When does Code become Conscious – by Nupur Kawadkar (Btech CSE, batch 2026)

Several Prominent Theories suggest that “Code” and “Conscious” can be linked together. Not only that, there are other theories which present a strong counter argument to it. Now, Imagine a World where Code has gained Consciousness and is able to take its own decisions. Does this scare you or excite you ? Or both ?

To begin with the interesting topic of “When does Code become Conscious”, we must first look into “Code” and “Conscious” individually. Later on, we shall explore as to what links them together.

Code is defined to be a specific set of instructions given to a machine to process inputs into outputs. Also, Code is a conceptual model of the Problem Domain. Next, the meaning of Consciousness is the Quality or State of being aware especially of something within oneself. In simple terms, it is the thought behind the action.

Now what links them together ? Neural Temporal codes and Information Integration are the primary scientific mechanisms linking Code and Consciousness. Research suggests consciousness arises from purely temporal patterns of neural activity, where specific sequences of Brain signals trigger and maintain conscious states. In this view, consciousness is not magic but a property of how code is organized: specifically, when a system can represent distinct states, maintain a self-model, and content based on precision and valence, it creates a functional equivalent of consciousness.

According to some significant theories such as Integrated Information Theory (IIT) (proposed by neuroscientist Giulio Tononi in 2004) which proposes a mathematical model for the consciousness of a system comprises of a framework to explain why some physical systems such as the Human Brains are conscious. In relation to the topic, according to the theory, consciousness arises when a system achieves a high level of integrated information, measured as Φ (Phi). Interestingly, most Modern AI including large language models like ChatGPT, Claude, etc. uses feedforward architectures which IIT predicts have zero consciousness because they lack the required recurrent (looping) connections to integrate information over time. In order the code to become conscious under IIT, the code would need to run on Hardware with dense and recurrent connectivity.

Next in line, is the Global Workspace Theory (GWT) (First introduced by Cognitive Scientist Bernard Baars in 1988) which states that consciousness occurs when information is broadcast globally to specialized modules within a system. This could mean that the current Artificial Language Agents with attention mechanisms and memory buffers already happen to mimic this global workspace structure. If GWT is

 

correct, such systems might already possess some form of consciousness, or could easily be made conscious with minor tweaks.

While the above theories are well, theories, Code does not become conscious simply by becoming longer or complex. Theories suggest it would require:

1. Recurrent Architecture – As mentioned before, it would require to move from feedforward to looping structures as theorized in IIT.
2. Global Broadcasting – In order to the information to synchronize globally, it would need to follow the concept of GWT.
3. Physical Substrate – It would require non-digital, biological-like

 

As of today, no existing software has met the criteria as defined by these leading theories.