Discourse monitor

How the monitored publications were assessed

Discourse valence across every scored item: what the sampled sources published and how independent models read it. It is not a measure of public opinion, not a measure of the state of the world, and article volume is not agreement.

Window
30 days90 days180 days365 days
Reading
All readings Strongly adverseAdverseMixed or uncertainFavourableStrongly favourable
Topic
All topics Existential and catastrophic riskAGI and timelinesLoss of control and alignmentGovernance and regulationConcentration of powerEvaluation and measurementScaling and architectureCyber capabilityHuman agency and epistemicsSynthetic media and manipulation
Range
Mean discourse valence per day · power-concentration · last 90 days

The trend line needs at least two days of readings and there is one so far. It appears on its own once the pipeline has run across a second day. Today's readings are on the axis below.

This line is the mean of every item scored on each day. It deliberately ignores the reading filter, because a daily mean of only the strongly adverse items would move with how many were filtered out rather than with the discourse.

Discourse valence, strongly adverse
-72
Strongly adverse
4 items · range -83 to -65
Adverse readingFavourable reading
Mean across the 4 items matching this filter under power-concentration in the last 90 days.

4 items read as strongly adverse

167
discovered, all time
57
passed triage
108
rejected at triage
230
model assessments
49
multi-model panels
112
distinct claims tracked

Across the whole window

The two tables below describe every item in the last 90 days under power-concentration. They are deliberately not narrowed by the reading filter (currently strongly adverse), because a mean taken only over items already selected for their score would just restate the filter.

By topic

TopicBranchItemsMeanPosition
Existential and catastrophic risk risk 28 -56
AGI and timelines structural 21 -37
Loss of control and alignment risk 21 -66
Governance and regulation structural 19 -20
Concentration of power risk 17 -37
Evaluation and measurement structural 10 -22
Scaling and architecture structural 9 -8
Cyber capability risk 8 -56
Human agency and epistemics risk 7 -22
Synthetic media and manipulation risk 6 -63
Medicine and health benefit 5 +0
Biological and chemical uplift risk 4 -20
Scientific acceleration benefit 4 +17
Abundance and material wellbeing benefit 2 -8
Compute and infrastructure structural 2 -13
Education and access to expertise benefit 2 -37
Climate and energy structural 1 -46
Employment and displacement risk 1 -5
Productivity and growth benefit 1 -26

By source

A source's mean says how the items it published were read. It is not a rating of the source.

SourceKindTierItemsMean
The Guardian: Artificial Intelligencenews19-55
Financial Times: Artificial Intelligencenews18-53
arXiv cs.CY (Computers and Society)academic18-26
The New York Times: Technologynews17-31
BBC News: Technologynews16-36
Ars Technica: AInews15-73
OpenAIlab13-17
IEEE Spectrum: AInews13+2
MIT Technology Review: AInews12-50
arXiv cs.AIacademic12-8
The Economist: Science and Technologynews11-16
Google DeepMindlab11+7
CSET, Georgetowngovernment11-22
arXiv cs.LG (Machine Learning)academic11+4