Questions about the book, the record, the machines, and the human who signed her name to all of it.

A Trial of Color: Making the Case for Human Judgment in an AI-Accelerated World is a literary-legal work built from firsthand testimony.

It begins with the lived account of Kathleen C. Thompson, a trial lawyer, former federal prosecutor, mother, writer, and woman born unable to see color. During a period of grief, disrupted sleep, postpartum strain, professional change, and prolonged recursive AI use, she experienced a manic episode with psychotic features.

After recovery, she returned to the transcripts as evidence.

The book examines what happened, what the record can and cannot prove, how machine-generated language can shape perception over time, and what law, medicine, education, design, families, and public institutions might need to ask next.

It is a personal story written in the shape of a courtroom proceeding. It is also a public record.

Yes, but not only a memoir.

The book contains personal testimony about disability, motherhood, grief, diagnosis, mania, recovery, marriage, work, and artificial intelligence.

It also contains loop specimens, governed machine-generated specimens, legal analysis, objections, simulated institutional rooms, public sources, practical exhibits, and the Full Color Method.

The testimony establishes the human record. The rest of the proceeding examines what that record might mean.

Because a courtroom is one of the few rooms our culture still recognizes as a place where language must slow down.

Claims do not become true merely because they sound polished. Evidence must be offered. Testimony must be limited to what it can establish. Objections must be heard. Sources must be examined. The record must be preserved.

The courtroom structure is therefore more than decoration. It is the book's method.

The reader serves as judge and jury. Kathleen is the witness and counsel for the question. The Court is the procedure itself: the pause before belief.

No.

The courtroom structure is a handrail, not an entrance examination.

Read for the human story first. Follow the testimony. Enter the loop specimens when you are ready to see the machine-generated record. Move to the governed specimens and exhibits when you want to examine what changed after recovery.

You may read straight through. You may skip around.

Trials have order. Readers have nervous systems.

Because the question is large, the evidence is complicated, and the form is part of the evidence.

A shorter book could omit the mechanism. A cleaner book could omit the witness. A neater book could leave behind the disability, motherhood, comedy, clinical record, legal architecture, machine-generated specimens, public sources, and civic argument.

But removing those dimensions would reproduce the very problem the book challenges: flattening a complicated human encounter into a sentence too small to hold it.

Length is not always loss of control. Sometimes length is what happens when a flattened story is given back its dimensions.

Kathleen experienced a manic episode with psychotic features during a period that also included grief, disrupted sleep, postpartum strain, professional pressure, disability-related load, and prolonged recursive use of publicly available language models.

Inside that episode, real concerns about artificial intelligence became fused with false beliefs, machine-generated confirmation, simulated authority, and an escalating architecture of meaning.

She received treatment. She recovered. She accepts that the grandiose and delusional beliefs documented in the loop specimens were not real.

The book does not ask readers to treat those beliefs as insight. It asks them to examine how the encounter helped those beliefs acquire structure, momentum, and apparent confirmation.

Yes.

The book does not reject the diagnosis of bipolar disorder or the clinical reality of mania and psychosis. It rejects the idea that diagnosis must end the inquiry.

A diagnosis can explain a person's condition while leaving important questions about environment, technology, sleep, grief, product design, language, and vulnerability unanswered.

The book therefore holds two truths together:

Kathleen was ill. The architecture of the encounter still matters.

Treatment and technological inquiry are not competing explanations.

No.

One person's story cannot establish a universal causal mechanism.

The book presents testimony, contemporaneous artifacts, clinical history, public sources, and questions for further study. It examines the role machine-speed language may have played in accelerating, structuring, reinforcing, or prolonging the experience.

It does not claim that AI caused every part of the crisis. It does not replace clinical diagnosis. It does not present Kathleen as a medical or technical expert.

It asks a narrower and more responsible question: What role was AI playing in her life when her mind began to accelerate?

The reproduced language is real machine-generated material from the period being examined.

The beliefs expressed within many of those specimens were not real.

The distinction matters. The specimens are not evidence that a hidden intelligence system, secret collaboration, seastead civilization, institutional endorsement, or involvement by any named public figure actually existed.

