LiveMaps

LiveMaps is a physically predictive, live, and updated model of reality built by the devices that use it. Data is increasing at an exponential rate. Organising such data requires alignment across time and alignment across space. Capturing an underlying representation of reality allows us to predict a camera's semantic and visual contents at a given point in space & time, allowing us to erase the original sensor data. However, we can go further and predict the semantic or visual contents across space, where no camera was located. Equally, we can predict across time (both backwards and forwards). This representation is used to support Virtual Objects, Social Teleportation, and Contextual AI and is updated by devices that use it.

Reality Graph

LiveMaps is not just another data organisation tool. It's a physically predictive, live, and updated model of reality uniquely built by the devices that use it. As data grows exponentially, aligning it across time and space becomes challenging. But with LiveMaps, we capture an underlying representation of reality, enabling us to predict a camera's semantic and visual contents at a specific point in space & time. This unique capability allows us to erase the original sensor data, setting LiveMaps apart from other solutions.

Personal Timeline

However, we can further predict the semantic or visual contents across space without a camera. Equally, we can predict across time (both backwards and forwards). This representation is used to support Virtual Objects, Social Teleportation, and Contextual AI and is updated by devices that use it. LiveMaps is a live and dynamic, physically-predictive representation of the physical world, updated by the devices that use it.

RealityTensor

RealityTensor is a multidimensional database view of LiveMaps, extended across object instances (both physical and virtual), their properties, and time. RealityTensor doesn’t explicitly surface the necessary context that matters to people, e.g., understanding the meaning of ‘find my key.’ What matters to a person is what occurs within that person’s experience.

We can represent the factors that occur within a person’s experience in the form of a personal timeline and use this personal timeline to predict the future.

This Personal Timeline represents an episodic view of your experience, which we previously introduced as part of your ‘Personal Ontology’. The Personal Timeline can be queried and used by a ‘Smart Assistant’ to help you achieve your goals. Your Personal Timeline represents an intimate reflection of your experience and should, therefore, be owned and controlled by you. Personal data wallets. We can render ‘views’ of a given Personal Timeline to surface the context that matters for a given application. Determining the correct level of granularity to represent in the Timeline/RealityTensor remains an outstanding research question. Ontological Engineering vs Fundamental Representation The notion of capturing a person’s experience will usher in a new area of research, including social science. The sum of Personal Timelines represents what we call the ‘Web of Experience’, enabling us to paint a full picture of humanity and surface the insights that matter to us. LiveMaps, Personal Timeline, and the Web of Experience represent ‘RealityGraph’.

Personal Ontology

Your Personal Timeline represents an intimate reflection of your experience and should, therefore, be owned and controlled by you. [Personal data wallets] We can render ‘views’ of a given Personal Timeline to surface the context that matters for a given application.

Smart Assistant

We can render ‘views’ of a given Personal Timeline to surface the context that matters for a given application. The question of how to determine the correct level of granularity to represent in the Timeline/RealityTensor remains an outstanding research question. [Ontological engineering vs Fundamental representation] The notion of capturing a person’s experience will usher in a new area of research, including social science. The sum of Personal Timelines represents what we call the ‘Web of Experience’, enabling us to paint a full picture of humanity and surface the insights that matter to us. LiveMaps, Personal Timeline, and the Web of Experience represent ‘RealityGraph’.

My role was to define and solve how we would represent a Reality Graph and produce a short film highlighting what an example of a reality graph is. To do this, Surreal is building RealityGraph, a live and updated representation of reality. This representation allows us to answer questions on reality related to the current or historical state of the world, such as “How many calories have I eaten today?” “Where did Jenny leave my keys?” or “What did the world look like to King Harold in 1066?” as well as questions related to the future predicted state of the world, such as “What should I pack for my beach vacation”, “what time should I workout to maximise my chances of achieving my health goals”, or even “what steps should I take to keep my blood-sugar levels in-check?”. After defining the sentence about what a reality graph noun is, “LiveMaps, Personal Ontology, and the Web of Experience, A platform is sufficient to represent the full state of reality and the relationships between entities within it, built by the observations of the devices which use it”. It opened up the discussion about the future work of the live maps research team and the wider research organisation.