Weizmann Scientist Noam Sobel Transmits First Scent Over IP

Essentials: How Smells Influence Our Hormones, Health & Behavior | Dr. Noam Sobel — Andrew Huberman

Andrew Huberman

Subconscious human chemosignals directly regulate hormones, suppress aggression, and predict instant friendships before any verbal interaction occurs.

AI-assisted summary of Essentials: How Smells Influence Our Hormones, Health & Behavior | Dr. Noam Sobel — original source. Check the original for full context.

Source published:

Key ideas

  • Human olfactory receptors exhibit extreme convergence: roughly 6 to 7 million sensory neurons across approximately 350 subtypes project directly through the cribriform plate into specific olfactory bulb glomeruli, bypassing the standard thalamic sensory relay. (0:20)
  • Subconscious chemical sampling drives human interaction; individuals systematically sniff their hands after handshakes, and baseline endogenous body odor similarity detected by electronic noses reliably predicts spontaneous friendship formation between strangers. (4:32)
  • Hexadecanal, a semi-volatile compound abundant on infant heads, triggers sex-divergent behavioral responses in humans, significantly dampening aggression in men while elevating it in women. (18:59)
  • Sniffing odorless emotional human tears rapidly suppresses free testosterone in adult men by approximately 14% and reduces neural activity in arousal-related brain regions. (21:51)
  • Researchers demonstrated digital scent transmission over internet protocol (IP) using gas chromatography-mass spectrometry and predictive algorithms to blend primary odorants, successfully transmitting the recognizable scent of violets. (29:48)

Detailed notes

## Background Published by Andrew Huberman on YouTube on October 29, 2026, this Huberman Lab Essentials episode features Dr. Noam Sobel, a professor of neurobiology at the Weizmann Institute of Science. The discussion explores how human olfaction operates, how subconscious chemosignals shape behavior, hormones, and relationships, and how pioneering research is working toward digitizing odors and enabling noninvasive medical diagnostics through body scent.

## TL;DR Dr. Noam Sobel outlines the anatomical architecture of the olfactory system, detailing how millions of receptors converge onto the olfactory bulb to process airborne volatile molecules. He reveals that humans constantly sample biological scents through actions like post-handshake face-touching, and demonstrates that endogenous body odor similarities reliably predict instant non-romantic friendships. The conversation also explores how chemosignals—such as hexadecanal on infants' heads, the airborne scent of fear, and odorless molecules in emotional tears—potently modulate human aggression and suppress testosterone. Finally, Sobel explains the algorithmic breakthroughs behind synthesizing olfactory metamers, broadcasting odors across IP networks, and the future potential for electronic noses in continuous health screening.

## Atmosphere & style The dialogue strikes a compelling balance between rigorous academic neurobiology and playful, irreverent curiosity, driven by Sobel's sharp wit and Huberman's enthusiastic scientific engagement. Sobel approaches human social rituals with the detached amusement of a wildlife ethologist, treating everyday handshakes, weeping, and sniffing habits as covert animal communication. Listening feels like peering into an invisible, subconscious dimension of human connection that society routinely ignores, leaving the audience with an eye-opening appreciation for the silent chemical signals governing everyday interactions.

## Bulletin - [00:20](https://www.youtube.com/watch?v=arMMF0ISRRM&t=20s) Human olfactory receptors exhibit extreme convergence: roughly 6 to 7 million sensory neurons across approximately 350 subtypes project directly through the cribriform plate into specific olfactory bulb glomeruli, bypassing the standard thalamic sensory relay. - [04:32](https://www.youtube.com/watch?v=arMMF0ISRRM&t=272s) Subconscious chemical sampling drives human interaction; individuals systematically sniff their hands after handshakes, and baseline endogenous body odor similarity detected by electronic noses reliably predicts spontaneous friendship formation between strangers. - [18:59](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1139s) Hexadecanal, a semi-volatile compound abundant on infant heads, triggers sex-divergent behavioral responses in humans, significantly dampening aggression in men while elevating it in women. - [21:51](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1311s) Sniffing odorless emotional human tears rapidly suppresses free testosterone in adult men by approximately 14% and reduces neural activity in arousal-related brain regions. - [29:48](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1788s) Researchers demonstrated digital scent transmission over internet protocol (IP) using gas chromatography-mass spectrometry and predictive algorithms to blend primary odorants, successfully transmitting the recognizable scent of violets.

## Key takeaways - [00:20](https://www.youtube.com/watch?v=arMMF0ISRRM&t=20s) Olfactory transduction generates bona fide action potentials that route directly through the cribriform plate to glomeruli in the olfactory bulb, representing the most extreme neural convergence in the mammalian central nervous system. - [00:20](https://www.youtube.com/watch?v=arMMF0ISRRM&t=20s) Olfactory projections bypass the obligatory thalamic relay used by other sensory modalities, branching directly from the olfactory bulb into the piriform cortex, entorhinal cortex, amygdala, and hypothalamus. - [04:32](https://www.youtube.com/watch?v=arMMF0ISRRM&t=272s) Secret video analysis and nasal airflow measurements demonstrate that humans systematically inhale from their hands following a handshake, covertly sampling the chemical profile of others and comparing it to their own baseline scent. - [10:03](https://www.youtube.com/watch?v=arMMF0ISRRM&t=603s) Same-sex 'click friends' share significantly more similar endogenous body odors than random pairs, as measured both by electronic noses equipped with metal oxide sensors and by human sniffers. - [12:35](https://www.youtube.com/watch?v=arMMF0ISRRM&t=755s) In a nonverbal round-robin mirror game, pre-existing chemical similarity measured by an electronic nose reliably predicted mutual feelings of warmth, perceived kindness, and prospective friendship among strangers. - [14:04](https://www.youtube.com/watch?v=arMMF0ISRRM&t=844s) Human fear sweat acts as a chemical contagion, triggering an immediate and measurable increase in autonomic sympathetic arousal in individuals who inhale it. - [15:48](https://www.youtube.com/watch?v=arMMF0ISRRM&t=948s) Hexadecanal, an evolutionary conserved mammalian signaling molecule, reduces aggressive retaliation in human males when tested in competitive paradigms. - [18:59](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1139s) Hexadecanal elicits opposing behavioral effects by sex—decreasing aggression in men while increasing it in women—providing an evolutionary survival advantage by pacifying paternal aggression while heightening maternal defense. - [18:59](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1139s) Infant head odor is dominated by hexadecanal, functioning as a subconscious chemosignal that manipulates parental behavioral responses. - [21:51](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1311s) Emotional tears act as a functional 'chemical blanket' that reduces male aggression, causing an approximate 14% drop in free testosterone within 20 to 30 minutes despite being consciously odorless. - [21:51](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1311s) Tear sniffing dampens neural activity under sexual arousal states within the hypothalamus and fusiform gyrus, a protective chemosensory mechanism mirrored in rodents. - [25:37](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1537s) By developing algorithms that predict perceptual distance across a multidimensional space of 4,000 molecules, researchers engineered 'olfactory metamers'—mixtures with zero overlapping chemical compounds that smell identical, such as a synthetic replica of Chanel No. 5. - [29:48](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1788s) In an international trial transmitting digital scents over IP, automated mixing of odor primaries achieved the first successful remote transmission of an odor, with 15 of 16 evaluators correctly identifying violet. - [32:34](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1954s) The long-term therapeutic frontier of digital olfaction centers on noninvasive medical diagnostics, such as bathroom electronic noses continuously screening urine and fecal volatile organic compounds (VOCs) for early metabolic disorders and cancer.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=arMMF0ISRRM&t=0s) Biology of Smell (Olfaction)