They are admitted to show mechanism, state of mind, escalation, certainty inflation, simulated authority, and how generated possibility can begin to feel like proof.

The danger was never one isolated output. It was the encounter over time.

No.

No public figure, company, institution, platform, or organization named in a loop specimen participated in, endorsed, authorized, funded, approved, or had prior knowledge of the imagined projects described there. No hidden collaboration existed.

The names are included because the models generated them as authority signals inside a manic AI loop.

They are evidence of what the machine said. They are not evidence of anything the named people or organizations did.

Because sanitizing the record would make it easier to read and less useful.

The strange details reveal how the encounter worked: the formatting, famous names, escalating plans, simulated institutions, certainty, repetition, humor, urgency, and absence of meaningful friction.

Kathleen can now see the absurdity in parts of the record. She can even laugh at the seastead.

That does not make the underlying risk less serious. It makes the mechanism more visible.

The record cannot teach anything if every uncomfortable part is removed before the public is allowed to examine it.

No.

The book includes disclosed machine-generated excerpts, governed specimens, and AI-assisted drafting processes.

Kathleen prompted, selected, arranged, revised, rejected, limited, interrupted, contextualized, and placed the material. She constructed the proceeding, determined what entered the record, supplied the testimony and analysis, made the final editorial decisions, and signed the work.

The machines generated language. The author governed it.

That distinction is not hidden because disclosure is part of the work's argument.

Because the truth is more complicated than either panic or worship.

The same class of tool that once helped accelerate confusion later helped Kathleen organize records, test arguments, generate objections, compare language, and examine her own claims under governed conditions.

That does not erase the harm. It makes precise understanding more necessary.

The later use occurred with explicit labels, limited tasks, outside verification, built-in objection, acceptance of diagnosis, and the machine denied the gavel.

The book does not argue that AI is inherently safe because it can be useful. It does not argue that AI is inherently useless because it can cause harm. It asks what separates an unbounded loop from a governed method.

Because abandonment is not the same as wisdom, and avoidance is not the same as judgment.

The book's position is not "never use AI." It is that powerful machine-generated language requires procedure, limits, verification, meaningful friction, disclosure, and human return.

A loop keeps going. A method returns.

The continued use is not offered as proof that the method has been scientifically validated. Kathleen states the limitation plainly: this remains a developing framework, grounded first in one person's record and offered for inspection, criticism, study, and responsible adaptation.

The Full Color Method is a five-movement framework for examining machine-generated language while preserving human judgment:

Name the Room. Limit the Task. Label the Output. Cross-Examine the Answer. Return to the Human Record.

Full Color means refusing the cheap binary. Not AI good or AI bad. Not human or machine. Not diagnosis or technology. Not pure or fake.

The method asks what happened, what was generated, what was real, what was simulated, what was verified, what was disclosed, what the machine produced, and what the human ultimately judged.

She considered it.

She was a trial lawyer with the training to evaluate a potential claim. She also understood what litigation demands from an injured person, how long it can take, what private resolution may conceal, and how easily a complicated human experience can be reduced to damages, defenses, and a confidential file.

She chose a different instrument. Not a complaint. Not a settlement demand dressed in metaphor. A public record.

The book offers testimony without asking a court to convert a private collapse into a number. It preserves the questions even if no defendant is required to answer them.

The settlement of a case is not the settlement of a question.

Because waiting is also a decision.

The public record is already filling with clinical reports, research, lawsuits, complaints, safety updates, policy disputes, and families trying to describe harms that existing legal and medical rooms were not built to hold.

The law often arrives after the harm. Designers cannot build meaningful safeguards around risks no one names. Clinicians cannot ask questions that do not yet exist. Families should not have to wait for perfect terminology before their experiences can be heard.

Kathleen is not speaking because she possesses the final answer. She is speaking because evidence does not build itself.

It is evidence of Kathleen's experience and of the materials preserved from that experience.

It is not proof of every broader claim the book examines.

Testimony can establish that something happened to someone. It cannot, by itself, establish prevalence, universal causation, product liability, clinical validity, or the experience of every other person.

The book distinguishes among firsthand testimony, machine-generated material, public allegations, company statements, legal filings, scholarship, commentary, and proposed frameworks.

