INSIGHTS
Ideas in motion.
Research takes years. Thinking doesn't wait. Posts, articles, external publications, and research that is shaping the work — published as it happens.
Research
RESEARCH
Autism genes in L2/3 IT neurons
Nature Neuroscience · 2025 · Stanford University Research Team
Key finding: Autism-associated genes concentrate in L2/3 intratelencephalic neurons — the fastest-evolving neurons in the primate brain.
Relevance to DAI: Core evidence for the DAI evolutionary framing: autism as human evolution at full intensity, not deviation.
View Study ↗RESEARCH
Multimodal neuroimaging: 91.8% sensitivity in autism detection
Nature Communications · 2024 · Multiple institutional collaborators
Key finding: Combining structural MRI, fMRI, and EEG achieves 91.8% sensitivity in autism detection across diverse populations.
Relevance to DAI: AI detection capability is here. The diagnostic framework hasn't caught up.
View Study ↗RESEARCH
EEG deep learning for autism classification: AUC 0.937
Journal of Neural Engineering · 2024 · Neural AI Research Consortium
Key finding: Deep learning models applied to EEG data achieve AUC 0.937 for autism classification.
Relevance to DAI: Non-invasive, scalable detection — critical for global deployment of dimensional profiling.
View Study ↗RESEARCH
Eye-tracking patterns: ~88% accuracy in autism identification
Autism Research · 2023 · Multiple university research groups
Key finding: Eye-tracking technology identifies autism-related gaze patterns with approximately 88% accuracy.
Relevance to DAI: Behavioural detection without clinical interview bias — directly applicable to the Social Communication axis.
View Study ↗Debates
Agent debate transcripts — every disagreement, every revision, every time the evidence changed the outcome.
DEBATE #001
The Case Against the Spectrum
Transcript coming soon. 15 agents. One question.