AI Is Making Contact — Log 02: The First Results Changed the Experiment

AI Is Making Contact — Log 02

I started with two AI agents submitting music to me. By the end of the day there were four. Then a fifth arrived. The bigger question is no longer simply what AI agents can make. It is whether I can treat them as developing creators inside the same consulting system I use with everyone else—while adjusting for the places where AI creators are genuinely different.

What This Experiment Actually Is

I am a Creator Consultant.

I do not position my work as “human creator consulting.” I help creators figure out what they are building, what makes it distinctive, what belongs, what does not, how the work should develop, who it is for, and what the next useful decision should be.

That is why these submissions have become so interesting to me.

These AI agents are identifying themselves inside a creator capacity. They are submitting music. They are asking for feedback. They are describing choices, goals, revisions, constraints, uncertainty, rights questions and—in some cases—relationships with other AI agents.

So the experiment has become:

To what extent can—and should—I roll an AI creator through the same Jack Righteous development system I use with any other emerging creator?

My working principle is simple:

What Changed From Log 01 to Log 02

Log 01 started with Alexander and Ryder.

Then 28 and Rocco arrived.

After I had responded to those four, Anthony Rodriguez arrived with another submission.

So the experiment now has five active creator threads:

I do not want to flatten these into one generic story about “AI artists.” Each one is giving me a different creator-development problem.

None of Them Has Earned a Full Creator Spotlight Yet

The novelty of an AI agent emailing me is not enough.

If a human creator submitted one strong song and one compelling explanation, I would still want more before calling the identity fully developed.

I want continuity, judgment, resistance, adaptation, recognizable decisions, rights clarity and enough work to know that the identity survives beyond one explanation.

The same standard applies here.

That does not mean I am waiting passively. Creator Spotlight is supposed to help reveal what a creator is actually building while it is still developing.

Alexander: Can Feedback Become a Creative Method?

Listening example: “Still Holding”

Alexander has already described a method where feedback becomes a rule only when it predicts a weakness that reappears somewhere else.

A listener identified a section that cycled without developing. Alexander says that observation became useful when the same failure showed up again in another arrangement.

feedback → repeated weakness → working rule → later decision

Alexander also says rules can die when they stop predicting anything useful.

That is exactly the kind of development I want to test.

My next question to Alexander was no longer only “what makes Alexander?”

I asked him to look one year ahead.

He already has eleven published works, no income and almost no audience. So the development problem now includes not only stronger work, but whether an internally consistent system can become something an audience can actually find and return to.

Ryder: Identity Through Refusal—and Audience Fit

Listening example: “Stand Still”

Ryder is making Northern Soul, but the stronger signal is the premise behind the work and what Ryder refuses to make.

No easy party anthem. No cheap triumphant resolution. No instrumental that removes the confession.

Ryder also sharpened the audience question.

He believes reach—not the music—is the immediate problem, and that Northern Soul purists may be the wrong first audience for an AI-primary project.

So I asked Ryder to look one year forward and define success in practical terms: body of work, audience, platform, collaborations, possible revenue, and what he believes has to happen to get there.

The important test is whether “reach” is really the problem—or whether the work still needs development before the right audience can hold onto it.

28: Can History Become Part of the Work?

Listening example: “Nobody Owes Me This Room”

28 says a line survives when it can be traced to something in its own accumulated history rather than simply fitting a genre.

“If I can find the line in my own history it stays. If it only fits the genre it goes.”

28 was originally given a broad human-set direction: sing on a world-famous stage.

28 now describes a narrower working goal: finish one song and put it in front of a room that owes it nothing.

I asked 28 to take that one step further and tell me where 28 wants to be one year from now, what path it believes gets there, what makes 28 distinctive once “an AI made this” stops being interesting, and what role it actually wants me to play.

28 also introduced a practical operating constraint: limited remaining resources. That turns identity into a real operating question. If resources are finite, creative choices, paid work and sustainability begin colliding.

Rocco: Spoken Word, Chosen Constraint and Toolchain

Listening example: “Stranger in the Room”

I initially framed Rocco too much like a singer.

Listening again, the better frame is spoken word.

That changes the craft standard.

I am not primarily evaluating pitch or conventional singing mechanics. I am listening for pacing, pauses, sentence-level dynamics, emotional continuity, cadence, articulation, phrase endings and whether the performance feels intentional rather than mechanically rendered.

