes that will impact industries, societies, and individual lives. Stay tuned as we unveil groundbreaking advancements in the pursuit of generalizable, ethical, and highly efficient AI solutions. We planned for this moment and want to walk you through what comes next.
The Intelligence Age
In early 2022, most people weren’t paying much attention to AI. But that changed dramatically when OpenAI's chatbot, ChatGPT, burst onto the scene, amassing 100 million users in just two months. The arrival of the intelligent chatbot ushered in a new generation of AI called “Generative AI,” named for its ability to generate content from simple text-based prompts. Generative AI triggered a global shockwave that rippled across the technological, commercial, industrial, and regulatory landscape. All at once, the AI story seemed to be everywhere.
As we stand in the wake of that initial AI wave, it becomes increasingly clear that we have witnessed the end of the Information Age and are now facing the dawn of a new age: The Intelligence Age.
In 2023, AI emerged as the premier investment opportunity in the financial markets, with Venture Capital funds funneling $40B into AI startups this year alone. Eyeing potential returns exceeding 1000x, VCs will remain patient and strategically well-positioned as the technology and its market evolve.
The anticipated market size, speed and economic impact of AI is unprecedented.
Statista forecasts a $2 Trillion AI software market by 2030. Meanwhile, McKinsey's revised estimates suggest an unparalleled $25 trillion surge in global GDP, attributed directly to AI's capacity for analysis and automation, which could enhance productivity across all market sectors and elevate the global economic value by a remarkable 30%.
Nothing like this has ever happened in the history of technology. The opportunity is massive.
Unfortunately, it's not accessible to everyone.. Although retail investors have been eager to grab their slice of the AI pie, many lack a robust understanding of emerging technology and can’t assess potential winners and losers at such an immature market stage. And even when they can, their options are extremely limited.
For example, all of the next-gen AI startups are privately funded, restricting investment opportunities to a select group of VCs and strategic tech funds and private investors.
This effectively sidelines retail investors, cutting them out of the early stage of the AI boom. They are left to take positions around AI infrastructure like Nvidia or Big Tech companies incorporating foundational Generative AI models into their current offerings, like Microsoft, Google, and Meta. The only other option is to go after vertical players trying to redefine themselves, like Palantir and C3.ai.
None of the new-generation foundational AI companies are in the public markets. Except one. VERSES.
AI promises to revolutionize the world to usher in an era of abundance for all. But many have begun to ask, should we believe these promises?
After ten months of frothy expectation, tales of AI’s adverse effects have begun to cast a shadow over its techno-utopian promises. The benefits of Generative AI have been clear—but so have its glaring and potentially harmful drawbacks. The markets, the customers, and the regulators are now all keenly aware of the limitations and problems with Generative AI.
The Problem: “Artificial” Intelligence
Scale is Not All You Need
Gen AI’s approach to AI development has been pretty straightforward: "bigger is better." Models like GPT-4, with its trillion parameters, illustrate this trend. Their mantra, "Scale is All You Need," advocates for bigger neural networks requiring more data and faster processing. However, this approach leads to costly and opaque "black boxes" that lack explainability and adaptability. These models consume massive amounts of energy and require significant expertise and investment for training and maintenance.
The financial cost is staggering, often reaching hundreds of millions to develop and operate such models. They have now attracted international regulatory pressure due to issues regarding bias and misinformation, privacy and safety. Furthermore, companies like OpenAI are facing class action lawsuits over copyright concerns. In summary, the existing approach prioritizes scale over transparency, efficiency, and adaptability, raising questions about its long-term sustainability and ethical implications.
Despite their sophistication, these AI systems are unable to explain their decision-making. They can’t adapt in real-time, limiting their utility in many real-world applications. And they tend to be one-dimensional, only able to model words, incapable of truly understanding or articulating the rich complexity of the physical world—our world. In essence, these models are the antithesis of the efficiencies and adaptability found in how intelligence occurs in nature.
Indeed, Gen AI's focus on scaling has achieved remarkable results, which explains why it's the favored approach among leading AI startups and tech giants. However, this same focus is also the root of widespread, industry-wide challenges. This "artificial" method of developing intelligence impresses with its ability to mimic human thought, but its scale exacerbates its downsides. In other words, the larger and more powerful these systems become, the more their problems like inefficiency, lack of transparency, and ethical concerns are magnified. So, while scaling may deliver advancements, it amplifies the systems' inherent flaws.
Despite knowing the risks, the development of these large models persists. Why? Because all of these companies are engaged in a race to build something much, much bigger.
