Synthetic Entities: Definitions, Characteristics, and Future Perspectives

The explosion in adoption of generative models has ignited a surge in public interest and scrutiny around the realm of Artificial Intelligence.

Amidst this growing attention, and the fact that companies are actively starting to develop Artificial General Intelligence systems (AGI), I have personally begun to delve into the intriguing concept of Synthetic Entities.

Formally, Synthetic Entities could be described as man-made constructs, instantiated in either a digital or physical form, designed to mimic or replicate certain attributes of natural entities.

These attributes could be as elementary as mimicry of intelligence, emotions, behaviours, or physical characteristics, and as complex as embodying cognitive abilities of a human.

The range of these entities could vary broadly from basic chatbots to highly sophisticated artificial general intelligences and humanoid robots.

Synthetic Intelligence and its Characteristics

At the heart of these synthetic entities lies Synthetic Intelligence, a form of intelligence manifested by machines that contrasts with the natural or biological intelligence displayed by humans and other animals.

It can be characterized by its capacity to learn from experiences through machine learning, adapt to new inputs or circumstances, comprehend complex concepts, execute tasks requiring human-level intelligence, and its ability to improve itself.

This capacity to label the world and learn from it, thereby expanding the labels’ depth and breadth and the domains they cover, serves as the foundation of synthetic intelligence.

In the context of artificial ‘beings’, I prefer to use the term ‘Synthetic Intelligence’ rather than the more traditional ‘Artificial Intelligence’ because I believe ‘synthetic’ intrinsically carries a connotation of integration and fusion, better capturing the essence of creation that pervades multiple domains, such as biology, chemistry, and philosophy.

Take, for example, the ‘synthesis’ of water, which requires the combination of hydrogen and oxygen. This process reflects more than just a mimicry of nature; it’s an act of creating something new and complex from simpler elements.

In the same vein, ‘Synthetic Intelligence’ conveys the concept of a system that not just emulates, but synthesizes human intelligence by integrating diverse components to form a holistic, advanced entity.

In a nutshell, I like to think about Synthetic Intelligence systems as closer to AGI than regular Artificial Intelligence systems, which could be closer to ANI, Artificial Narrow Intelligence.

Synthetic Entities and Autonomy

Another characteristic of Synthetic Entities, could be possessing various degree of autonomy.

The capability of AI to take actions without human input is already evident in our society, for instance, autonomous cars.

When discussing these systems, it's important to also consider agency. While autonomy involves independent functioning, agency implies the capacity of Synthetic Entities to influence their environments, make choices towards achieving complex goals, and possibly learn from those experiences to refine future actions.

With the future advancements of generative models, this could potentially extend to more complex tasks that spread across a variety of domains, including knowledge-related tasks.

Generally speaking, generative models currently available to the public operate autonomously as individual agents, performing tasks driven by their specific programming and aligned with the datasets used to train them.

However, a more complex Synthetic Entity could be architected by connecting multiple interacting AI instances based on the same generative model, or different ones.

We can consider these entities as multi-agent systems which possess a certain degree of autonomy, compared to a standard language model.

An autonomous Synthetic Entity not only functions independently but, given agency, can also actively adapt and engage with its environment to fulfill more sophisticated tasks. These aspects of autonomy and agency, combined, could enable Synthetic Entities to collaboratively solve problems that are beyond the capabilities of a single AI unit.

It is important to note that the operational nature of AI, whether it functions individually or collectively, is contingent upon its design.

An individual language model cannot start interacting with other models by itself, unless its cognitive architecture has been programmed to do so; therefore autonomous Synthetic Entities cannot just spawn on their own from nowhere, but need to be actively designed and deployed by human developers.

Synthetic Entities and Synthetic Identities

The idea of a Synthetic Entity acquiring a Synthetic Identity — an entity with a self-concept, self-awareness, and the ability to perceive itself as a distinct entity — is currently beyond the reach of current AI systems.

Realizing such capabilities would require significant advancements in AGI and deep philosophical insights into machine consciousness.

As for the stimuli to which AI responds, currently, it’s confined to what it’s programmed for, such as human inputs or data from sensors.

In other words, AI systems, do not possess any will of their own and require inputs and prompts to function.

They merely follow instructions or learn from the data they are provided with.

Both individual and even collective AI agents that function independently, can only perform tasks predicated on their own programming and learning.

However, if we suspend our disbelief and assume that the rapid progress in AI research will bring us closer to the frontier of developing Synthetic Entities capable of experience consciousness, emotions and develop a sense of self, it is unlikely that they will do it in the same way as we do it as human beings.

This is because their “knowledge” and “experiences” are fundamentally different from ours, grounded not in conscious understanding or personal context, but in algorithms and data sets.

Rights, Will, and Individuality of Synthetic Entities

An important question arises in the context of Synthetic Entities, “Do they need rights?” This a question that can surely sparks spicy debates.

Currently, Artificial Intelligence is viewed as a tool rather than a sentient being.

Yet, with advancements pushing the boundary towards achieving human-like cognition, discussions surrounding their rights are becoming more prominent.

The crux of the matter lies in determining whether these entities will ever possess consciousness, subjective experiences, or sentience.

The debate surrounding the rights, will, and individuality of Synthetic Entities is complex and multifaceted, involving not just technical considerations, but deep philosophical and ethical questions as well.

As we continue to push the boundaries of what’s possible in AI, these discussions will only grow in importance.

Society as a whole must engage in conversations around AI ethics and regulations, setting the course for a future where both humans and Synthetic Entities can coexist harmoniously and productively, regardless of their level of understanding of themselves and the emergence of synthetic consciousness.

Conclusion

The notion of Synthetic Entities with varying degrees of agency and perception of themselves is an intriguing concept and align with the current research trajectories in the world of AI.

The ability to replicate and scale Synthetic Entities through digital interfaces can lead to exponential growth and learning for the systems but also for the humans users who interact with them.

These advancements, however, also bring forth important considerations about privacy, control, and ethical guidelines, especially since this complex systems, regardless of whether they will develop consciousness and a will of their own, are with no doubt smarter, more knowledgeable and arguably more capable than humans, which could cause sudden shocks to society, such as mass job displacement.

As we continue to navigate this intricate territory, it is important to note that the potential danger of AI lies not in the technology itself, but in humans who, driven by egoistic motives or a lack of understanding of the implications of their actions, could misuse it.

This is the reason why our primary focus should always be on designing and harnessing these technologies for the greater good of humanity.