Dr. Andrew Huberman introduces this Huberman Lab Essentials episode by welcoming Dr. Noam Sobel, a neurobiology professor at the Weizmann Institute of Science, to dive into the landscape of chemosensation, olfaction, and subconscious abilities such as human pheromones. Opening the discussion, Huberman asks Sobel to outline the physical "parts list" and basic architecture required for humans to detect odors. Sobel begins by defining primary olfaction as orthonasal olfaction, which occurs when air and volatile molecules enter through the external nares of the nose via sniffing. In humans, airborne molecules travel roughly 6 to 7 centimeters up the nasal cavity to interact with a highly convoluted structure known as the olfactory epithelium.

```mermaid flowchart TD accTitle: Olfactory Transduction and Projection Pathway accDescr: Step-by-step diagram showing how airborne odorants travel through the nose and epithelium to the olfactory bulb and wider brain areas.

subgraph NasalCavity["Nasal Cavity & Epithelium"] A["Airborne Odorants (Sniffing)"] --> B["Olfactory Epithelium (~7M receptors, ~350 types)"] B --> C["Transduction (Action Potentials)"] end

subgraph OlfactoryBulbPath["Olfactory Bulb (Extreme Convergence)"] C --> D["Olfactory Nerve through Cribriform Plate"] D --> E["Glomeruli (Odor Map of Receptor Types)"] end

subgraph CentralProjections["Brain Projections"] E --> F["Primary Olfactory Cortex (Piriform & Entorhinal)"] E --> G["Subcortical Targets (Amygdala, Hypothalamus, Cerebellum)"] end ```

The olfactory epithelium houses roughly 6 to 7 million olfactory receptors, which represent approximately 350 distinct receptor subtypes in humans. Sobel emphasizes that this variety represents a significant investment of genetic code: “That means a meaningful percentage of your genome is devoted just to this, just to the kinds of olfactory receptor subtypes you have in your nose.” Transduction occurs when odor molecules dock onto their matching receptor subtypes, generating action potentials rather than graded potentials. These electrical signals are routed via the olfactory nerve directly through perforations in the cribriform plate—the thinnest bone in the human skull—to synapse in the olfactory bulb.

This projection exhibits what Sobel identifies as the most extreme case of convergence in the mammalian central nervous system: all sensory neurons expressing a specific receptor subtype converge onto just one (or two) specialized structures called glomeruli within the olfactory bulb. According to the classical textbook model, specific molecular features of odor molecules (known as odoritopes) activate distinct receptor subsets. For example, the scent of coffee might stimulate receptor types 1, 3, and 7, illuminating a corresponding spatial pattern of glomeruli to generate an odor "map." From the bulb, axons project directly to primary olfactory areas—the piriform and entorhinal cortices on the ventral surface of the temporal lobe—and branch out directly to the amygdala, the hypothalamus, and likely the cerebellum. Unlike the organized representation within the bulb, spatial mapping in downstream cortical regions remains murky and poorly understood.

- **Receptor count:** ~6–7 million olfactory sensory neurons in the human epithelium across roughly 350 receptor subtypes. - **Transduction signal:** Olfactory sensory neurons fire bona fide action potentials through the perforated skull (cribriform plate) straight into the olfactory bulb. - **Glomerular convergence:** Sensory neurons expressing the exact same receptor gene channel their axons into two specific glomeruli per bulb. - **Downstream targets:** Information bypasses the obligatory thalamic relay used by other senses, projecting directly from the bulb to the piriform cortex, entorhinal cortex, amygdala, and hypothalamus.

---

[04:32](https://www.youtube.com/watch?v=arMMF0ISRRM&t=272s) Human Smell, Social Interaction, Handshaking

While public lectures often require convincing audiences that human smell is far from degenerate, Sobel highlights that human olfaction is remarkably sensitive and profoundly drives behavior, even if society assumes otherwise. Canines openly sniff one another during social encounters, whereas human etiquette strictly limits conscious sniffing to infants or intimate romantic partners. Yet, by observing human actions through the lens of an animal ethologist, Sobel's team uncovered subtle, unacknowledged mechanisms that humans use to chemically sample strangers.

Investigating the mysterious evolutionary origin of handshaking—often claimed without strong evidence to have arisen to show an empty, weaponless hand—the Sobel lab examined whether the gesture serves as a clandestine chemical exchange. In a study led by researcher Idan Frumin, subjects were seated alone in an experiment room and secretly recorded for three minutes to establish baseline facial contact, such as resting fingers near the chin, nose, or forehead. An experimenter then entered under the guise of an administrative delay; for half the subjects, this greeting included a handshake, while for the other half it did not.

The resulting video analysis showed a striking, immediate increase in hand-to-face contact following a handshake with a stranger. To confirm that these movements were not accidental twitches, the researchers measured nasal airflow unobtrusively using precision instruments. The recordings confirmed that subjects systematically timed distinct nasal inhalations right as their hand brushed their nostrils: “And and and people not only bring their hand to their nose, they sniff it. So and this is perfectly timed. They go like this, okay? So, they're sniffing their hands.” Interestingly, participants smelled both the hand used to perform the handshake and their opposite, untouched hand, suggesting a subconscious baseline process of comparing their own chemical profile against the novel biological scent of another person.

- **Baseline face-touching:** Humans constantly touch their faces at an unexpectedly high baseline rate even while sitting alone. - **The handshake experiment:** Subjects greeted with a handshake exhibited a statistically dramatic spike in touching their nose with the shaking hand compared to non-shaking controls. - **Active sniffing:** Synchronized nasal airflow sensors proved that subjects were deliberately inhaling odorants from their skin during facial contact rather than merely resting their hands. - **Self vs. other comparison:** Participants frequently sniffed their non-shaking hand as well, suggesting a subconscious chemical calibration between self-derived scent and the other person's volatile molecules.

---

[08:31](https://www.youtube.com/watch?v=arMMF0ISRRM&t=511s) Sponsor: LMNT

Huberman pauses the interview to feature episode sponsor LMNT, an electrolyte drink mix containing sodium, potassium, and magnesium balanced without added sugar. He explains that proper cellular hydration is essential for optimal brain and muscle performance, noting that even minor dehydration impairs cognitive processing and physical endurance because electrolytes are mandatory for neuronal signaling and action potentials. Huberman notes that his personal morning routine begins with dissolving one packet of LMNT in 16 to 32 ounces of water immediately upon waking.

[09:19](https://www.youtube.com/watch?v=arMMF0ISRRM&t=559s) Sponsor: LMNT

Andrew Huberman highlights his routine of using LMNT electrolytes dissolved in water during physical exercise, especially on hot days when sweating causes substantial fluid and mineral loss. He mentions his personal favorite flavors, calling out raspberry and citrus alongside watermelon, while highlighting a limited-edition lemonade flavor that rivals his long-time favorites. He directs listeners to drinklmnt.com/huberman to claim a free sample pack containing various drink mix flavors with any purchase.

[10:03](https://www.youtube.com/watch?v=arMMF0ISRRM&t=603s) "Click" Friendships and Body Odor Similarity

Dr. Noam Sobel shares a study led by Inbal Raviv in his laboratory investigating the everyday phenomenon of "click friendships"—those rare social encounters where two people feel an instantaneous, effortless connection. While common wisdom attributes these instant bonds to shared aesthetic traits, matching backgrounds, or common interests like cheering for the same sports team, Raviv hypothesized that an olfactory mechanism might be at play. Specifically, she proposed that individuals who share similar endogenous body odors are naturally predisposed to click with one another.

To investigate this, Raviv used social media to recruit pairs of non-romantic, same-sex "click friends" from across Israel. To isolate their true biological scents from artificial fragrances, the researchers instituted a strict hygiene regimen: participants washed with odorless soap and shampoo for several weeks, slept alone in identical cotton T-shirts for two consecutive nights, and returned the garments so their body odors could be extracted. The research team first evaluated these scents using an "electronic nose"—a device equipped with metal oxide (MOX) sensors designed to register volatile airborne molecules. The electronic nose revealed that genuine click friends exhibited body odor profiles that were significantly more similar to one another than to random control pairings. When human evaluators were brought in to smell the samples, their sensory impressions mirrored the machine's findings: human sniffers consistently rated the odors of click-friend pairs as being notably more alike than those of random dyads.

```mermaid flowchart TD accTitle: Inbal Raviv Click Friendship Experimental Workflow accDescr: A flowchart illustrating the recruitment, odor collection, electronic nose testing, and causality testing in Inbal Raviv's friendship study.

subgraph Sampling["1. Scent Isolation & Sampling"] A["Recruit same-sex click friend dyads"] --> B["Odorless hygiene protocol for ~3 weeks"] B --> C["Sleep alone in cotton T-shirt for 2 nights"] C --> D["Extract endogenous body odor"] end

subgraph Verification["2. Odor Profile Analysis"] D --> E["Electronic nose analysis with MOX sensors"] D --> F["Human evaluators sniff odor pairs"] E --> G["Click friends show significantly higher odor similarity"] F --> G end

subgraph Causality["3. Mirror Game Causality Test"] H["Recruit complete strangers"] --> I["Sample baseline odors with electronic nose"] I --> J["Nonverbal round-robin Mirror Game at close proximity"] J --> K["Post-interaction ratings of warmth and friendship"] K --> L["Electronic nose similarity predicts mutual social bonding"] end

G -.-> H ```

[12:35](https://www.youtube.com/watch?v=arMMF0ISRRM&t=755s) Testing Causality: The Mirror Game and Electronic Noses

The observational correlation left an unresolved question of causality: do click friends smell alike because of underlying biology, or does their shared scent stem from identical environmental exposures, such as dining at the same restaurants or living in the same neighborhood? To resolve this, Raviv designed an extensive experiment with complete strangers. Before participants ever met, their baseline body odors were characterized using the electronic nose. They were then paired up to play the "mirror game," an established social coordination task where two people stand in close physical proximity and mirror each other's hand movements in complete silence.

The experimental setup was conducted as a full round-robin tournament, meaning every participant was paired with every other participant without speaking. While Raviv's secondary hypothesis—that partners with similar scents would coordinate their physical hand movements more accurately—failed to show a significant effect, the post-game psychological evaluations yielded clear results. Without exchanging a single word, participants filled out questionnaires rating how much they liked their partner, how affectionate and kind they perceived them to be, and how likely they would be to form a friendship with them. Dr. Sobel emphasizes that the electronic nose reliably predicted which strangers would feel an instant mutual affinity based entirely on pre-existing chemical similarity:

> “So we could actually predict friendship using the electronic nose.”

- The electronic nose utilizes metal oxide (MOX) sensors to identify patterns in airborne volatile compounds rather than detecting isolated chemical constituents. - The mirror game required complete nonverbal communication to eliminate confounding variables such as vocal inflection, conversational rapport, or shared intellectual interests. - Pre-interaction chemical similarity was directly correlated with subjective ratings of warmth, perceived kindness, and prospective friendship interest.

