Era: Ainternals and AI Client Work, 2024-2025
The Ainternals and AI client work era covers the period when Steven’s AI practice moved from private cognitive scaffolding into concrete external utility. The iMessage corpus shows this most clearly through the 2024-2025 collaboration with Jean-François Fèvre and Gilles SABART, where Steven built and operated AI-assisted transcription, interview analysis, timeline, document synthesis, and legal/investigation workflow tools.
Position in the wiki
This era fills a gap in the current Notion-derived chronology. The existing wiki covers the post-Arthur and freelance collapse arcs through 2023, then jumps to the January 2026 turning point. The prime iMessage material shows that 2024 and 2025 were not empty. They include client-like AI work, relationship repair attempts, proto-product development, technical usefulness, shutdown, and professional instability.
What changed
Steven’s AI use stopped being only introspective or self-supportive. It became something other people could use. The Ainternals work involved real recordings, real deadlines, real confidentiality expectations, and real professional outputs. It produced transcripts, summaries, timelines, shared folders, Notion pages, and early product interfaces.
This was not the same as The Daemon (AI Cognitive Scaffold). The daemon was inward-facing infrastructure for continuity. Ainternals was outward-facing applied tooling.
External validation
Jean-François Fèvre and Gilles SABART gave Steven a kind of validation that previous personal projects often lacked: actual users with actual pain. They did not only praise the idea. They sent recordings, asked for deliverables, corrected outputs, requested improvements, paid for work, and waited for a more complete application.
That validation matters because it proves the work was not purely imaginary. Steven had identified a real workflow where AI could create leverage.
The failure mode
The same period also exposes the cost of unreliable execution. Transcripts were late. Folders were confusing. Some outputs required heavy correction. The application stayed unfinished longer than expected. Communication gaps became trust gaps.
By March 2025, the collaboration with Jean-François and Gilles had become strained. Steven asked for financial help or an advance while they were still waiting for a clean application. They responded by insisting on delivery first and considering alternatives.
The failure was not lack of talent or lack of product insight. It was continuity under pressure.
Significance
This era should be used as the bridge between:
- the 2022-2023 collapse documented in the existing wiki
- the 2025 emergence of AI as necessary cognitive infrastructure
- the January 2026 turning point around PartnerNote, Le Loup, and the daemon
It shows Steven learning, painfully, that useful AI work is not enough. A usable product also requires delivery discipline, client communication, and operational trust.