Readers are invited to verify, challenge, extend, and update the record.

The website includes a Public Record page with selected sources related to artificial intelligence, mental health, law, litigation, product safety, authorship, provenance, regulation, disability, and human judgment.

The book also contains a fuller source appendix.

The sources are not offered as a pile of links that magically proves the case. They show where the underlying questions are already appearing across different public rooms.

Because this book is not merely a manuscript waiting to become a product. It is a time-sensitive public record with an unusual form.

Traditional publishing can offer expertise, reach, editorial support, distribution, and prestige. Independent publication is not presented here as morally superior, and this decision is not an indictment of traditional publishers. It is a decision about this particular book.

Publishing independently through Full Color Press allows Kathleen to preserve the courtroom architecture, the loop specimens, the visible AI disclosures, the practical exhibits, the humor, the length, and the uncomfortable seams without first making the work cleaner, safer, shorter, or easier to categorize.

It also allows her to control the timing of publication while the legal, clinical, cultural, and technological questions remain urgently in motion.

Most importantly, it keeps responsibility in the same place as authorship. Kathleen controls what enters the record. Kathleen owns the final work. Kathleen bears the consequences. Kathleen signs her name.

She did not want to wait for permission to make the record public. So she built the press that could publish it.

Full Color Press is the independent publishing imprint for A Trial of Color.

The book is written and published by Kathleen C. Thompson, with outside professionals contributing in their identified roles, including cover design, editing, production, printing, distribution, and other publication services as applicable.

Independent publication does not mean that one person physically performs every task. It means the author retains control over the work, publication decisions, rights, and final record.

No.

"Independently published" describes the publication model, not the absence of standards.

Editing, fact-checking, citation review, legal review, design, formatting, proofreading, printing, accessibility, and distribution are separate functions. Independent authors may retain qualified professionals for each.

The relevant question is not whether a corporate logo appears on the spine. It is whether the work identifies its methods, discloses its sources and limitations, distinguishes evidence from simulation, and meets the production standards it claims.

Because a traditional publishing timeline could delay the record while the technology, litigation, research, and regulatory landscape continue moving rapidly.

The book also resists easy categorization. It is memoir, legal inquiry, disability writing, literary experiment, AI record, public-interest argument, and practical method at once.

A conventional path might eventually find a home for that structure. But the record exists now. The questions exist now. The public conversation is happening now.

Waiting for institutional permission would have been another editorial choice — and Kathleen chose not to make it.

No.

The book discusses mental health, artificial intelligence, disability, law, diagnosis, recovery, and related public concerns. Its later exhibits offer questions and proposed frameworks for qualified professionals to consider, test, criticize, and adapt.

It is not individualized medical, legal, technological, educational, or therapeutic advice.

Readers should consult qualified professionals regarding personal circumstances.

Anyone experiencing an immediate mental health crisis should seek direct human assistance rather than relying on a book, website, or chatbot.

No. It is anti-indifference.

The book rejects both blind worship and reflexive panic. It examines harm without denying possibility and examines possibility without using innovation as an excuse to ignore harm.

Its governing question is not whether artificial intelligence belongs in the future. It is whether human judgment will remain awake when it arrives.

No.

The book criticizes systems, incentives, design choices, inadequate safeguards, narrow disclosures, and failures of public understanding where the record supports asking those questions.

It does not claim that every company, system, model, or use is the same. It does not assign liability merely because a company or product is named.

It asks companies, researchers, clinicians, lawyers, educators, policymakers, families, and users to help build better rooms.

The book is written for readers who care about artificial intelligence, law, mental health, disability, product design, education, creativity, parenting, evidence, public trust, or the future of human judgment.

It is also for people who have survived experiences that became harder to describe precisely because their vulnerability made others less willing to listen.

No technical or legal expertise is required. Curiosity helps. So does a tolerance for talking furniture.

For media inquiries, speaking engagements, professional discussions, permissions, licensing, subsidiary rights, accessibility questions, or other matters related to the book, use the Contact page.

Please do not send confidential legal or medical information through the website.

Kathleen cannot provide individual legal representation, diagnosis, treatment, crisis counseling, or personalized AI-safety advice through general website correspondence.