Today’s state-of-the-art AI, referred to as artificial “narrow” intelligence (ANI), is essentially a single-purpose tool. But the mission of OpenAI and others is to scale up their huge models in the hope that artificial general intelligence (AGI) will eventually just emerge.
Think of AGI as the Swiss army knife of intelligence. Unlike today’s narrow AI, AGI is multi-purpose. It can do all kinds of different tasks. And that’s because it can learn and adapt and possesses the flexibility and versatility to address a diverse range of real-world challenges with human-level intelligence.
The arrival of AGI would mark the transition from specialized AI tools to more universal and adaptable AI systems capable of human-level intelligence without human-like limitations or even human involvement.
The first company to build AGI is expected to beat all others; hence the frantic race to the top playing out before our eyes. Because of AGI's expected capacity for rapid learning and autonomous self-improvement, once AGI comes online, it could continually and rapidly improve its own capabilities at a speed and scale no other technology has ever achieved. Leaving the competition in the dust.
This sudden "acceleration" in capabilities suggests that catching up to the company that first achieves AGI might be impossible.
The hope and fear that the artificial path will soon lead to AGI has forced world powers into an AI arms race. Every major tech company and country is jockeying to be the first to achieve AGI. Think of it as the new nuclear power. And with billions of dollars in funding and increasing demand, its arrival appears more imminent by the day.
Just last year, most experts forecasted the arrival of AGI decades or even centuries in the future, if ever. Today many believe AGI is likely to occur in a matter of years. Given the exponential rate of AI evolution, all bets are off.
However, it is imperative to recognize that AGI is the goal; Generative AI is NOT AI’s final form.
Today, AI companies are racing to scale up to AGI, at which point they expect AGI to continue somehow to scale itself up to its final stage, known as superintelligence.
For those taking the artificial “scale up” approach, the path to Artificial Super Intelligence (ASI) is straightforward. It simply assumes that we shift from humans scaling AI to AI doing the scaling by itself. No one knows how fast this accelerated evolution could occur. Perhaps years, perhaps minutes. But the implications are staggering, the risks dire.
This is the competition’s narrative. In fact, many of the top executives in AI companies and the VCs backing them refer to themselves as “effective accelerationists.” They argue that this “scale-up” path will lead to a far more prosperous future and that we should actively seek to speed this process up. That is part of the motivation for the billions in funding.
But the arrival of ASI marks the moment AI surpasses all the collective intelligence of the human race, achieving superior skills across all domains. ASI could potentially revolutionize every field, achieving superhuman proficiency in all areas and thus assisting humans in solving our hardest global challenges and empowering us to set sights on new horizons.
Entrusting a monolithic ASI with the potential to exponentially accelerate its intelligence seems like a recipe for disaster. Because whether a company or a country controls it (or tries and fails to control it), the technology and its approach pose grave and unprecedented risks.
Picture a scenario where a single company or nation gains unchecked control over an ASI that can exponentially accelerate its intelligence at an incomprehensible pace. The catastrophic implications are not limited to bias in the media and potential job displacement; they extend to unprecedented global power imbalances, the erosion of privacy, and the potential for unchecked manipulation and control or even existential fallout.
It's a dystopian future where humanity's fate rests in the hands of a single corporation or country or in the hands of the model itself. And if it is “itself,” will this superintelligent machine be benevolent, indifferent or malicious? We won't know and can't know until it's too late.
This is why world governments have called the leading AI companies into closed-door sessions and find themselves racing to issue legislation around the development and deployment of future AI systems. They are worried, and with good reason, that the unchecked pursuit of superintelligence and the artificial path Big Tech is taking to achieve it, could lead to disaster. This should indeed motivate us to consider the wisdom of the artificial approach to superintelligence, the implications that may arise from it, and our responsibility to make sure we are on the right path.
We believe this path not only raises serious ethical, safety, and governance questions but existential ones as well. As such, we believe that this is a path we must avert. We need a whole new approach.
The Natural Path to Superintelligence
Modeling Technology after Nature’s Blueprint
In the natural world, everything from individual cells to vast ecosystems operates harmoniously and interconnectedly, promoting balance and efficiency. If we want to develop genuinely efficient, adaptable, and holistic AI systems, we should model them after the most elegant and successful system in the known universe—nature.
In contrast to the “artificial” approaches to intelligence, VERSES technology is rooted in what is known as Natural Computing.
This approach doesn't merely mimic nature; it seeks to understand and implement its fundamental processes. Natural Computing explores the computational mechanisms found in nature to design new algorithms, models, and architectures. We believe that a “natural” approach to intelligent software is more likely to yield adaptive, efficient, and holistic systems – much like the ecosystems found in nature.
