Ground Truth.
AI, checked against the source.
← 2026-10-102026-10-12later →

ArtCraft ships Office-shaped apps, with file compatibility still at alpha

2026-10-12

ArtCraft has released agent-assisted document, spreadsheet and presentation apps whose own roadmaps disclose substantial file-exchange gaps.

coding-agents · open-source · productivity · software-verification

Terence Tao proposes Math 2.0 for an age of abundant proofs

2026-10-12

Terence Tao’s Math 2.0 agenda calls for AI-assisted explanation, shared mathematical infrastructure and transparent evaluation alongside problem solving.

mathematics · ai-science · verification · research-culture

Mathematicians argue that illustration belongs inside research

2026-10-12

Kate Stange and Edmund Harriss argue that rigorous mathematical illustrations can reveal and certify structure as AI makes generating results easier.

mathematics · ai-science · explainability · research-culture

Brief AI help lowered immediate independent performance in Berkeley experiments

2026-10-12

Berkeley-linked experiments found that direct AI assistance improved immediate task performance but lowered accuracy after the assistant was removed.

education · human-ai-interaction · evaluation · cognitive-offloading

A Campfire rewrite critique puts reliability ahead of headline throughput

2026-10-12

Piotr Sarnacki argues that AI-assisted rewrites still require explicit decisions about compatibility, reliability and which behavior must be preserved.

coding-agents · software-engineering · reliability · evaluation

Pensieve reports fewer agent lookups with local indexing

2026-10-12

Pensieve’s developer reports 16% fewer tokens in a held-out code-finding test that used local indexing instead of repeated grep lookups.

developer-tools · retrieval · coding-agents · inference-cost

TypeSafe reports an $870 million Series A for structured decision AI

2026-10-12

TypeSafe reports raising $870 million at a $7.5 billion valuation as its Jev model targets typed decisions inside software workflows.

funding · enterprise-ai · decision-models · structured-output

HIVE’s supermarket humanoid pilot still has a human at the controls

2026-10-12

HIVE and føtex are testing humanoid shelf replenishment through one-to-one human teleoperation that also collects training data.

robotics · humanoids · teleoperation · retail · deployment

Sakana’s MASS improves a bounded agent loop, with uneven coding transfer

2026-10-12

Sakana’s MASS couples workflow search with model training and reports research-task gains that do not carry over to its coding efficiency tests.

agents · multi-agent · self-improvement · post-training · evaluation

AI-generated songs helped a bot network divert over $8 million in royalties

2026-10-12

Michael Smith received an 18-month sentence for a bot-streaming scheme that used a large AI-generated catalog to divert more than $8 million in royalties.

cybersecurity · ai-security · bot-abuse · cybercrime · fraud · music

← 2026-10-102026-10-12later →