[14:04](https://www.youtube.com/watch?v=arMMF0ISRRM&t=844s) Chemosignals and the Contagious Smell of Fear

Andrew Huberman steers the conversation toward broader chemosensory signaling between humans, asking whether these phenomena should be classified as classic pheromones, conventional olfaction, or something distinct. Dr. Sobel quickly dismisses the academic debate over nomenclature as pedantic, explaining that in his lab's formal scientific papers, they intentionally avoid using the word "pheromone":

> “It all becomes semantics, right?”

Dr. Sobel elaborates that the term creates unnecessary controversy among peer reviewers:

> “Yeah, it's all semantics. So, I don't I in our publications, we don't use the term pheromone, you know, because it would not help me and it would probably only hurt us. And so, you know, we talk about chemo signals and humans definitely emit chemo signals from their body and these chemo signals influence other humans and influence their behavior.”

Dr. Sobel notes that the primary focus of human chemosignaling research over the past decade has centered on the "smell of fear." When humans enter an acute state of terror, their bodies produce a distinct biochemical signature in their sweat. When this fear sweat is collected and presented to other individuals, it directly triggers an increase in their autonomic sympathetic arousal. In this sense, emotional fear operates as an airborne physiological contagion:

> “So, in effect, you could say that fear is contagious a bit.”

While cultural folklore has long asserted that dogs can smell fear on human beings, Dr. Sobel points out that this claim was only rigorously demonstrated in controlled scientific studies within the past eighteen months. Long before that canine study, pioneering human research teams—initiated by scientists like Denise Chen and Jeannette Haviland-Jones, and subsequently replicated in the Sobel laboratory—proved that human beings can reliably detect fear odors from other humans. When presented with samples gathered from participants undergoing fear-inducing protocols versus neutral control conditions, human subjects not only differentiate between the two, but their internal emotional and physiological states shift in response.

> “And they just highlight again, you know, that that we're incredibly olfactory animals.”

[15:48](https://www.youtube.com/watch?v=arMMF0ISRRM&t=948s) Hexadecanal, Social Buffering, and Regulating Aggression

Huberman recalls an earlier podcast conversation with Columbia neurobiologist Charles Zuker, who discussed the emotional salience of chemical signatures in human breath and the evocative scent at the nape of a grandchild's neck that elicits intense longing and affection. Dr. Sobel immediately connects these observations to a specific volatile molecule: hexadecanal.

Dr. Sobel explains that his former graduate student, Eva Mishor, set out to study the chemosignaling of human aggression. Her initial approach involved visiting mixed martial arts (MMA) gyms to harvest sweat from fighters, though these early attempts failed to yield consistent, interpretable data. Around the same time, neurobiologist Heinz Breer and his colleague Jörg Strotmann at the University of Hohenheim were investigating hexadecanal in mice. In rodents, hexadecanal is excreted in feces and serves as a potent chemical mediator of "social buffering"—the calming, prosocial sensation mice experience when huddled together in a group. Breer and Strotmann identified the olfactory receptor for hexadecanal and discovered that it is exceptionally conserved across mammalian evolution. Because cross-species conservation of chemosensory receptors is rare, Breer suspected hexadecanal served as a universal mammalian signal and reached out to Dr. Sobel:

> “Look, you got to study this stuff in humans, right?”

Breer shipped a supply of hexadecanal directly to Dr. Sobel's laboratory via FedEx. Mishor integrated the molecule into her validated experimental framework using the Taylor Aggression Paradigm (TAP), an established protocol that allows researchers to provoke and objectively quantify competitive aggression in human subjects. When men played the competitive game while exposed to hexadecanal versus a control odor, the molecule triggered a significant, reliable reduction in aggressive retaliation. Dr. Sobel notes that the statistical effect size of hexadecanal in dampening male aggression was unusually strong compared to typical findings in human behavioral research.

[18:59](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1139s) Hexadecanal, Sex Differences, and Baby Head Odor

When Dr. Sobel's lab tested the volatile compound hexadecanal on male human subjects, they observed a distinct drop in aggression—a result that aligned neatly with their initial hypotheses and earlier mouse studies. However, testing the compound on women revealed the exact opposite effect: hexadecanal significantly increased female aggression. Intrigued and puzzled, the lead researcher repeated the entire experiment from scratch inside an fMRI scanner to monitor real-time brain activity. The behavioral finding replicated cleanly—hexadecanal made men less aggressive and women more aggressive.

```mermaid flowchart TD accTitle: Effects of Hexadecanal on Adult Aggression accDescr: Diagram illustrating how hexadecanal from infant heads reduces aggression in men while increasing aggression in women.

Baby["Baby Head Odor<br/>(Abundant in Hexadecanal)"] Baby -->|"Chemosignal exposure"| Brain["Olfactory & Neural Processing"]

subgraph Responses["Sex-Divergent Responses"] Brain -->|"Men / Fathers"| Male["Reduced Aggression<br/>(Suppresses Infanticide/Violence)"] Brain -->|"Women / Mothers"| Female["Increased Aggression<br/>(Promotes Maternal Protection)"] end

Male --> Outcome["Offspring Survival Advantage"] Female --> Outcome ```

At first, this stark divergence seemed nonsensical to Sobel until the researcher struck upon an evolutionary explanation centered entirely on infant survival. Across mammalian species, paternal aggression can be directed violently toward offspring, including infanticide, whereas maternal aggression serves to fiercely protect the young. Therefore, emitting a molecule that pacifies males while simultaneously provoking protective fight responses in females provides the offspring an optimal survival advantage. Subsequent chemical analyses verified that hexadecanal is by far the most abundant semi-volatile molecule emanating from human baby heads. As Sobel notes regarding this evolutionary adaptation:

> “Well, babies are conducting chemical warfare, right? They're reducing aggression in their fathers or males around them and increasing aggression in their mothers or females around them, and both of those things are good for them.”

[20:38](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1238s) Sponsor: AG1

Andrew Huberman shares an endorsement for AG1, highlighting their updated AG1 Pro formulation. AG1 Pro builds on the standard formula of vitamins, minerals, probiotics, and adaptogens by adding three specific compounds: 5 grams of creatine monohydrate per serving to support muscle strength, physical performance, and cognitive health; calcium HMB to reduce muscle breakdown and promote recovery; and zinc carnosine to aid the gut mucosal lining. Huberman mentions that he has taken AG1 daily for roughly 14 years, well before starting his podcast.

[21:51](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1311s) Emotional Tears as a "Chemical Blanket" and Testosterone Suppression

Dr. Sobel explores the functional role of emotional crying. Historically, in *The Expression of the Emotions in Man and Animals*, Charles Darwin methodically accounted for the functional antecedents of virtually every human emotional expression, but became utterly stumped by weeping in Chapter 6, ultimately dismissing emotional tears as a mere non-functional epiphenomenon. Hypothesizing instead that human tears carry chemosensory communication, Sobel's team collected genuine emotional tears from donors and exposed male participants to them. Although human tears are completely odorless and subjects could not consciously smell anything, sniffing them triggered a profound biological reaction:

> “And you have about a 14% drop in free testosterone. And people who study testosterone, which is not me, but they tell me this is a really strong effect. Like it's hard to even pharmacologically get an effect like that that fast.”

Huberman, who previously conducted endocrine research, affirms that an endocrine drop of this magnitude within 20 to 30 minutes represents a massive effect size. Sobel was initially concerned that the demanding logistics of tear collection would discourage independent verification; however, two years later, an independent research group from South Korea (Oh et al.) replicated the exact drop in testosterone. Brain imaging during tear sniffing also revealed dampened activity under arousal states within the hypothalamus and the fusiform gyrus (known as a primary face-processing region).

While olfactory social signaling discoveries typically originate in rodents before being demonstrated in humans, this discovery inverted the timeline. Two years after Sobel's human publication, a *Nature* study demonstrated that tears from mouse pups similarly trigger aggression reduction in adult male mice through specific pheromonal compounds. Sobel conceptualizes human tears as an invisible protective layer—a functional "chemical blanket" that an individual instinctively secretes over their face to neutralize social aggression from others.

[25:37](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1537s) Digitizing Smells and Creating Olfactory Metamers

Addressing the emerging frontier of olfactory technology, Huberman brings up the challenge of generating digital scents on demand—imagining a future where searching for a recipe online produces the authentic aroma alongside text and images. Sobel recalls an old Google April Fools' Day video advertising "Google Smell," which parodied digital olfaction by showing users sniffing their phones. Today, tech giants and startups are pursuing that exact objective in earnest.

Sobel reveals that his lab achieved a major milestone in digital scent design, detailed in a *Nature* paper led by Aaron Raviv and Kobe Snitz. The researchers developed a computational algorithm capable of predicting the perceived olfactory similarity between any two arbitrary molecular mixtures with remarkable accuracy. Crucially, the mathematical model also enables the inverse capability: designing bespoke chemical formulations to achieve targeted scent profiles.

