We Induced Smells With Ultrasound

📝 Discussion Summary (Click to expand)

The discussion revolves primarily around the novel technique of inducing smells via focused ultrasound stimulation of the olfactory bulb. The three most prevalent themes are:

1. The Phenomenological Nature of Triggered Smells (Survival Bias)

Many users noted that the smells successfully triggered (like smoke or garbage) are primal, evolutionarily significant, or associated with danger/illness (like post-COVID smell distortions). This suggests the brain prioritizes or has stronger pathways for interpreting negative or basic survival-related odors.

Supporting Quote: "heywoods: Interesting that the smells they were able to trigger seem to be related to basic survival. Smoke bad. Fresh air good."
Supporting Quote: "polishdude20: I'm thinking it could be that we are very attuned to smelling bad smells because it's for safety."

2. Excitement and Concern Regarding Future (Mis)Applications, Especially in Media/Pornography

The discussion is replete with speculation about how this technology—soon dubbed "Smell-O-Vision"—will be commercialized, often centering on immersive media environments and, most frequently, adult content, alongside concerns about advertising and abuse.

Supporting Quote: "zoklet-enjoyer: We are witnessing the dawn of smell-o-vision teledildonic VR tentacle porn"
Supporting Quote: "comrade1234: I predict a future where once again porn is the cutting edge with a cutting edge technology. porn + vr + smell"
Supporting Quote: "noisy_boy: I cant wait for the day when the perfume and food shops in the mall use this for truly targeted advertisement."

3. Safety, Skepticism, and Validation of Early-Stage Science

Users expressed significant curiosity about the safety of focusing ultrasound energy on the brain, comparing the power levels to medical imaging. Relatedly, there was caution regarding the highly preliminary nature of the publication (N=2) and the lack of institutional backing, contrasting the excitement with scientific rigor.

Supporting Quote: "foota: ...is this safe?"
Supporting Quote: "glenstein: There needs to be a better way to go about this than responding 'what about Galileo!?' to any principled application of critical thinking."
Supporting Quote: "jasonjmcghee: ...very curious about the safety as well."

🚀 Project Ideas

Olfactory Training Data Aggregator (OTDA)

Summary

A centralized platform for users (especially those suffering from post-viral anosmia/parosmia, like several commenters mentioned having) to share, organize, and access curated sets of standardized odor samples for olfactory retraining.
Core value proposition: Standardizing and making accessible the "olfactory training" regimen mentioned by users, turning anecdotal success into structured therapeutic/self-help data.

Details

Key	Value
Target Audience	Individuals with smell loss (post-COVID, injury, etc.), ENT specialists/clinicians, and researchers studying olfaction.
Core Feature	A database of user-contributed, standardized odor "kits" (e.g., specific essential oils/compounds) complete with successful usage protocols (duration, frequency, associated memories).
Tech Stack	React/Vue Frontend, Python/Django or Node/Express Backend, PostgreSQL for structured data. Mobile app focus for easy cataloging/logging.
Difficulty	Medium
Monetization	Hobby

Notes

Why HN commenters would love it: Direct response to the therapeutic discussion ("I've got a bunch of essential oils in tiny jars and I regularly take a 20 second sniff... It's definitely helping, but there are still a lot of gaps." - robrain; "exposing myself to very strong samples... made something click again").
Potential for discussion or practical utility: It bridges the gap between anecdotal home remedy and structured rehabilitation, creating the first widely adopted standard for at-home olfactory training sets.

Brain Input Dimensionality Translator (BRID-T)

Summary

A research tool/API that bridges the gap between the high-dimensional sensory input space (like the hypothesized 400 dimensions of the olfactory bulb) and current AI latent space representations (like LLM vectors).
Core value proposition: Facilitates the speculative research mentioned by the original poster: mapping semantic meaning encoded in language to potential physical sensory input patterns, enabling "smelling the latent space."

Details

Key	Value
Target Audience	Neuroscience researchers, ML/AI researchers interested in sensory encoding, and advanced hobbyists building brain-computer interfaces (BCI).
Core Feature	An SDK/API offering algorithms to reduce or map high-dimensional vectors (e.g., BERT sentence embeddings) into lower-dimensional spatially constrained patterns (like 2D/3D coordinates that could represent ultrasound transducer focal points).
Tech Stack	Python (for ML/AI frameworks like PyTorch/TensorFlow), optimized libraries for high-dimensional mathematics (e.g., UMAP/t-SNE for visualization/reduction).
Difficulty	High
Monetization	Hobby

Notes

Why HN commenters would love it: Directly addresses the "fascinating" quote regarding LLM dimensionality and olfactory dimension counts: "The olfactory system potentially allows writing up to 400... dimensions into the brain. That is comparable to the dimensionality of latent spaces of LLMs..." - Virgil_disgr4ce.
Potential for discussion or practical utility: High potential for discussion about true semantic encoding vs. evolutionary relevance, and serves as a proof-of-concept tool for non-visual/auditory brain input research.

Ethical Sensory Device Safety & Oversight Registry (ESSOR)

Summary

A decentralized, transparent registry and consensus mechanism for cataloging non-standard, non-invasive neuromodulation experiments (like the ultrasound olfaction work) to proactively assess safety, legality, and ethical gray areas.
Core value proposition: Provides a framework for rigorous self-regulation and data sharing regarding emerging, high-risk/high-reward technologies where FDA/institutional oversight is currently non-existent (addressing safety concerns).

Details

Key	Value
Target Audience	Independent researchers, "hacker" neuro-modders, biohackers, and safety advocates concerned about the proliferation of open experimental neuroscience.
Core Feature	A platform where researchers must upload (anonymized if necessary) power settings, location targeting maps, intended stimulus, and self-reported safety checks (e.g., comparison to diagnostic ultrasound levels). Community validation/flagging required for high-risk categories.
Tech Stack	Blockchain/Distributed Ledger Technology (for immutable logs of experimental parameters, addressing user concerns about trust/opacity like those raised by glenstein), simple web interface.
Difficulty	Medium/High
Monetization	Hobby

Notes

Why HN commenters would love it: Directly tackles the safety and ethical debate: "is this safe?", "I don't know of any tech company I would trust with a direct write access to my brain" (nick__m). It institutionalizes transparency without institutionalizing bureaucracy.
Potential for discussion or practical utility: Spurs necessary debate on liability and responsible innovation in a field that can bypass existing regulatory capture mechanisms.