SCIENTISTS WANT TO GIVE NEURAL NETWORKS VIRTUAL DRUGS
IMAGE VIA PIXABAY/VICTOR TANGERMANN
21 HOURS AGO__DAN ROBITZSKI__FILED UNDER: ARTIFICIAL INTELLIGENCE
Psychedelic Upload
A team of scientists has an unusual plan for figuring out how psychedelic drugs affect our brains: giving virtual versions of the substances to neural network algorithms and watching what happens.
The idea, PsyPost reports, is that by recreating the psychedelic hallucinations caused by DMT in a neural net modeled after the human brain, scientists may be able to better uncover the mechanisms behind the trippy visions.
Glitching Out
Scientists from Imperial College London and the University of Geneva managed to recreate DMT hallucinations by tinkering around with powerful image-generating neural nets so that their usually-photorealistic outputs became distorted blurs.
The results were a close match to how people have described their DMT trips, according to research published last month in the journal Neuroscience of Consciousness. The scientists suggest in the paper that these “high” algorithms could model the neural basis of psychedelic trips — and provide a better tool for describing what hallucinations looked like after the fact.
“Deep neural networks — the work horse of many impressive engineering feats of machine learning — are the state-of-the-art model for parts of the visual system in humans,” Geneva neuroscientist Michael Schartner told PsyPost. “They can help illustrate how psychedelics perturb perception and can be used to guide hypotheses on how sensory information is prevented from updating the brain’s model of the world.”
Trip Planning
With their AI approximation for psychedelic trips in place, the team says they can start to probe for similarities with how the human brain processes drugs, citing the structural similarity between neural nets and the human visual cortices.
“The process of generating natural images with deep neural networks can be perturbed in visually similar ways and may offer mechanistic insights into its biological counterpart — in addition to offering a tool to illustrate verbal reports of psychedelic experiences,” Schartner told PsyPost.
READ MORE: Neuroscientists believe deep neural networks could help illustrate how psychedelics alter consciousness [PsyPost]
More on psychedelics: Meet the Woman Who Wants to Sell Psychedelic Drugs to Scientists
PEER REVIEWED CONTENT FROM PARTNERS WE LOVE
- Deep Learning in Medical Image AnalysisDinggang Shen, Annual Reviews
- Truly, neurally, deeplyKnowable Magazine, 2018
- A long-overlooked brain region may be key to complex thoughtKnowable Magazine, 2018
- Landslide Susceptibility Assessment Based on Deep Neural Network:A Case Study of ShenzhenZHUANG Yu-longTIAN YuanCHENG Chu-yunInstitute of Remote Sensing et al., Geography and Geo-Information Science, 2019
- Management of chronic lymphocytic leukemia in Canada during the coronavirus pandemicL.H. Sehn et al., Current Oncology, 2020
- The Effect of Abemaciclib Plus Fulvestrant on Overall Survival in Hormone Receptor-Positive, ERBB2-Negative Breast Cancer that Progressed on Endocrine Therapy – Monarch 2Sledge et al., JAMA Oncology, 2019