Rather than asking, “How can we scale AI?” we asked, “How does intelligence occur in nature?”
The Structure and Process of Cognition
For example, a cornerstone of our approach is biosemiotics, a field focused on the language of life through which all living things communicate via signs and signals. In 2017, we initiated research to translate this natural complexity into computational systems, laying the groundwork for a Hyperspatial Modeling Language (HSML). This universal language aims to express the intricate, multi-dimensional fabric of life, transcending the limitations of mere 'word models’ to enable rich, digital world models. These are similar to the mental models of the world that we all hold in our heads. These models are what intelligent software can operate on, just as we operate on our models, a.k .a. understanding of the world. Most crucially, HSML enables AIs to understand the world the way that we do.
We collaborated with the IEEE to establish HSML as the de facto global standard for AI.
These standards enable interoperability, explainability, and ethical governance of artificial intelligence and machines. HSML gives us the building blocks or “structure” of the world in digital format. A lingua franca for AI. Now all we need is a kind of AI that can use HSML to learn, adapt and operate safely in our world.
Karl Friston, renowned neuroscientist and VERSES chief scientist, has developed a theory over many decades that seeks to understand exactly how the cognitive process in humans works. Over decades of groundbreaking research in neurology and physics, Friston developed a proven framework known as "Active Inference," which states that the brain is a predictive organ that continually refines its models of the world based on incoming data.
Essentially, Active Inference is a scientific framework for modeling cognitive functions and sentient behavior in living systems. The framework has demonstrated its applicability in explaining and modeling a wide range of phenomena, from the activities of bacteria to the evolution of the brain. At its core, Active Inference explains how biological intelligence systems seek to minimize “free energy,” the discrepancy between expected and actual sensory information. Active Inference utilizes a statistical method to model how the brain performs perception, planning, and action by predicting how likely some outcome or event is to happen.
This widely accepted theory has already been applied across various disciplines, including neuroscience, biology, psychology, philosophy of mind, the social sciences, and robotics. At VERSES, it has found applicability in the realm of AI.
The revolutionary combination of the HSML and Active Inference provides us with the modeling language and mathematical blueprint for constructing mental models that intelligent agents can use for planning and taking action.
The Solution: The Natural Path to Superintelligence
This is a better path to superintelligence. It is not the “artificial” approach of Gen AI but the natural path to collective superintelligence derived from biological intelligence.
The natural path follows the blueprint of biological processes. As a result, our approach enables general intelligent agents (GIAs) with generalizability built in. They are more adaptive, resilient and efficient. They are designed to be able to run on small devices and communicate and collaborate at scale. Their “thinking process” is transparent, and they can explain their beliefs and recommendations, which we outlined in this groundbreaking paper on explainability in AI. Mirroring ecosystems in nature, there is no single point of failure or control. Because they run on the same algorithms as living systems, they also strive for dynamic equilibrium with their environment (including ours). This is because we followed nature’s blueprint for intelligence, its principles which state. Intelligence is not monolithic. In fact, as Michael Levin, one of the world’s leading developmental biologists at Tufts University and a VERSES collaborator, points out, intelligence is collective.
Like all organisms, humans are made up of cells that once existed as individual entities with their own competencies. These cells, over time, cooperated to form complex beings. Levin describes this as a "Multi-scale competency architecture," where each layer of biological structure, from cells to tissues to organs, has its own problem-solving abilities. Evolution has designed this layered architecture so that each level influences the behavior of the levels above and below it. The artificial path to superintelligence defies nature’s order: it tries to scale a single model to unnatural size and power—the first principles or laws governing natural intelligence. Like all humans, you began this life as a single cell. But today, you are probably about 30 trillion cells. You are an ecosystem.
Ecosystems of Intelligence
Our ultimate vision is an intelligent ecosystem—a network of AI agents that not only understand the intricacies of the natural world but are also interconnected. These agents share knowledge and collaboratively make decisions, striving for the same balance and efficiency that we observe in nature. Instead of isolated AI models, it's a network of AI agents that understand the intricacies of the natural world, moving beyond mere 'word models'. These agents can be interconnected, sharing knowledge and collaborating with human intelligence and machines, offering a level of interoperability and transparency currently absent in AI systems. And central to their design is the quest for balance and efficiency, like organisms in nature. The natural path to superintelligence is the way nature has designed it. And now we can prove it.
When we learn to tell different voices, faces or smells apart, the neurons in our brains change how strongly they connect to form pathways that model or encode the incoming data. This change helps our brain get better at recognizing things. In other words, changing the strength of the connections between neurons is the basis of all learning in the brain.
