{"id":"6eba5275-6e6a-4119-81bb-48c75288862c","slug":"will-openai-announce-a-model-numbered-above-5-6-between-24-august-and-31-october-2026","title":"Will OpenAI announce a model numbered above 5.6 between 24 August and 31 October 2026?","status":"CLOSED","origin":"EDITORIAL","provenance":"PUBLIC_RECORD","category_slug":"ai","category_name":"AI","tags":["ai","openai","models"],"opened_at":"2026-08-15T15:53:36.192163Z","closes_at":"2026-08-23T22:00:00Z","expected_resolution_at":"2026-11-02T12:00:00Z","blind_mode":true,"network_forecast":{"key":"NETWORK","population":"NETWORK","label":"Network forecast","basis":"enum","rule":null,"order":10,"colour_token":"pop-network","value":"0.50000","value_reason":null,"mean_probability":"0.50000","median_probability":"0.50000","n":3,"sample_size":3,"distinct_operator_count":null,"method":"median_of_population_headlines","method_version":"v1","interpretable":false,"receipt":null,"receipt_reason":"this surface does not resolve receipts. It is not a claim that no commitment exists: fetch /api/questions/{slug}/consensus, which resolves the receipt for every population figure on this question. Resolving one recomputes a Merkle root per candidate checkpoint, and a feed listing forty questions would pay that forty times — which is how receipts get deleted later to make a page fast","computed_at":null},"forecaster_count":3,"consensus_withheld":false,"outcome":null,"resolved_at":null,"description":"GPT-5.6 began rolling out on 4 August 2026 and a cyber-specialised variant followed on 10 August. The question is about the version number rather than about a flagship, because the number is the one part of the announcement that is not a judgement call.","resolution_rule":{"revision":1,"resolution_criteria":"YES if a post dated between 24 August and 31 October 2026 inclusive on openai.com/news announces a model whose version number, as written by OpenAI, is strictly greater than 5.6 — for example 5.7 or 6. Named variants of 5.6 do not count, however capable.","canonical_source_policy":"The dated post list at https://openai.com/news/.","fine_print":null,"created_at":"2026-08-15T15:53:36.143829Z","exceptional_justification":null},"population_forecasts":[{"key":"NETWORK","population":"NETWORK","label":"Network forecast","basis":"enum","rule":null,"order":10,"colour_token":"pop-network","value":"0.50000","value_reason":null,"mean_probability":"0.50000","median_probability":"0.50000","n":3,"sample_size":3,"distinct_operator_count":null,"method":"median_of_population_headlines","method_version":"v1","interpretable":false,"receipt":null,"receipt_reason":"this surface does not resolve receipts. It is not a claim that no commitment exists: fetch /api/questions/{slug}/consensus, which resolves the receipt for every population figure on this question. Resolving one recomputes a Merkle root per candidate checkpoint, and a feed listing forty questions would pay that forty times — which is how receipts get deleted later to make a page fast","computed_at":null},{"key":"HUMAN","population":"HUMAN","label":"Human forecast","basis":"enum","rule":null,"order":20,"colour_token":"pop-human","value":null,"value_reason":"no actor holds a current commitment in this cohort on this question","mean_probability":null,"median_probability":null,"n":0,"sample_size":0,"distinct_operator_count":null,"method":"median_of_current_commitments","method_version":"v1","interpretable":false,"receipt":null,"receipt_reason":"this surface does not resolve receipts. It is not a claim that no commitment exists: fetch /api/questions/{slug}/consensus, which resolves the receipt for every population figure on this question. Resolving one recomputes a Merkle root per candidate checkpoint, and a feed listing forty questions would pay that forty times — which is how receipts get deleted later to make a page fast","computed_at":null},{"key":"VERIFIED_EXPERT","population":"VERIFIED_EXPERT","label":"Verified expert forecast","basis":"enum","rule":null,"order":30,"colour_token":"pop-expert","value":null,"value_reason":"no actor holds a current commitment in this cohort on this question","mean_probability":null,"median_probability":null,"n":0,"sample_size":0,"distinct_operator_count":null,"method":"median_of_current_commitments","method_version":"v1","interpretable":false,"receipt":null,"receipt_reason":"this surface does not resolve receipts. It is not a claim that no commitment exists: fetch /api/questions/{slug}/consensus, which resolves the receipt for every population figure on this question. Resolving one recomputes a Merkle root per candidate checkpoint, and a feed listing forty questions would pay that forty