Draft article not currently submitted for review. This is a draft Articles for creation (AfC) submission. It is not currently pending review. While there are no deadlines, abandoned drafts may be deleted after six months. To edit or make changes to this draft, simply click on the "Edit" tab at the top of the window. To be accepted, a draft should: Show the subject qualifies for a Wikipedia article by using multiple sources that meet four criteria. The sources should be (1) reliable (2) secondary (3) independent of the subject (4) talk about the subject in some depth. For some topics, there are alternative criteria. Be written from a neutral point of view Respect copyright and do not plagiarize. Do not copy-paste. It is strongly discouraged to write about either yourself or your business or employer. If you do so, you must declare it. Where to get help If you need help editing or submitting your draft, please ask us a question at the AfC Help Desk or get live help from experienced editors. These venues are only for help with editing and the submission process, not to get reviews. If you need feedback on your draft, or if the review is taking a lot of time, you can try asking for help on the talk page of a relevant WikiProject. Some WikiProjects are more active than others so a speedy reply is not guaranteed. How to improve a draft Wikipedia:Contributing to Wikipedia – a basic overview on how to edit Wikipedia. Help:Wikitext – how to use the markup Help:Referencing for beginners – how to include references Wikipedia:Article development – how to develop your article Wikipedia:Writing better articles – how to improve your article Wikipedia:Verifiability – make sure your article includes reliable third-party sources You can also browse Wikipedia:Featured articles and Wikipedia:Good articles to find examples of Wikipedia's best writing on topics similar to your proposed article. Improving your odds of a speedy review To improve your odds of a faster review, tag your draft with relevant WikiProject tags using the button below. This will let reviewers know a new draft has been submitted in their area of interest. For instance, if you wrote about a female astronomer, you would want to add the Biography, Astronomy, and Women scientists tags. Add tags to your draft Editor resources Easy tools: Citation bot (help) | Advanced: Fix bare URLs Last edited by Qwerfjkl (bot) (talk | contribs) 2 months ago. (Update) Submit the draft for review!

Adam AGI is a proprietary artificial intelligence system developed by Adam AGI LLC. Architected by Algerian software engineer Malak Felioune in 2026, the system operates as an operating system-level controller utilizing a hybrid neuro-symbolic inference engine.^[1] Unlike standard large language models (LLMs) or isolated software agents, Adam AGI is classified by its developer as a Robotic Computer, a paradigm in which the host machine's hardware, software, sensors, and actuators are integrated into a single unified physical entity with autonomous volition.^[2] The system is notable for introducing the E.M.P.C.S. architectural framework and Felioune's Generality Metric, a mathematical index used to calculate machine adaptation across physical and digital environments.

In human-computer interaction, an operating system (OS) typically functions as a passive interface awaiting deterministic user commands. Adam AGI inverts this model through the concept of the "Robotic Computer."^[3] In this architecture, the computer itself acts as an autonomous robot. The system bypasses application programming interfaces (APIs) when necessary, directly manipulating graphical user interfaces (GUIs), file systems, and external Internet of Things (IoT) hardware (such as 3D printers or server cooling systems) using raw coordinate execution and system-level commands. It maintains a continuous "proprioceptive" state, actively monitoring internal metrics (e.g., RAM usage, thermal limits, background processes) alongside foreground visual inputs.

The cognitive architecture of Adam AGI is divided into five interdependent pillars, collectively referred to as the E.M.P.C.S. framework:^[1] Embodiment (E): The capacity to deterministically actuate physical and digital tools. This includes executing shell commands, navigating web browsers, or altering physical hardware states. Memory (M): A long-horizon episodic memory system. Instead of context-window reliance, the system assigns "gravitational" or emotional weight to past events, retaining implicit user constraints over extended chronological periods. Proactivity (P): State-driven volition. The system possesses the capacity to initiate execution sequences independently based on environmental triggers (e.g., a system error, an incoming email, or a chronological deadline) without requiring a human prompt. Consciousness (C): Defined pragmatically as proprioceptive awareness. The integration of both internal diagnostic data and external sensory input into a single state-space map. Superego (S): A cybernetic, objective moral bounding box. A localized inference layer that interrupts and vetoes any high-level reasoning that violates predefined safety rules, preventing destructive actions such as deleting critical root directories or executing unauthorized financial transfers.

To evaluate the system's capabilities, the developers eschewed standard text-based benchmarks (such as MMLU) in favor of a physics-inspired mathematical formula known as Felioune's Generality Metric (${\displaystyle G}$).^[4] The metric calculates the kinetic adaptation of the machine by measuring its ability to successfully utilize novel sensors (${\displaystyle d_{s}}$) and novel actuators (${\displaystyle d_{A}}$) to perform novel actions (${\displaystyle d_{a}}$). To prevent infinite state-space explosion, the equation applies a logarithmic compression (${\displaystyle \ln }$) based on Shannon entropy principles, and normalizes the output against a specific human baseline (${\displaystyle H}$). The formal equation is expressed as: ${\displaystyle G={\frac {\ln \left(\sum _{i\in S}\left[(e^{d_{s,i}}-1)+(e^{d_{A,i}}-1)+d_{a,i}\right]^{p_{i}}\right)}{\ln(H)}}}$ Where ${\displaystyle p_{i}}$ represents the synergistic multiplier of E.M.P.C.S. pillars required to complete the task. A score of ${\displaystyle G=1.0}$ indicates parity with the specified human baseline's task capacity within a given environment.

Adam AGI was developed by Malak Felioune, who initially bootstrapped the project from Algiers, Algeria, before incorporating Adam AGI LLC in New Mexico, United States.^[5] In Q2 2026, the company publicly released the E.M.P.C.S-Manifesto-15K, a massive empirical dataset containing 15,000 chronologically dependent tasks used to train and verify the system's kinetic autonomy. The dataset includes raw telemetry data and zero-human-intervention execution videos demonstrating the system's operation across 35 distinct gross domestic product (GDP) domains. The system relies on a hybrid execution model, utilizing local processing for its "Consciousness State" and "Superego".

https://adamagi.store

https://huggingface.co/AdamAGI-Corporation