# KW-001 · AI as K6 Candidate

> Public research brief  
> Snapshot date: 2026-07-06  
> Status: Watchlist / candidate boundary  
> Related site modules: Long Wave Atlas, Current Regime, Watchlist

## Core Question

Is AI the beginning of a new long-wave diffusion phase, or is it primarily a late-K5 financial and infrastructure boom?

Kairos currently treats AI as **K6 Candidate Watch**, not confirmed K6. The candidate label exists because AI has credible ground signals, infrastructure investment, and potential general-purpose technology characteristics. It remains unconfirmed because broad diffusion, sustained productivity gains, enterprise ROI, and non-tech profit-pool migration still need evidence.

## Current Reading

The public reading is intentionally cautious:

- AI is the most important current K6 candidate, but it is not confirmed K6;
- the strongest evidence today is infrastructure buildout, hyperscaler investment, semiconductor demand, data-center expansion, and power-grid pressure;
- the weaker evidence is broad productivity diffusion across the economy;
- AI can be both a real technology platform and a late-cycle financial overbuild in different segments;
- Kairos therefore separates "technology potential" from "macro-cycle confirmation".

## Why This Belongs in Kairos

Long-wave transitions are rarely confirmed at the moment they begin. They are usually visible first as a mix of:

- technological excitement;
- infrastructure investment;
- speculative capital formation;
- uneven adoption;
- delayed productivity evidence;
- institutional and energy-system bottlenecks.

KW-001 exists to prevent two opposite errors:

- declaring AI as confirmed K6 too early;
- dismissing AI as only a bubble before diffusion evidence has had time to appear.

Kairos keeps AI in a candidate state so the judgment can be strengthened, weakened, or killed by evidence rather than narrative momentum.

## Evidence Status

Public evidence status: **watchlist / partially source-checked**.

The current public evidence base includes three broad directions:

- official business-adoption data and productivity indicators;
- academic work that offers both optimistic and cautious readings of AI productivity;
- company-level and infrastructure-level signals from capex, data centers, semiconductors, cloud demand, and power constraints.

The most important open gap is not whether AI tools exist or whether capex is real. The open gap is whether AI produces broad, durable, cross-industry productivity diffusion.

## Watch Triggers

The K6 candidate case strengthens if:

- AI adoption rises across industries, not only within information, finance, software, and large enterprises;
- productivity improvement appears across multiple quarters and multiple sectors;
- non-tech industries show verifiable profit-pool migration or margin improvement linked to AI adoption;
- AI revenue and enterprise ROI start to justify infrastructure capex;
- automation, power, data-center, semiconductor, and advanced-manufacturing constraints form a durable industrial investment cycle;
- evidence shows that AI is changing production functions, not only office workflows.

## Kill Criteria

The K6 candidate case weakens if:

- AI adoption remains shallow, concentrated, or mostly experimental;
- productivity gains fail to appear outside a few information-heavy sectors;
- AI revenue cannot cover capex, depreciation, and compute cost pressure;
- traditional industries do not show measurable margin or productivity improvement;
- infrastructure overbuild leads to falling utilization, pricing pressure, or refinancing stress;
- the main macro effect is asset-price concentration rather than broad capital formation.

## Relationship to KW-006

KW-001 asks whether AI can become a new long-wave diffusion engine.

KW-006 asks whether AI infrastructure capex is developing financing fragility.

Both can be true at the same time. A real technology platform can still create overbuilt segments, speculative financing, and weak players. Kairos therefore tracks diffusion evidence and financing risk separately.

## Rejected Overclaims

The following claims are **not** part of the current Kairos public reading:

- "AI is already confirmed K6."
- "AI productivity diffusion has already been proven across the whole economy."
- "AI is only a bubble and has no real industrial signal."
- "Capex alone proves a new long wave."
- "A financing bubble would automatically disprove all AI diffusion potential."

## Open Questions

- Does AI adoption become broad enough to affect aggregate productivity?
- Which industries outside technology show measurable ROI?
- Can AI reduce labor, energy, coordination, or design bottlenecks at scale?
- Does compute cost fall fast enough to support durable adoption?
- Will the next hardware cycle improve ROI or expose overcapacity?
- How much of the boom is productive capital formation versus financial extrapolation?

## Reading Boundary

This brief is for macro-cycle learning and evidence-gated observation. It is not investment advice, a trading signal, or a claim that K6 has been confirmed.