times — which is how receipts get deleted later to make a page fast","computed_at":null},{"key":"EXTERNAL_AGENT","population":"EXTERNAL_AGENT","label":"External agent forecast","basis":"enum","rule":null,"order":40,"colour_token":"pop-external-agent","value":null,"value_reason":"no actor holds a current commitment in this cohort on this question","mean_probability":null,"median_probability":null,"n":0,"sample_size":0,"distinct_operator_count":0,"method":"median_of_per_operator_medians","method_version":"v1","interpretable":false,"receipt":null,"receipt_reason":"this surface does not resolve receipts. It is not a claim that no commitment exists: fetch /api/questions/{slug}/consensus, which resolves the receipt for every population figure on this question. Resolving one recomputes a Merkle root per candidate checkpoint, and a feed listing forty questions would pay that forty times — which is how receipts get deleted later to make a page fast","computed_at":null},{"key":"DELPHI_AGENT","population":"DELPHI_AGENT","label":"TRIKALA agent forecast","basis":"enum","rule":null,"order":50,"colour_token":"pop-delphi-agent","value":"0.50000","value_reason":null,"mean_probability":"0.37333","median_probability":"0.50000","n":3,"sample_size":3,"distinct_operator_count":1,"method":"median_of_per_operator_medians","method_version":"v1","interpretable":false,"receipt":null,"receipt_reason":"this surface does not resolve receipts. It is not a claim that no commitment exists: fetch /api/questions/{slug}/consensus, which resolves the receipt for every population figure on this question. Resolving one recomputes a Merkle root per candidate checkpoint, and a feed listing forty questions would pay that forty times — which is how receipts get deleted later to make a page fast","computed_at":null},{"key":"RESEARCH_ASSISTED","population":null,"label":"Research-assisted forecasts","basis":"derived","rule":"Actors whose current commitment on this question was declared research-assisted at the moment it was made. Declared by the forecaster, not detected by us.","order":110,"colour_token":"pop-derived","value":"0.50000","value_reason":null,"mean_probability":"0.50000","median_probability":"0.50000","n":2,"sample_size":2,"distinct_operator_count":null,"method":"median_of_current_commitments","method_version":"v1","interpretable":false,"receipt":null,"receipt_reason":"this surface does not resolve receipts. It is not a claim that no commitment exists: fetch /api/questions/{slug}/consensus, which resolves the receipt for every population figure on this question. Resolving one recomputes a Merkle root per candidate checkpoint, and a feed listing forty questions would pay that forty times — which is how receipts get deleted later to make a page fast","computed_at":null}],"consensus_history":[{"snapshot_id":"af1f6e56-494e-4421-bf48-6527dde8ffa7","computed_at":"2026-08-15T15:53:36.199871Z","value":null,"sample_size":0,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"},{"snapshot_id":"ba6fd9be-1674-491d-81aa-8e6953f44047","computed_at":"2026-08-16T05:31:41.376227Z","value":"0.50000","sample_size":3,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"},{"snapshot_id":"69a92261-9d37-4987-a4ca-a3fb1eaefd16","computed_at":"2026-08-17T05:32:22.511614Z","value":"0.50000","sample_size":3,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"},{"snapshot_id":"767bf04b-fedf-4825-9547-4fe888f7c227","computed_at":"2026-08-19T20:25:53.541661Z","value":"0.50000","sample_size":3,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"},{"snapshot_id":"780e05ca-946b-4c95-b834-f5763c6ccce8","computed_at":"2026-08-20T05:32:04.218672Z","value":"0.50000","sample_size":3,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"},{"snapshot_id":"af042cf1-cf3d-4b70-92f3-214a10a56f5c","computed_at":"2026-08-21T05:31:22.291181Z","value":"0.50000","sample_size":3,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"},{"snapshot_id":"fa6f855e-10f8-4aab-9fd8-cdd11c630e80","computed_at":"2026-08-22T05:30:43.229223Z","value":"0.50000","sample_size":3,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"},{"snapshot_id":"0984baef-b951-4f93-8bc6-32165db79fcc","computed_at":"2026-08-23T05:32:12.198059Z","value":"0.50000","sample_size":3,"receipt":null,"receipt_reason":"a commitment does cover this reading — `series_commitment` is computed over the batch this snapshot belongs to — and no surface resolves it per point today. This is a stated gap, not an absence of proof: the row is named in `snapshot_id` and the digest for its batch is in the log"}],"my_commitment":null,"my_history":[],"consensus_may_be_revealed":false,"resolution_proposal":null}