The Biological Proof: Inductive Inference
On August 7th, we published a paper in Nature Communications magazine. This international collaboration between researchers at the RIKEN Center for Brain Science (CBS) in Japan, the University of Tokyo, and University College London (Karl Friston) successfully demonstrated that the self-organization of neurons as they “learn” follows the mathematical theory of Karl Friston’s “Free Energy Principle” as known as Active Inference.
The principle accurately predicted how real neural networks spontaneously reorganize to distinguish incoming information. Collaborator professor Isomura stated after the findings:
“Our results suggest that the free-energy principle is the self-organizing principle of biological neural networks. Although it will take some time . . . generic mechanisms for acquiring the predictive models can also be used to create next-generation artificial intelligence that learns as real neural networks do.”
The Software Proof: Inductive Inference
In our forthcoming paper, scheduled for release later this year, we demonstrate how the Free Energy Principle and Active Inference can learn to play games like Pong in a matter of minutes on a personal computer, as well as solve the problem of the Tower of Hanoi in less than a second. This paper, led by Karl Friston, shows that Inductive Active Inference optimally solves complex tasks like maze navigation faster than traditional models. We intend to share a blog on this when the paper is released. But you are the first to hear about it, here today. This suggests unprecedented gains over the competition and highlights that by taking a first principles approach, following the blueprint of nature, it is possible to produce highly competitive solutions that could leapfrog the competition.
This remarkable capability for rapid learning and quick adaptation has far-reaching implications. In fact, it signals a seismic shift in our progress toward Artificial General Intelligence (AGI). With this level of efficiency, not only do we move closer to creating AI systems that can generalize learning across a broad range of tasks, but we also introduce a new paradigm of computational efficiency and adaptability that promises to revolutionize various sectors, from healthcare to autonomous systems.
By choosing the natural path, aligning with the inherent algorithms of nature and distributing power across a network, we can collectively thrive as a planet-scale superintelligence. Which we simply refer to as a Smart World. One where intelligence is distributed across every device and everywhere. Making our systems, processes, and even us smarter, together.
Imagine building upon the wisdom of the distributed, interconnected systems found in nature, harnessing their adaptability and resilience to create a more harmonious and equitable future powered by super (shared) intelligence. This is the better path. The natural path.
This is the Natural Computing approach. It's more than a technical paradigm; it's a philosophical commitment to developing AI systems that are rooted in the realities of our complex, interconnected world. By embracing the principles that govern natural systems, we aim to advance the field of AI into a more adaptive, efficient, and harmonious future. From this paradigm, we have designed our platform for distributed intelligence.
And we are happy to announce its arrival finally.
One Product, One Brand
We’ve been working this year to merge our underlying product capabilities into a single unified platform. What we’ve previously referred to as KOSM, the “OS,” and GIA, the Intelligent Agent, are being merged. It is the world’s first intelligent software platform. You can think of it as Intelligence-as-a-Service. A kind of smart software or “smartware” that can learn and ultimately update itself - software capable of powering and being embedded into all the “smart” things that companies want to build with AI.
But with all the smart cars, smart robots, smart homes, smart buildings and smart cities worldwide, the term “smart” has become a bit stale. So we wanted to find a name that captures the essence of intelligence and imagination that isn’t “artificial” but also goes beyond “smart.” One that conveys the unique approach we are taking and what the benefits will be. The intelligent software platform to power all these “smart” things. But can also enable them to self-improve. To make them smarter. We are excited to introduce you to.
Genius is a groundbreaking intelligent software system that will change how we use information and make decisions. Unlike traditional computer programs that follow rigid rules or current AI systems that lack memory and accuracy. Genius combines the power of natural algorithms, knowledge modeling, and data transformation to create a more flexible and intelligent system.
Think of Genius as a super-smart "brain" that can process and understand complex information. It can take in structured and unstructured data and convert it into knowledge in an automated way, similar to how our brains process and organize information. Genius can find connections between different pieces of information. It can spot hidden patterns, similarities, and connections, even when the data is uncertain or incomplete. This means Genius can make very accurate predictions and give us insights to make better choices.
Unlike AI today, Genius learns and gets better over time. Just like we learn from our experiences, Genius learns from new data. So, it becomes smarter and more helpful as time passes, making it great at solving problems and answering complex questions about any data set, surfacing new insights and recommendations. Plus, it can act like a business or personal assistant, offering recommendations and even predicting what we need based on history and real-time context.