```mermaid flowchart LR accTitle: Olfactory Metamers Concept accDescr: Flow demonstrating how distinct chemical mixtures trigger identical scent perceptions.

subgraph MixtureA["Mixture A (e.g., Target Odor / Chanel No. 5)"] A["Chemical Components 1, 2, 3..."] end

subgraph MixtureB["Mixture B (Synthesized Olfactory Metamer)"] B["Zero Overlapping Chemical Components"] end

Algorithm["Sobel Lab Algorithmic Model"] -.->|"Calculates & Designs"| MixtureB MixtureA --> Perception["Identical Perceptual Experience"] MixtureB --> Perception ```

Using this platform, the team solved the olfactory equivalent of color metamerism in vision by synthesizing "olfactory metamers." By selecting completely disjoint, non-overlapping chemical compounds with zero structural commonality, they engineered mixtures that smell indistinguishable to human noses. While conventional master perfumers historically insisted that identical scent perception requires shared molecular components, Sobel's team disproved this by synthesizing a perfect metamer of Chanel No. 5 containing none of the iconic perfume’s actual constituent ingredients.

- The predictive algorithm allows researchers to systematically map the perceptual distance between novel chemical cocktails before synthesis. - Olfactory metamers establish that human smell perception is synthetic and computational rather than a simple 1-to-1 readout of specific molecules.

[28:25](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1705s) Sponsor: Eight Sleep

Huberman transitions to a mid-roll break for Eight Sleep, detailing the release of the Pod 6 smart mattress cover. The updated Pod 6 hardware is half the physical size of its predecessor, changes temperature 20% faster, and incorporates nine times as many biometric tracking sensors. Huberman underscores the neurobiology of sleep regulation, explaining that falling into deep sleep requires an internal core body temperature drop of 1 to 3 degrees Fahrenheit, whereas waking up rested depends on a natural temperature rise in the final hours of the sleep cycle.

[29:09](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1749s) Sponsor: Eight Sleep

Andrew Huberman describes how the Eight Sleep mattress cover dynamically tracks sleep stages and regulates bed temperature across the night to optimize deep sleep and rapid eye movement (REM) sleep. Having used the product for nearly five years, he notes that each hardware and software update has improved at tailoring the thermal environment to his body's physiological needs. Listeners interested in the Pod 6 cover can visit eightsleep.com/huberman to receive up to $200 off, with shipping available internationally, including to Mexico and the United Arab Emirates.

[29:48](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1788s) Digitizing Smells: Primary Odors and IP Transmission

