Remember the Day — news feed beginning 2026-09-13

Sweden voted on Sunday, September 13. Election-night reporting indicated a tight opposition lead, but the count was preliminary. Voters elect the Riksdag rather than appointing a prime minister directly.

Of 349 parliamentary seats, 310 are permanent constituency seats and 39 are adjustment seats. A party normally needs 4% nationally; a party reaching 12% in a constituency can qualify for permanent seats there. The drawing is a schematic, not a result.

The Speaker proposes a prime minister. The proposal is blocked if at least 175 members vote no; otherwise it is approved. This explains the government-formation procedure—it does not mean a government had been settled on election night.

A special collection of recent discoveries and Saturday event programmes—not ten developments that all occurred on September 12. Brain-flow measurements initially involve animals; fossil feathers offer clues rather than a complete explanation of extinction.

A fruit-fly wiring map is a research resource, not an explanation of consciousness. Bat longevity research offers leads, not a proven human anti-ageing treatment.

The running-mouse findings apply to the animals and study conditions. Possible organic chemistry in Venus’s clouds does not establish life.

The Smithsonian and Reno stories describe advertised September 12 programmes. They do not supply post-event attendance figures or confirmed outcomes.

Berlin’s industrial heritage festival runs September 12–October 4. Dublin’s Liffey Swim programme describes a 2.2-kilometre course; no winners or water-quality assurances are claimed.

On September 12, Anthropic CEO Dario Amodei committed the company to inviting an embedded external review team. The announcement describes a proposed arrangement, not reviewers already at work.

The proposed remit reaches beyond a finished demonstration into training processes and safety practices. Access would broadly resemble that of internal risk teams, subject to legal, contractual and privacy limits.

Reviewers would be able to publish key findings without company editorial control. Narrow confidentiality redactions would remain possible, but an unfavorable finding alone would not justify one. Implementation remains the test.