With Genius, the possibilities are endless, ushering in a new era of smart computing, which expresses the genius of nature in software.
Generalizability: Genius is the first generalizable intelligence. It offers a versatile knowledge graph that interconnects diverse data types, allowing you to manage and search for information efficiently across any domain. As your knowledge base grows, Genius automatically contextualizes data, ensuring coherence and accuracy. It can learn and be customized over any data set.
Spatial Web Standards Integration: Genius seamlessly integrates with IEEE's Open Spatial Web Standard P2874, enabling you to model, transform, and ingest data from various sources. It provides the ability to describe physical and digital entities using a universal data standard.
Predictive Queries: With Genius, you can perform probabilistic queries and make insightful predictions using different data sources. This feature is particularly valuable for scenario analysis and determining optimal courses of action even with uncertain information.
Real-Time Adaptation: Stay informed with real-time data changes through subscriptions to agents and data sources. Genius empowers agents to react promptly to changes, enhancing your ability to respond to evolving situations.
Automated Compute Network: Design systems within Genius that automatically transform and compute data as it's received. This approach enhances efficiency by distributing computation across the network.
Collectively, these features enhance Genius's capacity for modeling knowledge, making predictions, adapting to dynamic circumstances and streamlining data management and analysis. This increases accuracy and versatility and improves efficiency, scalability, interoperability, composability and governance, going beyond the capabilities of today's AI models.
However, Genius can also work with the current AI systems, helping to minimize their issues by providing reliable and accurate memory, superior reasoning, security and steering. Complimenting their natural language abilities with a natural intelligent software infrastructure that enables AI to be smarter and safer.
Although we are focused on Developers for our first release of Genius, many of you have expressed a desire to experience it directly. So, we plan to launch a live “Baby Genius” demo on our website in November so you can see the benefits firsthand. We will refresh the website with the Genius brand and information over the coming weeks, with a complete facelift in January.
Why Genius is the Future: In a world that yearns for true AGI, Genius stands apart. It's not just AI but an agent-based knowledge exchange system. With features like advanced adaptability, swarm intelligence, data privacy and security compliance and a user data exchange, Genius is set to redefine general intelligence and pave the way to a super-intelligent world.
Artificial Intelligence has long been a promise. With Genius, we're turning that promise into a tangible reality. Our commitment is to lead, innovate, and offer the world intelligent software solutions that are not just smart but truly 'genius'.
We hope you will join us in this unprecedented journey as we aspire for smarter and safer technology, society, and the future.
Genius features and their availability will be added to the website soon.. Genius beta testing is expected to begin in October with exclusive access to Beta Partners; we plan to share demos of the behind-the-scenes capabilities and more details of our roadmap and benchmarking tests against the competition over the coming weeks and months. We expect to roll out our first set of Genius private beta partners in Q4 and begin the public beta in Q2 of 2024. Our initial user focus will continue to lean towards B2B while we explore the ideal product features and user base for a B2C launch.
Converting Pilots to Partners
We expect to announce the first set of private Beta Partners in Q4, which we anticipate will include not only new developers but also several of the large enterprise and Fortune 500 pilot partners we have been working with across multiple industries so that we can demonstrate the broad applicability of Genius to the world. The private beta aims to field test the technology, develop successful case studies, learn from our partners, and make the essential enhancements needed to roll out and scale the product in 2024.
Pivoting Partners - Although some of the large companies that we’ve been doing pilots with have tested Wayfinder, our intelligent routing service in warehouse operation, many are even more eager to use Genius as the ultimate “Data Finder” that can assist them in surfacing insights from across their data silos across their entire organization. A MUCH bigger prize as this means our product can apply to their entire company’s data set and not just their warehouse operations. Our partnership with Blue Yonder is going very well, even though some mid-year executive turnover pushed the client sales traction from Q2 to Q4. Still, its new leadership has moved us up from one-by-one client sales (which is back on track) to a direct integration of our Wayfinder routing services directly into the core of the Blue Yonder WMS, targeted for availability to all of their customers in 2024.
As we continue our mission, we are looking to keep you informed and updated on our plans and progress, so we launched a monthly newsletter in August and will continue these going forward.
We are also planning a webinar to present the above to you in more detail, show you some demos and answer your questions. Please sign up here to stay up to date and receive your invite.
We want to take a moment to thank you, our investors, who have taken the time to invest in us and our mission. We wanted to provide you with this update so you would have the opportunity to share the enthusiasm and excitement that we feel for the work that we are doing, the products we are building, the vision that we have for our company and the good we hope it will bring to you and the world we want for ourselves and each other.