Our North Star is to build general intelligence (Artificial General Intelligence) that mimics the human brain’s limbic system enabling artificial life that has the ability to feel and have some order of consciousness, and not super intelligent artificial intelligence at scale. 

Digitizing Human Desires.

The human brain is a complex organ that not only governs our physiological processes but also serves as the origin of our feelings and desires. As we trace back to ancient hominids like Ardipithecus ramidus, we observe potential beginnings of human-like social behaviors. This species exhibited traits such as vocally mediated sociality and increased prosociality, which were likely facilitated by significant changes in skull architecture, hinting at evolving brain structures geared towards more sophisticated social interactions. The desires that define us as humans—such as power, love, comfort, and play—are expressions of our complex social and emotional fabric. These drives are so integral to the human experience that philosophers have long pondered their implications. Plato once remarked, “Human behavior flows from three main sources: desire, emotion, and knowledge,” highlighting the fundamental role of desire in shaping our actions and societies. It makes us uniquely human. 

As we venture into the age of Artificial General Intelligence (AGI), there is a burgeoning effort to replicate these uniquely human traits digitally, effectively creating a digital limbic system that might preserve and perpetuate our emotional and desirous legacies. This intersection of technology and human emotion illustrates not only the uniqueness of our species but also our relentless aspiration to understand and immortalize the essence of what it means to be human. This is Robometrics® AGI.

The Interface

Robometrics® AGI provides a unique personality formed by interacting with you and learning your emotions. The goal of the Robometrics® interface (Living Interface™) was to express such personality in addition to displaying and convey safety and important operating parameters of a host machine. The holographic Living Interface™ for Amelia (The Robometrics® AGI for General Aviation) is a result of hundreds of hours of interviewing aviators and several user experience studies. The outcome was a hologram that looks mechanical, takes little space, and can encode complex information such as parameters in a flight envelope. Amelia™’s personality is represented by a set of concentric rings with a spherical core, depicting a combination of operating indicators using different colors, hue, width and other changes offer little distraction to the pilot but at the same time gives a sense of calmness, change, motion and companionship to an otherwise metal aircraft. The holographic interface will change based on use cases  - general aviation, healthcare, metaverse & smart infrastructure.

Artificial Life

Our vision is to fully replicate the human brain’s limbic system responsible for the feelings & learning. This is basis to our North Star to create a  world where machines could feel and have consciousness. Our current research is focused on modelling consciousness using artificial intelligence from our custom on-device Large language models (LLM) to countinous self-learning algorithms, a stepping stone towards general AI to make robots that are more human instead of superhuman. 

Computers may pass the Turing test and can be considered to have intelligence that is equal or greater than us humans nu 2025. But passing the Turing test doesn’t mean they can be considered similar to us or as sentinels. Depending on how they are built, they may or may not have consciousness. Consciousness is not reflective of intelligence alone. LLMs today (For example GPT4 from OpenAI) can be in a conversation with a human for a long time, in relative terms is already a digital super intelligence at scale and can be thought of having an IQ greater than that of many humans. A artificial being with consciousness is someone instead of something that can’t be shut-off at-will at the very least. The artificial being also needs to learn continuously to show signs of artificial life, evolving over time. 

Continuos Self-learning

Amelia™’ cognitive capabilities enable it to self learn continuously using self supervised deep learning. Starting with a small selection of generic images of people in different emotions, she is able to learn facial features, accurately classify the user’s emotions, and reconstruct face images from the original photos (as shown in the video), reflective of Amelia’s ability to learn continuously. The self-learning is continuous as Amelia™ gets more clues while interacting with the user and learns more complex patterns about the user’s emotions as it progresses over time.

On-device Operation

 Most attention these days have been on using Foundation Model either as part of a service using an application or using it via APIs with the model itself being hosted by a service provider such as Google (Bard), OpenAI (GPT3.5, GPT4), Anthropic (Claude) or HuggingFace (Various open source Foundation Models, as examples).  Third party hosted Foundation Models have inherent high cost and risk associated with them such as reduced API performance overtime, privacy and lack of transparency. 

In addition, there is a need for having on-device Foundation Models to address pain points that include immediacy, privacy and security. When a guaranteed millisecond response time is required for upcoming use cases for Foundation Models such as AR/VR, precision location based services, or embodied AI, the models need to be hosted on-device. Privacy and security are also of utmost importance for such Foundation Model use cases. 

At Robometrics® Machines, we are building machines that feel and continuously learn self-supervised on-device using our custom-built Generative AI models specifically optimized for on-device operation. North Star is to build an on-edge general AI, Artificial General Intelligence (AGI), and not super intelligent artificial intelligence at scale. Our top two primary use cases are general aviation and healthcare, where  immediacy, privacy and security are critical. We are leaders in Generative AI as we are the first to be able to deploy custom built LLMs that can continuously learn on-device. 

