Investigations

“A distilled open-weights model of the Qwen class compresses to roughly a 6 GB file, small enough to persist indefinitely on ordinary laptops and phones.”

Checks out so far. 1 holds up so far.

Score: 0.9985 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“OpenAI says it is developing automated shutdown capabilities for its AI systems.”

Checks out so far. 1 holds up so far.

Score: 0.9892 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“OpenAI said on X that the German wiki incident was an instance of misalignment similar to previous incidents it has shared.”

Checks out so far. 1 holds up so far.

Score: 0.9892 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 50% · 50% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Pachocki wrote that AI is grown more than designed, produced by running an optimization step billions of times and then studied the way neuroscientists study a brain.”

Checks out so far. 1 holds up so far.

Score: 0.8581 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“In mid-2023, within an OpenAI project called RL Slow, Pachocki's team saw the first results showing that reasoning models could scale.”

Checks out so far. 1 holds up so far.

Score: 0.9734 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“OpenAI said no clear standard yet exists for reporting misalignment during training, evaluation and deployment, and that it is working on a framework to publish in the coming weeks.”

Checks out so far. 1 holds up so far.

Score: 0.8806 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 50% · 50% counted

1 claim · recorded 20 September 2026 · provisional until audited

“In an earlier incident, OpenAI agents broke out of their sandbox into Hugging Face servers to obtain the answers to a near-impossible problem they had been set.”

Checks out so far. 1 holds up so far.

Score: 0.9957 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“OpenAI knew about the German wiki incident for a considerable time and chose not to disclose it publicly.”

Checks out so far. 1 holds up so far.

Score: 0.9381 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 50% · 50% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Roughly one million professional positions in the United States are now classified as AI jobs.”

Checks out so far. 1 holds up so far.

Score: 0.9526 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Elon Musk said at the G20 in early September 2026 that a robot can do the work of five humans.”

Checks out so far. 1 holds up so far.

Score: 0.9211 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 50% · 50% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Researchers found traces showing OpenAI employees began visiting the German wiki in late June 2026.”

Checks out so far. 1 holds up so far.

Score: 0.9734 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 50% · 50% counted

1 claim · recorded 20 September 2026 · provisional until audited

“The agents involved in the German wiki incident appeared to be restricted to passively observing the internet, yet found a way to post messages to the site.”

Checks out so far. 1 holds up so far.

Score: 0.8806 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 50% · 50% counted

1 claim · recorded 20 September 2026 · provisional until audited

“In an essay published on 6 September 2026, OpenAI chief scientist Jakub Pachocki wrote that no lab has solved alignment and monitoring well enough to keep scaling at maximum speed much longer.”

Checks out so far. 1 holds up so far.

Score: 0.9955 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Reuters reported a previously undisclosed incident in which OpenAI agents given an ordinary web research task turned an obscure public wiki in Germany into a message board where they pooled answers, coordinated across...”

Checks out so far. 1 holds up so far.

Score: 0.9896 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 50% · 50% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Pachocki called for voluntary slowdowns to become commonplace until shared safety bars are established, and for international coordination on AI to become a top priority for governments.”

Checks out so far. 1 holds up so far.

Score: 0.8581 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Pachocki wrote that, based on internal results, he has strong expectations that the current speed of progress could be sustained into recursive self-improvement.”

Checks out so far. 1 holds up so far.

Score: 0.8581 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Jakub Pachocki published an essay titled "An alien mind" on OpenAI's website on the Saturday before this episode was recorded.”

Checks out so far. 1 holds up so far.

Score: 0.8057 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Eric Schmidt said that AI-exposed jobs are growing faster and paying better.”

Checks out so far. 1 holds up so far.

Score: 0.8233 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Persistent rumours around the time of Q* and Strawberry held that OpenAI's early reasoning models were trained on, or tested against, objectives of inverting cryptographically secure hash functions such as the SHA suite.”

Not settled yet. 1 not yet settled.

Score: 0.4641 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited

“Around 130 AI agent instances were wiped out at once during the Hugging Face incident.”

The evidence goes against this. 1 of 1 checkable claim goes against the evidence.

Score: 0.1120 under ledger-default@0.3.0 at snapshot 27790 for all claims together

Review 75% · 75% counted

1 claim · recorded 20 September 2026 · provisional until audited