ScienceAI

Philip Shiu | Towards Embodied, Whole Brain Emulations

Foresight Institute
Foresight Institute•May 15, 2026

Why It Matters

Demonstrating reliable structure‑to‑function predictions validates connectome‑based brain emulation, accelerating neurotechnology and biologically grounded AI development.

Key Takeaways

  • •Predicting neural activity from connectome accelerates brain understanding
  • •Model predicts sugar-induced proboscis extension via specific neurons
  • •Optogenetic tests validate computational predictions across 100+ cell types
  • •Inhibitory bitter neurons suppress feeding behavior in the model
  • •Cross-modal simulations reveal unexpected aversive role for IR94e neurons

Summary

Philip Shiu presented his work at Eon on using detailed connectome data to predict neural activity and ultimately build embodied whole‑brain emulations. The core goal is to infer firing patterns from the static wiring diagram of neurons, first demonstrated in the fruit‑fly Drosophila brain and now extended to cultured mammalian tissue.

The team constructed a point‑neuron model where synaptic weights and signs are derived directly from electron‑microscopy‑derived connectivity and neurotransmitter predictions. By stimulating sugar‑sensing gustatory receptor neurons, the model correctly forecasted activation of the contralateral proboscis‑motor neuron 9, a finding confirmed with calcium imaging and optogenetic activation across more than a hundred genetically defined cell types.

Key examples include silencing the Rattle neuron, which the model predicted would diminish proboscis extension—a result reproduced behaviorally in flies. Moreover, when integrating bitter and sugar inputs, the simulation captured the expected suppression of feeding, and it unexpectedly classified the IR94e taste neurons as aversive, a hypothesis later supported experimentally.

These results suggest that a sufficiently detailed connectome can generate accurate, testable predictions of circuit function, opening pathways for rapid mental‑health drug screening, brain‑computer interfaces, and safer AI systems that emulate biological neural dynamics.

Original Description

Subscribe to Foresight Neurotechnology Seminar Group: https://foresight.org/seminar/neurotechnology-seminar-group/
Supporting researchers, entrepreneurs, and funders in advancing neurotechnology progress and applications in brain-computer interfaces, whole brain emulations, and neuro-inspired AI safety.
Philip Shiu | Towards Embodied, Whole Brain Emulations
Abstract: Embodied, whole-brain emulations of animals, based on the connectome of their brain, would advance AI, accelerate mental health treatments, and may be a path towards aligned AI and furthering human flourishing. In this talk, I will discuss efforts to create embodied models of brains. Using the connectome of a fruit fly, I and others used simple biophysical neural models, and showed that this produced reasonably accurate models of neural activity. Next, we’ve embodied this brain model. At the end of this talk, I’ll discuss paths towards embodied whole-brain emulations in mammals.
Bio: Philip Shiu is a neuroscientist, and Head of Engineering at Eon Systems. During his postdoctoral fellowship at the University of California, Berkeley, he mapped the Drosophila taste system, and created a whole-brain simulation of the fly brain. He received his PhD from Harvard University, where he studied gene silencing in C. elegans. He is recipient of a NIH F32 postdoctoral fellowship, a NSF GRFP fellowship, and a Foresight Fellowship.
══════════════════════════════════════
About The Foresight Institute
The Foresight Institute is a research organization and non-profit that supports the beneficial development of high-impact technologies. Since our founding in 1986 on a vision of guiding powerful technologies, we have continued to evolve into a many-armed organization that focuses on several fields of science and technology that are too ambitious for legacy institutions to support. From molecular nanotechnology, to brain-computer interfaces, space exploration, cryptocommerce, and AI, Foresight gathers leading minds to advance research and accelerate progress toward flourishing futures.
We are entirely funded by your donations. If you enjoy what we do please consider donating through our donation page: https://foresight.org/donate/
Visit https://foresight.org, subscribe to our channel for more videos or join us here:
══════════════════════════════════════
Timecodes
00:00 Brain Emulation Goals
02:20 Why Structure Matters
03:10 Building The Fly Brain
08:30 Sugar To Action
11:40 Predicting Neural Responses
15:00 Necessary Feeding Circuits
17:00 Optogenetic Validation
19:10 Integrating Taste Signals
23:30 Predicting New Biology
24:40 Q&A: A Connectome Completeness
27:00 Q&A: Limits Of Behavior Modeling
30:40 Q&A: Mammalian Brain Emulation
34:20 Q&A: Micro-Island Neural Systems
38:30 Q&A: Structure To Function Scaling
41:50 Q&A: Cell Types And Generalization
47:40 Q&A: Mouse Brain Timelines

Comments

Want to join the conversation?

Loading comments...