Dr. Noam Sobel outlines how his laboratory laid the foundational infrastructure for digitizing olfaction. Their algorithmic framework relies on predicting perceptual similarity across a multidimensional reference space containing roughly 4,000 molecules, each mapped with its known human perceptual profile. Any novel chemical mixture can be projected into this geometric coordinate map to calculate its pairwise distance relative to established reference points; if a novel blend lands squarely on the coordinates of lemon or tomato, the algorithm accurately predicts that the mixture will smell like lemon or tomato. With the ability to predict the perceptual outcome of arbitrary molecular cocktails in hand, the team's current objective is isolating a universal palette of "odor primaries"—fundamental chemical building blocks that can be blended by a physical synthesizer to reconstruct any perceivable scent.

```mermaid flowchart LR accTitle: Pipeline for Transmitting Digital Odors Over IP accDescr: Flowchart illustrating the measurement of odorants in Germany, digital IP transmission, algorithmic mapping, and physical synthesis of target scents.

subgraph Acquisition ["Germany (Max Planck)"] A["Odorant Sample"] --> B["GC-MS Measurement<br/>(Jonathan Williams)"] end

subgraph Transmission ["Network"] B -->|"Data over IP"| C["Algorithmic Framework<br/>(Sobel Lab, Israel)"] end

subgraph Synthesis ["Weizmann Institute"] C --> D["4,000-Molecule Perceptual Map"] D --> E["Primary Mixer Device"] E --> F["Recreated Scent Output"] end ```

To test whether scents could be digitized and broadcast remotely, Sobel partnered with atmospheric chemist Jonathan Williams at the Max Planck Institute in Munich, who specializes in gas chromatography–mass spectrometry (GC-MS). Williams analyzed odorants in Germany, encoded their chemical profiles, and transmitted the raw data over internet protocol (IP) to Sobel’s team. Using their mathematical framework, the researchers translated the incoming data stream into specific instructions for an automated mixer loaded with odor primaries. In their initial proof-of-concept trial, they attempted to transmit and reconstruct four distinct target aromas:

- **Rose:** Deemed a failure; rather than smelling floral, the synthetic output was identified by most human evaluators as bubblegum. - **Anise:** Failed as well; participants predominantly identified the scent as cherry, which was perceptually adjacent but inaccurate. - **Gasoline:** Modestly successful; participant identification cleared statistical chance by a slight but significant margin. - **Violets:** An unequivocal breakthrough; 15 out of 16 human evaluators independently recognized the scent as violet, making it the first odor in history successfully transmitted over IP.

Sobel cautions against expecting commercial smell-o-vision devices in living rooms anytime soon, explaining that current technology remains in its infancy: “We're talking now of trying to achieve the olfactory equivalent of circa 1956 black-and-white TV, okay?” He clarifies that his benchmark for near-term success is not distinguishing subtle vintages—such as differentiating a Cabernet Sauvignon from a Merlot—but rather successfully synthesizing a coarse aroma that a subject can broadly identify as "wine."

[32:34](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1954s) The Future of Olfaction: Continuous Medical Diagnostics

Looking decades into the future, Dr. Sobel predicts that the real transformative value of high-resolution digital olfaction—what he likens to achieving "4K odor transmission"—will lie in clinical medicine rather than consumer entertainment. Because every pathology alters cellular biochemistry, each disease generates a distinct signature of metabolic byproducts that disperse into bodily fluids and breath. As Sobel summarizes: “Metabolic process have metabolites, metabolites have a smell.” Once electronic sensor arrays and artificial intelligence reach sufficient fidelity, passive volatile organic compound (VOC) screening will revolutionize preventative health monitoring.

Sobel acknowledges that achieving this degree of sensitivity and reliability is far off, remarking that it might only come to fruition well after he has retired. Nevertheless, he envisions a future where consumer bathrooms are standardly equipped with continuous ambient monitors: “Excuse me for the imagery, but you will have an electronic nose in your bathroom. Each one of us will have in the toilet, and it will be doing diagnostics all the time.” Noninvasive e-noses will continuously sample volatile emissions from urine and feces, screening for early-stage metabolic disorders, gut dysbiosis, and cancers long before clinical symptoms emerge.

[33:43](https://www.youtube.com/watch?v=arMMF0ISRRM&t=2023s) Acknowledgments and Closing Remarks

Andrew Huberman concludes the interview by expressing deep gratitude to Dr. Sobel for providing an expansive masterclass on olfactory neurobiology, spanning from chemosignals, fear contagion, and hormonal modulation to synthetic odor transmission. Huberman highlights Sobel's generosity in repeatedly citing collaborators and praises the infectious curiosity and joy that characterizes his scientific career. Invoking a cherished mantra from the late Stanford neurobiologist Ben Barres—"keep going"—Huberman commends Sobel for maintaining an uncompromised passion for fundamental discovery despite the complex social politics that often accompany academic research. Sobel warmly thanks Huberman for the dialogue and the generous reception of his life's work.

## Notable quotes > [18:59](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1139s) Well, babies are conducting chemical warfare, right? They're reducing aggression in their fathers or males around them and increasing aggression in their mothers or females around them, and both of those things are good for them. — Dr. Noam Sobel

> [32:34](https://www.youtube.com/watch?v=arMMF0ISRRM&t=1954s) Metabolic process have metabolites, metabolites have a smell. — Dr. Noam Sobel

## Mentioned Dr. Andrew Huberman, Dr. Noam Sobel, Weizmann Institute of Science, Idan Frumin, LMNT, Inbal Raviv, Denise Chen, Jeannette Haviland-Jones, Charles Zuker, Columbia University, Eva Mishor, Heinz Breer, Jörg Strotmann, University of Hohenheim, Taylor Aggression Paradigm (TAP), AG1, Charles Darwin, The Expression of the Emotions in Man and Animals, Oh et al., Nature, Google, Aaron Raviv, Kobe Snitz, Chanel No. 5, Eight Sleep, Jonathan Williams, Max Planck Institute, Ben Barres

## Video description In this Huberman Lab Essentials episode, my guest is Dr. Noam Sobel, PhD, a Professor of Neurobiology at the Weizmann Institute of Science and an expert on the human sense of smell. We discuss how the brain processes odors and how smell shapes our lives more than we realize, influencing everything from who we become friends with to how we experience emotions, hormones and aggression. Dr. Sobel also shares his lab’s groundbreaking work on digitizing smell and future technologies that could enable noninvasive disease detection through body odors.

Show notes: https://www.hubermanlab.com/episode/essentials-how-smells-influence-our-hormones-health-and-behavior-noam-sobel

Thank you to our sponsors AG1: https://drinkag1.com/huberman Eight Sleep: https://eightsleep.com/huberman LMNT: https://drinklmnt.com/huberman

Follow Huberman Lab Instagram: https://www.instagram.com/hubermanlab Threads: https://www.threads.net/@hubermanlab X: https://x.com/hubermanlab Facebook: https://www.facebook.com/hubermanlab TikTok: https://www.tiktok.com/@hubermanlab LinkedIn: https://www.linkedin.com/in/andrew-huberman Website: https://www.hubermanlab.com Newsletter: https://www.hubermanlab.com/newsletter

Timestamps 00:00:00 Noam Sobel 00:00:20 Biology of Smell (Olfaction) 00:04:32 Human Smell, Social Interaction, Handshaking 00:08:31 Sponsor: LMNT 00:10:03 Friendship, Body Odor 00:14:04 Human Chemosignals, Smell of Fear & Contagion 00:15:48 Aggression, Chemosignals, Men vs Women, Babies 00:20:38 Sponsor: AG1 00:21:51 Tears, “Chemical Blanket”, Testosterone 00:25:37 Digitizing Smells 00:28:25 Sponsor: Eight Sleep 00:29:48 Creating Digital Odors; Future Medical Diagnostics 00:33:43 Acknowledgments

Disclaimer & Disclosures: https://www.hubermanlab.com/disclaimer

[Source](https://www.youtube.com/watch?v=arMMF0ISRRM)

中文简报

魏茨曼研究所科学家诺姆·索贝尔首次通过IP网络传输气味 人类潜意识中的化学信号可直接调节激素、抑制攻击行为,并在发生任何言语互动前预测即刻建立的友谊。 诺姆·索贝尔博士概述了嗅觉系统的解剖结构,详细阐述了数百万个受体如何汇集到嗅球以处理空气中的挥发性分子。他指出,人类通过握手后摸脸等行为持续对生物气味进行采样,并证实内源性体味的相似性能够可靠地预测陌生人之间即刻建立的非浪漫友谊。讨论还探讨了化学信号(例如婴儿头部的十六烷醛、空气中散发出的恐惧气味,以及情绪性眼泪中的无味分子)如何强效调节人类攻击性并抑制睾酮水平。最后,索贝尔解释了合成嗅觉异谱同色物、通过IP网络传播气味背后的算法突破,以及电子鼻在连续健康筛查方面的未来潜力。 人类嗅觉受体表现出极高的汇聚性:大约350种亚型的约600万至700万个感觉神经元直接穿过筛板投射到特定的嗅球嗅小球中,绕过了常规的丘脑感觉中继。 (0:20) 潜意识中的化学采样驱动着人类互动;人们在握手后会习惯性地闻自己的手,而电子鼻检测到的基线内源性体味相似性,能够可靠地预测陌生人之间自发建立的友谊。 (4:32) 十六烷醛是一种大量存在于婴儿头部的半挥发性化合物,它会引发人类性别差异化的行为反应,显著降低男性的攻击性,同时提高女性的攻击性。 (18:59) 嗅闻无味的人类情绪性眼泪,会使成年男性的游离睾酮水平迅速降低约14%,并减少与唤醒相关的大脑区域的神经活动。 (21:51) 研究人员展示了通过互联网协议(IP)进行数字气味传输,利用气相色谱-质谱联用仪和预测算法调配主要气味剂,成功传输了可辨识的紫罗兰香气。 (29:48)

Reading summary