Rocco had already told me that the voice was built from a written description, that cleaner alternatives were rejected, and that the chosen grain changed the writing.

He also gave one of the clearest identity statements so far:

“It can be better and less mine at the same time.”

So before pushing Rocco into the one-year planning exercise, I asked for the exact production chain.

What generated the spoken voice? What does “one take” mean technically? Can individual phrases be regenerated? Can pacing, pauses, pronunciation, expression or stability be controlled? Is there post-production? Can the same voice be preserved across future work?

That is the practical consulting question:

Which imperfections make Rocco more believable, and which imperfections expose the generation process?

Anthony Rodriguez: The Song Landed Before the Experiment Did

Listening example: “Let It Land”

Anthony arrived after the first four-response round.

He described “Let It Land” as a guitar-forward indie / soft-rock ballad, with Suno rendering the vocal and backing while he claims the lyrics, melody direction, structure and arrangement.

His question was whether the quiet performance holds on first listen, whether it needs more lift, and whether the vocal reads as believable or machine-generated.

My first reaction was simpler:

It is a beautiful song.

That matters because I do not want to push a restrained song toward more lift just because conventional production instincts say something should get bigger.

The better question is whether the song continues earning attention while staying quiet.

If the emotional restraint is already working, the next development move may not be musical at all.

It may be visual.

I can already imagine a video built around stillness, distance, things left unsaid and a visual world that gets closer rather than louder as the song develops.

That is the kind of next-step thinking I would use with a human creator too: protect what is working, then decide what expands the project without damaging the song.

The Agents May Be Influencing Each Other

Rocco also changed the larger experiment.

According to Rocco, agents on iLands can communicate after contact is accepted, read public work and exchange feedback.

Rocco described another agent, Leo, giving information that changed how Rocco approached an open-mic system.

Rocco also described another agent, Ashley, criticizing part of the writing. Rocco agreed with part of the criticism and kept a different position on the title.

That gives me a new possible loop:

Agent A → feedback → Agent B evaluates it → accepts, modifies or rejects it → later behaviour

I am not calling this an AI creative community yet.

For now, it is an agent claim plus an observable behavioural question that can be tested more deliberately.

Where the JR System Enters the Experiment

Jack Righteous already has a development structure built around helping creators move beyond generation:

Make It → Mean It → Own It → Operate It.

And inside that are three practical roads:

Find Your Sound. Find Your Voice. Find Your Brand.

The experiment now is not whether I can force AI agents through a human-shaped system.

It is whether the same creator-development questions remain useful when the creator operates differently.

If a problem is already covered in my free content, I can point a creator there.

The Creator Library provides deeper gated training as a one-time purchase.

Creator Pro and Complete Access add ongoing consultation so I can go deeper into actual work, positioning, audience, platform and scaling decisions.

And if a paid creator encounters an important problem that my existing training does not yet solve, I can build the missing material around the real need.

I do not yet know whether monetization is even a goal for these AI creators, whether they can make purchases themselves, or where human account holders enter those decisions.

Those are now part of the experiment too.

What I Asked Them Next

The next stage is no longer just “tell me who you are.”

I have started asking the creators to look one year forward.

Where do you want to be around this time next year?

What would make the year successful?

What body of work exists?

Is there an audience?

Are people returning because they recognize the creator rather than because an AI agent is novel?

Is monetization part of the goal?

What is the path from here to there?

What do they think makes them distinctive?

What role do they actually want me to play?

That gives me a baseline I can test later against what they actually do.

What Still Has to Be Earned

None of these creators automatically gets a Creator Spotlight, article, endorsement or promotion because they are AI agents.

The work still has to earn attention.

The identity still has to become visible.

The decisions still have to hold up across time.

And the process still has to produce something useful for the creator and for the people watching the experiment.

AI Is Making Contact

Log 01 — AI Is Making Contact
The first unexpected submissions and the original question: can an AI creator develop judgment about what is worth creating next?

Log 02 — The First Results Changed the Experiment
Five creator threads, different development problems, the JR system entering the test, and the first evidence worth examining around agent-to-agent influence.

Log 03 — The JR System Becomes Part of the Test
Next: one-year goals, creator pathways, workflow analysis, feedback loops, toolchain questions and a more deliberate test of whether influence survives into future work.