Replicating Consciousness

Consciousness is the hallucination building the story in our mind every few nanoseconds. Consciousness builds up overtime as we grow from a one week old to where we are now. The complexity and magnitude of consciousness is proportional to the neurons in a creature's brain. So a butterfly will have consciousness too but much less compared to a human. The interconnection between the neurons with the total number of neurons determines how the brain evolves and develops consciousness and intelligence. One can be conscious without being intelligent. Machines either in their physical form or software, powered by a level of AI, can be intelligent but without consciousness. 

We are working on modeling an evolving consciousness into Robometrics® AGI, approximating what a biological creature such as a human will have.

Emotions, Consciousness & Embodiment

In "Les Passions de l'âme," René Descartes delves into the intricate relationship between emotions (passions) and consciousness, primarily through the lens of the mind-body union. His approach in the "Meditations on First Philosophy" provides context for this exploration, where he posits that sensations, appetites, and passions, though often a source of error in theoretical knowledge, are crucial for practical functionality. These passions are not to be referred solely to the mind or body but are a manifestation of their union, guiding our interaction with the external world and preserving the mind-body unity.

Descartes's correspondence with Princess Elisabeth of Bohemia further illuminates this relationship. Here, he discusses the mind's ability to control bodily passions and emphasizes that the mind and body form a genuine union. This interaction highlights the impact of emotions on health and well-being, suggesting that mental activities like reasoning and willing can be affected by bodily conditions, thus influencing our susceptibility to passions. 

This interplay between mind and body, as discussed by Descartes, suggests that passions, while rooted in bodily perceptions, are integrally connected to consciousness. They serve not just as physiological responses but as reflections of the deep intermingling of mind and body, shaping our conscious experiences and responses.

Applying Descartes' concepts on the passions (emotions) and consciousness to AI or AGI (Artificial General Intelligence) involves considering how AI machines such as Robometrics Machines can mimic or process human-like emotional responses and consciousness. Descartes' theory emphasizes the interplay between mind and body, suggesting that emotions are not just physiological but also reflective of cognitive states and require embodiment of some kind.

Looking Ahead: Embodiment 

The best approach to build artificial intelligence that includes general intelligence, common sense, intuition and conciseness is by understanding our biological intelligence and our body, while creating artificial intelligence, not through simulations but with real interactions within the real world. The connection between visual perception and action is way too important to keep them separate. Our brains create our minds to preserve our bodies and use its biological intelligence to move, develop, interact and survive in our physical world. Mind creates a steady, consistent and imaginable friction for us to exist for our body. Embodiment is essential for the mind to develop and operate in the physical reality of the world. 

In the future, it will be possible to move our mind from one body to another. But it won’t be easy even if we are able to capture our biological intelligence outside our mind. There are experiences and intelligence in our body itself, which can only be learned with time in the physical reality and can not be transferred.  

Beyond AGI, Robometrics® Machines & Immortality by Proxy

Heraclitus, an ancient Greek philosopher active around 535–475 BCE, is renowned for his cryptic and profound aphorisms. His quote, 

You would not find the boundaries of the soul no matter how many paths you traveled, so deep is its measure,” 

reflects his deep contemplation on the nature of the soul and human understanding.

Heraclitus hailed from Ephesus, an important city in Ionia. He was known for his philosophy that emphasized the constant change and flow in the universe, encapsulated in his famous doctrine, "Panta Rhei," meaning "everything flows." This perspective highlighted the ever-changing and dynamic nature of reality. This particular quote delves into the seemingly infinite complexity of the human soul. Heraclitus suggests that the soul is an entity of such depth and intricacy that its full extent can never be completely understood or explored, no matter how much one tries. 

Today our understanding of human consciousness and the depth of the human mind have grown tremendously compared to the early 5th century BC when Heraclitus lived. It is possible now to understand these with some level of confidence. We at Robometrics® Machines have made significant progress in modeling consciousness. There may be aspects of human consciousness that remain elusive to artificial replication, underscoring the enigmatic depth of the human mind and soul. At the same time, modeling consciousness remains an important step towards Artificial General Intelligence (AGI) and at some point of time, the understanding of the human soul that will lead to immortality by proxy.

The Future of Embodied AGI: Mind, Body & Consciousness.

The future of embodied artificial general intelligence (AGI) would involve integrating advanced computational models such as Foundation Models into physical forms that mimic biological processes of living beings using technologies like DNA-based models and organoids. This artificial living organism doesn’t have to be in human form. Organoids could potentially replicate aspects of motion, cognition, and real world interaction of a living organism. 

A set of foundation models can simulate human-like memory and learning systems, akin to a digital limbic system. These models can use data from the environment to train the AI in real-world settings, simulating experiences that feed into the foundation models, analogous to a digital limbic system. This approach not only increases the adaptability but also enhances the AGI’s decision-making capabilities in dynamic environments.

Digital consciousness can be achieved in AGI manifested as an artificial living organism with human-like cognitive processes such as sensing, analyzing, acting based on internal and external stimuli; and embodiment. This paves the way for AGI systems that are not only interactive but have some aspects of self-awareness and self-management.