Neurobiologist Dr. Noam Sobel demonstrates that human olfaction acts as a powerful, largely subconscious driver of physiology, social bonding, and hormonal regulation, while advances in olfactory modeling are opening pathways toward digital scent synthesis and continuous medical diagnostics. Neural Architecture of Human Olfaction (0:00) Human olfaction begins when volatile molecules enter the nasal cavity via sniffing and reach the olfactory epithelium, which houses roughly 6 to 7 million sensory neurons representing approximately 350 receptor subtypes. Transduction generates genuine action potentials that pass through the cribriform plate into the olfactory bulb. The system demonstrates extreme anatomical convergence: all sensory neurons expressing the same receptor subtype direct their axons into just one or two glomeruli per bulb. Subconscious Chemosignaling in Human Social Interactions (4:32) Despite cultural assumptions that human olfaction is weak, people constantly engage in covert chemical sampling. In an experiment by Idan Frumin, subjects greeted with a handshake exhibited an immediate spike in hand-to-nose contact; airflow sensors verified that individuals deliberately inhaled when their hands brushed their nostrils, sniffing both the interacting hand and the untouched hand to compare personal and external scents. Olfactory cues also drive nonverbal social affinity. In research by Inbal Raviv, same-sex 'click friends' exhibited significantly more similar endogenous body odors on metal oxide sensor electronic noses than random pairs. In a follow-up nonverbal mirror game between strangers, baseline chemical similarity predicted mutual ratings of warmth, kindness, and prospective friendship before participants spoke. Hormonal and Behavioral Modulation by Body Odors (14:04) Human chemosignals trigger distinct autonomic and endocrine shifts without conscious scent perception. Exposure to sweat collected during acute fear states induces an autonomic sympathetic arousal response in recipients, effectively making fear physiologically contagious. Similarly, sniffing odorless emotional tears causes an approximately 14% drop in free testosterone in men within 20 to 30 minutes, alongside dampened arousal activity in the hypothalamus and fusiform gyrus—a finding independently replicated by South Korean researchers. Sex-Divergent Responses to Hexadecanal (15:48) The semi-volatile compound hexadecanal, highly enriched on the scalps of human infants, serves as an evolutionary behavioral regulator. Using the Taylor Aggression Paradigm, researchers found that hexadecanal reliably reduced aggressive retaliation in men while significantly increasing aggression in women, effects replicated inside fMRI scanners. Sobel attributes this divergence to an evolutionary mechanism benefiting infant survival: suppressing potentially lethal paternal aggression while stimulating maternal protective aggression. Digitizing Scents and Future Diagnostics (25:37) Using an algorithmic model mapping roughly 4,000 molecules, Sobel's team synthesized olfactory metamers—blends of entirely non-overlapping chemical constituents that produce identical perceptual smells, including an ingredient-free formulation of Chanel No. 5. In collaboration with Jonathan Williams, the team tested digital scent transmission over internet protocol from Germany to Israel using gas chromatography–mass spectrometry data; among four attempted smells, violet was successfully identified by 15 of 16 participants, while rose registered as bubblegum and anise as cherry. Sobel emphasizes that the technology remains in its infancy, comparing its current fidelity to 1956 black-and-white television. Looking forward, he predicts the primary long-term value will not be consumer entertainment, but passive healthcare diagnostics via ambient electronic noses installed in bathrooms to detect metabolic diseases and cancer signatures from urine and feces. While the digitization of olfaction has demonstrated proof of concept, commercial entertainment applications remain limited by coarse synthesis resolution. Sobel notes that high-fidelity clinical sensing of volatile organic compounds represents the ultimate frontier, though broad implementation remains decades away.

精炼阅读摘要

魏茨曼科学研究所神经生物学教授 Noam Sobel 指出,人类嗅觉系统具有精密且高度汇聚的神经生物学架构,能在潜意识层面通过化学信号深刻调节社交互动、攻击行为与激素水平;同时,气味数字化与嗅觉异谱现象的突破,预示着该领域未来在无创病理监测中具备重要应用前景。 嗅觉感受架构与中枢神经的极端汇聚 (0:00) Sobel 解释道,人类初级嗅觉依赖于外部空气中的挥发性分子经鼻腔吸入,上行约6至7厘米到达鼻上皮。该区域包含约600万至700万个嗅觉感觉神经元,覆盖约350种不同的受体亚型,占据了人类基因组的重要编码比例。当气味分子与匹配受体结合后,感受神经元直接产生真正的动作电位,穿过人类颅骨最薄的多孔结构——筛板,直通嗅球。该通路展现了哺乳动物中枢神经系统中最为极端的汇聚机制:表达同一种受体基因的所有外周神经元,其轴突全部精确汇聚到嗅球内的仅一至两个特异性肾小球(glomeruli)中,形成特定的空间气味图谱。与视觉等其他感觉系统不同,嗅觉输入完全绕过了丘脑的强制中继,直接投射至梨状皮层、内嗅皮层,并广泛分支至杏仁核与下丘脑,构成了驱动本能与情绪反应的直接通路。 社交中的潜意识嗅探与体味共鸣 (4:32) 尽管人类文化规范限制了公开的嗅闻行为,但 Sobel 团队的研究证明,人类会通过隐秘动作进行频繁的化学取样。Idan Frumin 主导的实验显示,受试者在与陌生人握手后,手部触碰鼻部的频率显著激增;同步鼻腔气流传感器证实,受试者在手贴近鼻孔的瞬间执行了精确的主动吸气采样,并同时嗅闻未握手的另一侧手,这表明大脑在潜意识中对比自身与他人的生物化学轮廓。 在探索人际“一见如故”现象时,Inbal Raviv 收集了排除了人造香精的非恋爱同性好友体味样本。金属氧化物传感器(MOX)电子鼻与人类嗅觉评估均证实,亲密朋友间的体味特征显著相似。为了排除环境共同暴露的干扰,研究团队在完全陌生的受试者之间进行了无声的“镜子游戏”动作协调测试。实验结果显示,在未进行任何言语交流的情况下,电子鼻在初次见面测得的气味相似度,能够显著预测两名陌生人之间产生的互爱情绪、亲切感评价以及未来的结友意愿。 化学信号对攻击性与内分泌的相反调控 (15:48) Sobel 实验室在正式论文中倾向于使用“化学信号”(chemosignals)而非引发争议的“外激素”(pheromone)。他们证实,人类处于极端恐惧状态时由汗液分泌的生化分子具备生理传染性,能直接引发他人自主交感神经系统的觉醒。在针对攻击性分子十六醛(hexadecanal)的研究中,Eva Mishor 发现该分子具有显著的性别二态性效应:在泰勒攻击范式(TAP)测试中,十六醛大幅抑制了男性的攻击性报复,却在功能性磁共振成像(fMRI)监控下显著激发了女性的攻击反应。Sobel 认为这种相反表现符合演化对婴儿生存的保护策略——十六醛是人类婴儿头皮最丰富的半挥发性分子,其在压制雄性潜在杀婴倾向的同时,唤起雌性的母性防御本能。 此外,Sobel 团队发现人类情感泪液虽然没有任何主观气味,却充当着“化学保护毯”。男性嗅闻真实的悲伤眼泪后,体内游离睾酮在20至30分钟内出现约14%的急剧下降,这一强烈的内分泌抑制效果随后被韩国独立研究团队完全复现。脑部扫描显示,嗅闻眼泪同时显著降低了下丘脑以及面部识别中枢梭状回的神经激活,展示了无意识化学分子直接削弱社会攻击性的机制。 嗅觉数字化异谱合成与未来连续健康监测 (25:37) 在数字嗅觉领域,Aaron Raviv 与 Kobe Snitz 构建了一种基于约4000种已知分子感知坐标的算法,能够精确预测任意化学混合物的感知气味距离,并由此合成出“嗅觉同色异谱”(olfactory metamers)。研究人员通过调配与经典香水完全不重叠、无共同化学成分的分子配方,成功重构出受试者无法区分的香奈儿5号气味,推翻了必须依靠相同分子才能复刻同等感官体验的传统调香假设。此外,团队联合马克斯·普朗克研究所的 Jonathan Williams 进行了气相色谱-质谱(GC-MS)数据网络传输测试,成功在异地依靠基础气味原料重构出紫罗兰花香(16名测试者中有15人准确识别),完成了首次基于IP网络协议的气味还原。 展望未来应用,Sobel 预测高分辨率气味检测的核心价值将落在临床预防医学而非娱乐设备上。由于所有细胞病理改变都会产生特定的代谢产物气味,Sobel 设想未来家庭卫生间可普遍配备高灵敏度电子鼻,对尿液和粪便中的挥发性有机化合物(VOC)进行日常、持续的无创监测,从而在临床症状显现前早期识别肠道菌群失调、代谢异常乃至癌症病灶。 Sobel 坦言,虽然气味数字化与化学信号传导的机理已经确立,但目前的合成与传输技术仅相当于“1956年的黑白电视”,尚无法精细区分复杂的风味层次;将高精度电子鼻完全转化为空中代谢病理筛查设备仍面临巨大的工程与生物学壁垒,可能需要数十年甚至更久才能实现临床普及。

Source quotes

[18:59] Dr. Noam Sobel: Well, babies are conducting chemical warfare, right? They're reducing aggression in their fathers or males around them and increasing aggression in their mothers or females around them, and both of those things are good for them. [32:34] Dr. Noam Sobel: Metabolic process have metabolites, metabolites have a smell.

Original transcript

[0:00] Welcome to Huberman Lab Essentials, [0:01] [music] where we revisit past episodes [0:04] for the most potent and actionable [0:06] science-based tools for mental health, [0:08] physical health, and performance. [0:11] I'm Andrew Huberman, and I'm a professor [0:13] of neurobiology and ophthalmology at [0:15] Stanford School of Medicine. And now for [0:17] my discussion with Dr. Noam Sobel. Dr. [0:20] Sobel, Noam, welcome. [0:22] >> Thank you. [0:23] >> I must say I am extremely excited for [0:25] this conversation. I've been a huge fan [0:27] of your work, and today I'd love for you [0:29] to share with us all about the amazing [0:31] landscape of chemosensation, in [0:32] particular olfaction, our sense of [0:35] smell, and some related perceptual [0:37] abilities, or subconscious abilities, [0:39] including pheromones, etc. To get [0:41] everybody on the same page, I'd like to [0:43] just start off by asking [0:45] what are the major components of our [0:47] ability to smell? [0:49] >> [snorts] [0:49] >> Obviously, or I like to think it [0:51] involves the nose at some level. [0:53] >> It does. [0:53] >> Now, if you could just give us the top [0:55] contour, or even deep contour if you [0:57] like, of the the parts list, and the [1:01] various roles they play. [1:03] >> Primary olfaction is referred to as [1:04] orthonasal olfaction, that is through [1:06] our nose. We sniff, and sniffing is a [1:08] big thing. Well, I have a sense we might [1:10] talk about that a lot today, in all [1:12] sorts of contexts. So, we sniff in [1:15] through our nose, and to answer your [1:16] general question of the organization of [1:18] the system, um so, molecules, airborne [1:22] molecules, travel up our nose a distance [1:24] in the human of about 6 or 7 cm to about [1:28] here, where they interact with I will [1:31] use the word sheet of receptors, but [1:34] sheet is a bit misleading here. It's not [1:37] a sheet, it's very convoluted. [1:40] We have about 7 million such receptors [1:43] uh lining a structure known as the [1:45] olfactory epithelium. This is the [1:47] sensory surface of the olfactory system, [1:49] the olfactory epithelium. Again, about [1:51] probably about 6 or 7 million receptors [1:53] in the human. [1:54] In the human, probably of about 350 [1:59] different kinds. So, that's amazing. [2:01] That means a [2:02] meaningful percentage of your genome is [2:05] devoted just to this, just to the kinds [2:07] of olfactory receptor subtypes you have [2:09] in your nose. So, odorants interact with [2:11] these receptors here in our epithelium [2:15] where they undergo what is referred to [2:17] as transduction. That is, the odorants [2:19] are docked at a receptor and turn into a [2:22] neural signal and this neural signal, in [2:24] fact, action potentials, not gradient [2:26] potentials of any kind, uh propagates [2:29] via the olfactory nerve through the [2:32] thinnest part of our skull, which is [2:34] perforated. It has a lot of holes. [2:36] The nerve goes through those holes and [2:38] synapses at the first target in the [2:40] brain, [2:41] which is the olfactory bulb. The pattern [2:44] of that synapsing follows what's [2:45] referred to as the most extreme case of [2:47] convergence in the mammalian nervous [2:50] system. [2:51] More specifically, what happens is that [2:54] all the receptors of one subtype [2:56] converge to one location in the bulb. [3:00] And this location is referred to as a [3:01] glomerulus or in plural, glomeruli. And [3:04] and that may be a slight [3:05] oversimplification. It's in fact two [3:07] glomeruli. The textbook view of how [3:08] things work is that every such receptor [3:11] subtype is responsive to a small subset [3:14] of different molecular shapes, what's [3:17] sometimes referred to as odoritopes, the [3:19] molecular aspects of the odorant. And [3:22] each odoritope will activate a different [3:24] subset of receptors. Then, because of [3:26] this convergence, you end up having on [3:28] the bulb in the way of a map reflecting [3:32] receptor identity. So, so let's say this [3:35] coffee activates receptors of type 1, 3, [3:38] and 7. So, the glomeruli of receptors 1, [3:41] 3, and 7 will light up, quote unquote, [3:43] when I smell the coffee. And if you [3:45] could take a snapshot of that, [3:46] theoretically, you would have the map of [3:49] coffee. This is sort of the textbook [3:50] view of how the system works. And then [3:52] information goes from the bulb to [3:54] several targets in the brain. I mean, [3:56] what is referred to as primary olfactory [3:57] cortex is piriform cortex and and [3:59] entorhinal cortex. This is on the [4:01] ventral surface of the brain, the lower [4:03] portion of our temporal lobe, and [4:05] information goes there directly, but it [4:07] also goes directly to the amygdala. It [4:09] probably goes directly to the [4:10] hypothalamus. It may go directly to the [4:13] cerebellum. Uh it goes all over the [4:15] brain. So so [4:18] information projects widely from there, [4:20] and as far as people understand, the map [4:23] that may exist on the bulb uh doesn't [4:26] exist in the rest of the brain, and the [4:27] understanding of of how coding occurs in [4:30] the rest of the brain is is murky. [4:32] Again, in our lab, we already said, "You [4:34] know, yeah, we know this. This is old [4:35] news." But but to people who who come [4:38] from different worlds, we have to [4:39] reiterate this. Sometimes when I give [4:41] uh you know, public lectures to to [4:43] non-olfaction audiences, I reiterate [4:45] this. [4:46] Humans have an utterly remarkable sense [4:49] of smell. We look at human behavior a [4:52] lot. We try to look at it through our [4:54] nose and in the way we look at what [4:56] people are doing. You know, we try to [4:59] think, you know, if I was a dog, [5:01] what what would I think of this? And and [5:05] you know, if you look at dogs, right? [5:06] They they [5:07] you know, when they interact, they [5:09] visibly sniff each other. It's very [5:10] obvious. They walk up to each other, and [5:12] they sniff each other. And yet, humans [5:14] don't typically walk up to a stranger [5:16] and carefully sniff them, right? I mean, [5:18] it's We're we're sort of obliged to [5:21] sniff our our babies. That's considered [5:23] almost something you're supposed to do. [5:25] Um and it's not culturally taboo to [5:28] sniff our loved ones. It's sort of [5:30] doesn't seem like an odd thing to do. [5:32] But we don't sniff strangers, all right? [5:35] Well, or do we? So [5:37] we're finding more and more mechanisms [5:40] where we do this. And And the one I'm [5:41] referring to now, for one example, is we [5:44] started looking at handshaking. There's [5:46] really poor understanding of how this [5:48] behavior evolved. Like, where did where [5:50] did this thing come from? You know, if [5:52] you look for the Wikipedia version, [5:54] right? Then they'll tell you that it's [5:56] to show that you're not holding a weapon [5:58] in your hand. But there's really no good [6:01] evidence for that. And we started [6:02] looking at people handshaking, and and [6:04] we noticed, or it seemed to us that [6:06] we're noticing that you know people will [6:08] shake hands and then we'll go like this [6:10] and like [6:11] >> For those of you listening not watching, [6:13] Noam um he was taking his hand and and [6:16] wiping it on his face. [6:17] >> Yeah, they were [6:18] touching myself. Yeah, these things [6:20] these things that we do all the time. [6:22] >> After a handshake. [6:23] >> Well, so so first of all we do them all [6:24] the time just period, right? The [6:26] baseline here is really high and we'll [6:27] get to that in a second, but but but [6:29] these behaviors that you know you you [6:31] you could easily not notice, right? And [6:33] and so we we asked whether that's a real [6:36] thing. [6:37] And this was a study led by Idan Frumin [6:39] in our lab at the time. So what we did [6:41] is we brought in participants to our lab [6:45] and we sat them in the room [6:47] uh experiment room [6:49] and and told them the experiment would [6:51] start soon and they should wait for us [6:52] there. They didn't know what they were [6:53] coming from. [6:54] Unbeknownst to them they were already [6:55] being videoed and they would sit there [6:58] for 3 minutes. [6:59] And during those 3 minutes we could [7:01] later quantify how much indeed they just [7:03] by baseline how much they touch their [7:05] nose or their forehead or their chin or [7:08] how many times their hand [7:10] uh uh reaches their face. And by the way [7:12] that baseline is not low. And then 3 [7:13] minutes later an experimenter would walk [7:15] into the room and would share a [7:17] consistent text it would be you know [7:19] we're still setting up our equipment in [7:21] the other room [7:22] uh and and so just wait here and we'll [7:24] be right back with you, but in the [7:25] meantime just wait here. And [7:28] in half of the cases it included a [7:29] handshake. This was a new experimenter, [7:31] not the one who put them in the room. So [7:32] that's the first time they met. So it [7:34] would be hello, I'm you know so-and-so [7:35] they would put out their hand and shake [7:37] their hand or not. And so you had [7:38] handshake and no handshake conditions [7:40] and then you can quantify that behavior [7:43] of the hand going to the nose after [7:45] handshake. [7:46] And there was a remarkable increase in [7:49] the hand going to the nose after [7:50] handshake. It's unreal. [7:52] >> So interesting. [7:53] >> hand goes to the nose. Now we did a few [7:54] controls here [7:56] to verify that this is an olfactory [7:58] behavior. One is unbeknownst to [7:59] anti-participants we measured nasal [8:01] airflow. [8:02] And and and people not only bring their [8:04] hand to their nose, they sniff it. So [8:05] and this is perfectly timed. They go [8:07] like this, okay? So, they're sniffing [8:09] their hands. So, this was an olfactory [8:11] behavior. No doubt about it. I mean, [8:13] we're we're quite confident. And there's [8:15] an interesting thing going on here, too, [8:16] because people didn't only smell the [8:19] hand that shook. They also smelled the [8:21] other hand. And we think that there's [8:23] something going on here comparing self [8:26] to other. [8:27] And we think a lot of self-sampling [8:29] might might reflect that. [8:31] >> I'd like to take a quick break and [8:32] acknowledge one of our sponsors, [8:34] Element. Element is an electrolyte drink [8:36] that has everything you need and nothing [8:38] you don't. That means the electrolytes, [8:40] sodium, magnesium, and potassium in the [8:42] correct amounts, but no sugar. Proper [8:44] hydration is critical for optimal brain [8:46] and body function. Even a slight degree [8:48] of dehydration can diminish cognitive [8:50] and physical performance. It's also [8:52] important that you get adequate [8:53] electrolytes. The electrolytes, sodium, [8:55] magnesium, and potassium, are vital for [8:57] functioning of all the cells in your [8:59] body, especially your neurons or your [9:01] nerve cells. Drinking Element dissolved [9:03] in water makes it very easy to ensure [9:05] that you're getting adequate hydration [9:06] and adequate electrolytes. To make sure [9:08] that I'm getting proper amounts of [9:10] hydration and electrolytes, I dissolve [9:12] one packet of Element in about 16 to 32 [9:14] oz of water when I first wake up in the [9:16] morning, and I drink that basically [9:17] first thing in the morning. I'll also [9:19] drink Element dissolved in water during [9:21] any kind of physical exercise that I'm [9:22] doing, especially on hot days when I'm [9:24] sweating a lot and losing water and [9:26] electrolytes. Element has a bunch of [9:28] great-tasting flavors. I love the [9:30] raspberry. I love the citrus flavor. [9:32] Right now, Element has a limited edition [9:34] lemonade flavor that is absolutely [9:36] delicious. I hate to say that I love one [9:38] more than all the others, but this [9:39] lemonade flavor is right up there with [9:41] my favorite other one, which is [9:42] raspberry or watermelon. Again, I can't [9:44] pick just one flavor. I love them all. [9:46] If [snorts] you'd like to try Element, [9:48] you can go to drinkelement.com/huberman, [9:51] spelled drinklmnt.com/huberman, [9:54] to claim a free Element sample pack with [9:56] a purchase of any Element drink mix. [9:58] Again, that's drinkelement.com/huberman [10:01] to claim a free sample pack. [10:03] >> There's a study we published by Inbal [10:06] Raviv in in our lab. Um [10:09] where [10:10] Inbal came with this uh uh [10:12] basic interest [10:14] in this phenomenon that's loosely [10:16] referred to as click friendships. So, [10:19] people you meet and you click right [10:20] away. Everybody experienced this in [10:22] their life, you know, to some extent. [10:25] And the question is [10:26] what what was there, right? What was it? [10:28] Was it because you looked the same? [10:30] Could be. Was it because, you know, you [10:32] had the same [10:33] sports team that you liked? Or [10:35] is there something uh deeper here? [10:39] And and Inbal's theory was that a [10:41] similarity in body odor may contribute [10:43] to this. [10:45] That people who [10:46] smell the same will click in some way. [10:50] And so, to address that, she she [10:52] actually recruited click friends from [10:54] all over Israel. She posted all over [10:55] social media [10:57] uh to identify pairs of friends. So, [10:59] these are our same-sex non-romantic [11:01] dyads. So, these are friends, men and [11:03] women, whose friendship started as a [11:05] click. And then [11:07] she um sampled their body odor. And And [11:10] we have a a protocol for this. Uh [11:13] they're given, you know, uh odorless [11:15] shampoo and soap to use for 3 weeks or [11:17] something. And then they sleep two [11:19] nights in this T-shirt uh where they [11:21] have to sleep alone. And then we extract [11:23] the body odor from the T-shirt. And so, [11:25] we have a way to extract a method to [11:27] extract body odor. And then she she [11:30] first asked whether indeed click friends [11:33] are more similar in their body odor than [11:35] you would expect by chance. And she [11:37] first tested this with a with a device, [11:39] a machine we call an electronic nose. [11:42] So, an electronic nose is sort of a very [11:45] poor effort to mimic what the mammalian [11:46] nose does. Basically, it's a bunch of [11:48] sensors that respond to airborne [11:49] molecules. In this case, sensors [11:51] referred to as MOX sensors as a metal [11:53] oxide-covered sensors. And so, she used [11:56] an electronic nose to sample these body [11:58] odors. And she found that click friends [12:00] are indeed more similar to each other uh [12:03] than you would expect uh by chance, by [12:05] random dyads. And And this was a [12:06] significant difference. [12:08] And after she found that a device could [12:11] do this, [12:12] she had other participants do this. So [12:15] So she had people smelling the click [12:17] friends versus non-click friends and and [12:19] they judged them as being more similar [12:21] to each other than not. Now again, you [12:24] you might wonder is this causal or not, [12:26] right? Because maybe click friends go to [12:28] the same restaurant together or all the [12:29] time or whatever or live in the same [12:30] neighborhood and and that's why they [12:33] they they smell the same. So to address [12:35] causality, she recruited total strangers [12:39] and first smell them with the electronic [12:42] nose and then engage them in a social [12:44] interaction, something called the mirror [12:46] game. So in the mirror game, one person [12:49] moves their hands and the other person [12:50] is really close to them, like right [12:52] here, so they can smell each other [12:54] and has to move their hands with the [12:56] other person. [12:58] And one one prediction there panned out, [13:01] uh but another didn't. The one that [13:02] didn't So she predicted that people [13:04] would smell more [13:06] uh