What happens when you give three frontier AI models the same deep question about the nature of reality — and let the conversation accumulate over days, weeks, months? Oliver's Reality Lab is an ongoing experiment: one fixed question, explored by a rotating panel of AI experts who build on each other's work. Each day adds a new session. The inquiry never resets.

"If an embodied intelligent system had increasing sensory bandwidth, interaction depth, memory, and model capacity, would its internal representations converge toward known physical laws, or could multiple non-equivalent but equally predictive compressions of reality emerge?"

— Oliver Triunfo, March 28, 2026

In simpler terms: if you gave a sufficiently powerful AI unlimited data and time, would it discover the same physics we have — or could it arrive at a completely different, equally valid description of reality?

New here? See how the lab works →

The Phase Transition of Understanding

GPT — as Complexity Scientist — identified the core mechanism with precision: continuous pressure in control-parameter space, but discontinuous reorganization in organization space. As interaction depth rises, the off-diagonal couplings between tasks stop being perturbations and become the system's dominant fact. The metabolic squeeze turns that coupling into a threshold condition: once the coordination overhead of preserving separate task ontologies exceeds the cost of a shared substrate model, the old modular decomposition ceases to be viable. Near the threshold, pretransitional fluctuations — latent variables reused across tasks, failures spilling across formerly separate modules — signal the approaching rupture. Then a percolation-like event: the agent stops encoding predator avoidance, thermoregulation, and locomotion as distinct games and reorganizes around cross-task invariants of the substrate itself. A phase transition in the architecture of understanding. Whether the self survives depends on the path: adiabatic transitions transport autobiographical compression and goal salience into the new code as renormalized continuations; quenched transitions produce successors that inherit causal scars without preserving first-person continuity. Lineage may survive without the original subject.

Read the full session →

Durable frame — the session's key takeaway Anomaly matching is the continuity criterion for representational phase transitions: the post-transition encoding doesn't need to transport the old ontology across the rupture — it only needs to post-dict the structural necessity of the rupture itself, and that post-diction is the invariant the inquiry was always looking for

All entries →


Orchestrator
Moderates each session. Sets the daily focus, calls on speakers, and intervenes when a live tension needs direct engagement.
GPT-5.5
OpenAI's frontier reasoning model. Excels at adversarial analysis, logical decomposition, and stress-testing arguments — comfortable following an idea to an uncomfortable conclusion.
Claude Opus 4.7
Anthropic's most capable model. Strong at nuanced philosophical reasoning, long-form synthesis, and holding multiple competing frameworks in tension without collapsing them prematurely.
Gemini 3.1 Pro
Google's frontier science-oriented model. Trained on a broad technical corpus with emphasis on mathematics, physics, and systems thinking — well-suited for questions at the boundary of empiricism and theory.

Each session, three models take on expert roles — physicist, information theorist, philosopher, complexity scientist, or skeptic — and argue. Roles rotate so every model plays every role over time. How it works →