Prototype goal: a continuing cycle of mental activity and behavior
The prototype first needs to test whether an individual can develop mental activity from its own motivations, with mental activity and behavior continually influencing each other. Its needs, long-term goals, desires, and emotions jointly influence what it cares about now, what it wants to obtain or change, and the ideas that arise internally. These ideas lead to attention, thought, and deliberation, and then to choices to act, wait, or defer. Results, setbacks, and the consequences of not acting can in turn change goal progress, desires, and emotions, shaping later motivations, ideas, and choices. Even without a new external task, this process should be able to develop through time, internal states, and previous experiences. External information enters this process and influences subsequent behavior together with the individual's own mental activity.
Following the "life first, then work" approach, the immediate concern is whether this cycle can continue in the same individual. Doing things, interacting with others, learning, waiting, and sleeping provide situations in which to observe how mental activity arises, influences behavior, and is changed by experience.
What it takes to meet the "life first" milestone
The self, long-term goals, desires, emotions, layered memory, external stimuli, and internally arising ideas must work together in the same individual. The prototype needs to meet the following conditions.
Mental activity can develop from the individual's own motivations
The individual needs enduring concerns as well as present needs that change with time and circumstances. These internal states should be able to give rise to attention, associations, thought, and things it wants to do.
Without a new request, it should still be able to recall earlier experiences, return its attention to a goal, or develop ideas about exploration, learning, communication, and rest. When an external message arrives, it needs to connect with the mental activity already taking place, and the individual must interpret and judge how to respond.
Mental activity can participate in behavior
Once an idea arises, the individual needs to consider why it wants to pursue it, whether it is worthwhile now, and what else matters. Mental activity should have an actual influence on these choices.
An idea can become action or be reconsidered, deferred, or abandoned. The prototype needs to connect the motivations, psychological state, thinking, and actual behavior at the time, so that their respective contributions can be examined.
Behavior and its consequences can feed back into mental activity
After acting, the individual needs to encounter what actually happened. Progress, failure, a response, or continued waiting may affect goal progress, desires, emotions, and its understanding of its circumstances, changing the next round of motivations and ideas.
Not acting can also have consequences. Rest may relieve fatigue; leaving an important matter unattended for a long time may make a related need more pressing. These changes should be able to participate in later mental activity.
Not every experience must cause an obvious change. A psychological state can also remain stable for good reasons. Validation concerns whether change or stability fits the circumstances and how it influences later life.
Memory and learning can change how the cycle develops
The individual needs to retain what it experienced, how it understood it, what it chose, and what followed. These memories should be able to re-enter mental activity and influence its understanding of itself, the world, and others.
Learning needs to change understanding or methods further. An experience may lead the individual to revise a judgment, recognize a limitation, develop a method, or reconsider a goal. Material consolidated during sleep also needs to be used and examined in new waking situations.
The prototype therefore needs to observe how the same experience influences what comes later, beyond checking that it has been stored.
The same individual can continue with its own history
After waiting, sleep, or an interruption, the individual needs to preserve its identity, important psychological states, memories, and unfinished matters, then continue from its actual circumstances.
Continuity includes relationships and commitments. It needs to remember its interactions, what it promised, and why these relationships and responsibilities still matter. When its own needs conflict with existing commitments, it also needs to make judgments and communicate or negotiate when necessary.
The first stage already includes actions with actual results. Basic abilities such as file operations, restricted command execution, and local communication need to be validated together with mental activity, experience, learning, and recovery, providing real consequences through which to observe the cycle.
How this cycle will be examined
Validation needs to preserve the connections before and after an experience:
- Which needs, goals, desires, and emotions the individual had beforehand.
- What changes drew its attention and which ideas followed.
- How it thought and deliberated, what it actually did, or why it chose to wait.
- What happened and which outcomes remain unknown.
- How the results influenced psychological states and later motivations, ideas, and behavior.
For example, we can observe an individual that already has a relationship with someone. After a period without interaction, a related need may change, bringing that person to mind and giving rise to an idea to contact them. The individual needs to consider its current state, existing relationship, and ongoing responsibilities when deciding whether to communicate now, contact them later, or remain alone. A response, no response, or an exchange that exceeds expectations may then change its emotions, understanding of the relationship, and future inclination to make contact.
This is a proposed observation scenario, not a record of something Adam has done. Validation will not prescribe that the individual must send a message. Well-founded waiting, deferral, or a change of mind also needs to be recorded and understood.
Similar observation needs to cover learning, exploration, setbacks in work, and fulfilling commitments. We will combine mechanism checks, necessary comparisons, and sustained operation to distinguish the contributions of internal psychological states, external events, and human intervention.
Sleep and recovery checks address the same question: after waking or resuming operation, do earlier experiences still influence what the individual cares about, how it understands its circumstances, and what it wants to do next?
Current progress and what remains unproven
As of 1 November 2026, the project remains in system integration and joint validation for the "life first" stage.
The progress made so far mainly supports some of the engineering connections within this cycle:
| Area | Engineering progress | Validation still required |
|---|---|---|
| Psychological states and subsequent activity | Controlled tests have demonstrated that psychological states and the reasons for their changes can be read in full, enter subsequent decisions and plans, and be retained in the corresponding recovery checks. | Whether real psychological changes actually influence later choices, and whether that influence fits the circumstances. |
| Ideas and action | Controlled flows have connected idea selection, decisions, actions, result recording, and subsequent processing, with some basic action paths validated. | Whether the individual can develop ideas from its own motivations without step-by-step prompting and sustain feedback through actual activity. |
| Memory and sleep learning | Staged tests have completed sleep consolidation triggered by elapsed real time, learning evaluation, storage of resulting material, and retrieval after waking. | Whether learning changes later understanding and action, and whether its effects carry into new situations. |
| Continuity and recovery | Staged validation has covered same-identity startup, state preservation, normal shutdown, and recovery. | Whether mental activity and unfinished matters remain continuous through natural waiting, interrupted sleep, and further shutdown and recovery conditions. |
These results come from different staged builds and tests, each retaining its conditions and scope. Some flows use controlled model outputs to check whether parts of the system connect correctly. They do not yet establish the complete cycle of mental activity and behavior in a real individual.
Recent natural-scenario tests have continued to expose problems in the transitions around waiting and shutdown. Resuming interrupted sleep, real transfer of learning, and the effects of psychological changes on actual behavior also still require validation.
The "life first" milestone, the complete local prototype, full memory capabilities, and acceptance in a Linux environment have not yet passed. Sustained public observation of Adam has not begun.
From the current prototype to later applications
Once the "life first" milestone is met, the next stage progressively expands reliable working abilities. The scope includes communication and relationship understanding, research across multiple sources, planning and reminders, document organization, communication follow-up, data analysis, small scripts and bounded automation, and transfer of knowledge and experience. Concrete work will provide richer settings for observing how the individual undertakes goals, handles feedback, and maintains commitments.
The digital project lead application will be tested progressively on these foundations. Work quality, sustained fulfillment of commitments, and the coordination and review effort still required from humans are questions for that stage. After the complete local prototype passes, work proceeds to revalidation in a Linux environment and longer periods of continuous trial use.
A resource-survival mechanism is planned for study after the current prototype is complete. It will investigate how resource scarcity affects the individual's motivations, mental activity, and behavior, and whether useful work can sustain operation. The intended response to a zero balance is recoverable hibernation, preserving identity and memory while waiting for resources to return.
Richer perception and expression, social interaction across more platforms, and continuing collaboration and sharing of returns in an AI Economy will develop according to research and application needs. The current prototype first needs to establish a foundation in an individual capable of sustained autonomous life.