pragma.vision Technology observatory & strategic foresight

Verification register AI & Agents

Readiness verdict

GPT-4

A dated reading of what is claimed, reported, and independently verified in the current evidence.

As of
2026-08-20
Revision
1
Method
v1.3.0

Current reading

Readiness band and full integer triple

AI-assisted assembly · derived results

Readiness band

Emerging

Primary summary from verified readiness

Confidence
58% · fresh
Computed at
2026-08-20T10:56:14.973133+00:00
Claimed
20

Public ambition and stated capability

Reported
18

Observed practitioner reporting

Verified
15

Independently supported evidence

Gap
+5

Claimed leads verified

Evidence strength Growing

Decision

What the current evidence supports

Human editorial judgment · 2026-08-20

Track; not yet

Why
GPT-4 is the historical root of a lineage pv already actively tracks (GPT-4o), but OpenAI's own current deprecation schedule confirms the original model itself is legacy: its ChatGPT-facing family is fully retired and its API alias shuts down 2026-10-23, so there is no forward case for new adoption of GPT-4 proper.
Next
No new direct integration against the legacy gpt-4/gpt-4-0613 endpoint; pv already runs on the downstream GPT-4o/GPT-5.x lineage tracked elsewhere in the register, so audit for any residual internal or partner calls still pinned to gpt-4 and migrate them off before the 2026-10-23 API shutdown.

Constraints

Blockers

No named blocker is present in the current public projection.

Evidence summary

Derived counts

AI-assisted assembly

Total
5
Tier 1
0
Tier 2
3
Tier 3
2
Supports
2
Contradicts
2
Context
1
Latest observed
2026-08-19

Counts and dates only. Raw signals, private excerpts, trust records, and internal corpus material are not published here.

Publication record

Revisions

Initial public reading

  1. 2026-08-20 Reading moved from emerging to emerging.

Your opinion

Tell us anything.

What works, what doesn't, what's missing — especially about our watches, lenses, and the register itself. Anonymous is fine; leave an email if you'd like a reply.