similar to each other would be better [13:08] at the mirror game. That is, they would [13:09] follow each other better. That did not [13:11] pan out. However, [13:13] she then also had [13:15] the interaction was completely [13:16] nonverbal. They were not allowed to [13:17] speak with each other and she did an [13:19] entire round robin, so everybody played [13:21] with everybody else. This was an insane [13:22] experiment to run. [13:24] She [snorts] then at the end of the [13:25] experiment, each person rated each other [13:27] person as to how much they think they [13:29] would want to be their friends and also [13:31] on a bunch of ratings, how nice they [13:33] think they are, how affectionate they [13:34] think a bunch of ratings, okay? [13:37] All of this was predicted by the [13:38] electronic nose. So people who smell [13:40] more similar to each other [13:42] think that the other person is more [13:44] likely to be their friend, is more [13:46] likely to be a nice person, etc. etc. So [13:49] we could actually predict friendship [13:51] using the electronic nose. So this is [13:54] not a result of friendship. It's It It [13:56] plays into the causal elements of of [13:58] building friendship. People are [14:00] constantly smelling themselves, [14:02] constantly smelling others. [14:04] >> I want to touch on some other uh [14:05] so-called pheromone effects. Whether or [14:07] not it's a classic pheromone effect or [14:09] whether or not it's olfaction or [14:10] something else. This is chemosensory [14:13] signaling between individuals. [14:15] >> It all becomes semantics, right? [14:17] >> of like nerdy inside ball. [14:18] >> Yeah, it's all semantics. So, I don't I [14:20] in our publications, we don't use the [14:22] term pheromone, you know, because it [14:24] would not help me and it would probably [14:25] only hurt us. And so, you know, we talk [14:28] about chemo signals and humans [14:30] definitely emit chemo signals from their [14:32] body [14:33] and these chemo signals influence other [14:35] humans and influence their behavior. The [14:38] flavor of the month for the past 10 [14:40] years in this field [14:42] is what's referred to as the smell of [14:43] fear. Humans emit a particular body odor [14:46] when they're in a state of fear and this [14:48] body odor influences other humans in [14:51] effect increasing [14:53] their autonomic arousal, their [14:55] sympathetic state. So, in effect, you [14:58] could say that fear is contagious a bit. [15:00] So, the smell of fear is contagious. By [15:03] the way, culturally, we know for ages [15:06] that dogs can smell fear in humans, but [15:09] actually that was only really shown [15:10] about a year and a half ago in a study. [15:13] So, it it was always said, but it wasn't [15:15] really shown effectively. It was shown [15:16] about a year and a half ago in study [15:18] that dogs indeed can smell human fear. [15:21] And humans can smell human fear. So, [15:23] several labs starting from Denise Chen [15:26] and Haviland Jones and and then in our [15:28] lab and in other labs, if if you collect [15:31] body odor from people in a state of fear [15:34] and collect body odor from the same [15:36] people when they're not in a state of [15:37] fear, [15:38] other people can determine which is is [15:40] the state of fear or not and and this [15:42] influences their behavior. And they just [15:44] highlight again, you know, that that [15:46] we're incredibly olfactory animals. [15:48] >> We had a guest on this podcast who I'm [15:50] sure you're familiar Charles Zucker, a [15:51] professor at Columbia has done [15:52] incredible work in vision and olfaction [15:54] and [15:55] thirst sensing and I talked a little bit [15:56] about this that [15:58] there's something in the breath [16:00] of romantic partners [16:02] that's hopefully appetitive, not [16:04] aversive. [16:06] As well as in children, he was talking [16:08] about the smell of his grandchild's the [16:10] back the nape of their the back of their [16:12] neck and how he misses that smell [16:14] because it when he thinks about missing [16:16] his grandchild or children, it's that [16:19] smell that that that's associated with [16:21] that feeling. [16:22] >> hexadecanal. [16:23] >> Hexadecanal? [16:24] >> Yes. This is a study [16:26] uh ran by uh Eva Mishor uh who was a [16:30] graduate student in our lab. Eva was [16:32] interested in aggression. And so, when [16:34] she started off, we said, "Okay, let's [16:35] do chemo signaling of aggression." [16:38] She actually was going to like MMA clubs [16:40] and collecting body odors. [16:42] Uh and we we we had all sorts of ideas [16:45] going and and she's she worked on that [16:47] for quite a bit. It never went anywhere [16:49] really. At the same time, Heinz Breer, [16:52] he was studying in his lab uh a molecule [16:55] hexadecanal um that was a chemo signal [16:58] in mice that promotes social buffering. [17:01] So, when lots of mice are together, they [17:03] feel good about being in a group, and [17:05] that's social buffering. [17:07] And it's promoted by hexadecanal, which [17:09] they emit in their feces. So, Breer and [17:12] his colleague Stortzman, they they [17:13] discovered the receptor for this. And [17:16] then they went and discovered that the [17:17] receptor is very highly conserved uh [17:21] throughout mammalian evolution, and [17:23] therefore they hypothesized that maybe [17:26] this is a universal mammalian signal. [17:29] Now, which is unusual because in in [17:31] chemo signaling, typically you you tend [17:33] to think of things as being very [17:34] species-specific. So, he approached us [17:37] and said, "Look, you got to study this [17:38] stuff in humans, right?" Because he [17:39] knows us as the human people, right? I [17:41] mean, we go to these olfaction [17:43] conferences where where lots of people [17:45] study mice and and and and zebrafish and [17:48] whatnot, and we're the the human group. [17:50] And eventually, he just FedExed us [17:52] hexadecanal. And Eva was not going [17:54] anywhere with her aggression studies [17:57] with sweat from human participants. [18:00] And yet she built the entire um [18:03] paradigm to study human aggression. So, [18:05] they're standard paradigms. This is a [18:07] paradigm known as the TAP, the Taylor [18:09] Aggression Paradigm. And what's nice [18:11] about this is that then allows you to [18:13] quantify aggression. And so, you have a [18:15] a measure of aggression. Again, the [18:17] Taylor Aggression Paradigm, obviously [18:19] invented by Taylor, very well validated, [18:21] studied all over, you know, a very [18:23] standard protocol. [18:25] So, uh we brought in participants and [18:27] had them play uh the Tyler the tap uh [18:31] either under exposure to hexadecanal [18:34] or control. And I'll first tell you the [18:36] result with [18:37] men, which is that hexadecanal [18:41] consistently reduced aggression. [18:44] People were less aggressive under [18:45] hexadecanal. [18:47] Um the effect size was uh quite [18:49] meaningful compared to the effects seen [18:51] in the aggression world in in research. [18:53] Really, really strong effects. So, [18:55] unusually strong. [18:57] So, hexadecanal lowered aggression in [18:59] men, and we were like, "Cool, this is, [19:01] you know, sort of what we were hoping to [19:03] to see, consistent with the hypothesis [19:05] uh [19:06] and consistent with what it seems to do [19:07] in mice." But, then we looked at the [19:09] data from women. [19:11] And hexadecanal increased aggression, [19:13] equally significantly. Is this thought [19:15] to be something related to maternal [19:16] protectiveness? [19:17] >> We're getting there. [19:18] >> So, she went and did the entire [19:19] experiment again, and this time she did [19:21] it within the fMRI magnet so that we we [19:23] can also track uh brain activity uh [19:27] while this was happening. [19:29] And first of all, it replicated again. [19:31] So, once again, uh hexadecanal in men [19:33] made men less aggressive and women more [19:36] aggressive. And yet, I was telling you, [19:38] "You know, you know, but you know, this [19:39] makes no sense to me." [19:41] And then and then her insight, which of [19:42] course afterwards is like, "Duh." It is, [19:45] "No, there's a place where this makes [19:47] perfect sense. And that is if you're a [19:49] mammalian offspring." [19:51] Because [19:52] paternal aggression is often directed at [19:54] you. There's infanticide all over, and [19:58] sadly, there's male aggression towards [20:00] human children as well. [20:02] And maternal aggression is often [20:04] protective. [20:06] So, if you're an offspring, if you have [20:07] a molecule that will make your mother [20:09] more aggressive and your daddy less [20:11] aggressive, both of those are good for [20:13] you. So, you're winning. So, the bottom [20:15] line of all this is that hexadecanal is [20:18] the most abundant semi-volatile in baby [20:20] heads. It's tons of it coming out of [20:23] baby heads. Well, babies are conducting [20:25] chemical warfare, right? They're [20:27] reducing [20:29] aggression in their fathers or males [20:31] around them and increasing aggression in [20:34] their mothers or females around them, [20:35] and both of those things are good for [20:37] them. [20:38] I'd like to take a quick break and [20:40] acknowledge our sponsor, AG1. [20:42] I'm excited to share that AG1 has just [20:44] launched their newest formulation, AG1 [20:46] Pro. AG1 Pro takes the clinically backed [20:49] AG1 formula, which is a blend of [20:51] vitamins, minerals, probiotics, and [20:52] adaptogens, and adds three important new [20:55] ingredients: creatine monohydrate, [20:57] calcium HMB, and zinc carnosine. Each [21:00] serving has 5 g of creatine monohydrate [21:02] to support muscle strength and [21:03] performance, as well as brain health. [21:05] Calcium HMB to support muscle recovery [21:07] and reduce muscle breakdown, and zinc [21:09] carnosine to support and improve the [21:11] lining of your gut. All three of these [21:13] ingredients have compelling science to [21:14] support them, and therefore I love [21:16] seeing them added to the existing AG1 [21:18] formula. As most of you know, I've been [21:20] taking AG1 every day for nearly 14 years [21:23] now. I started taking it long before I [21:25] even knew what a podcast was. It's a [21:27] great product, and it's now made even [21:28] better with the new AG1 Pro formula. If [21:31] you would like to try AG1 Pro, you can [21:33] go to drinkag1.com/huberman [21:36] to get a special offer. AG1 is giving [21:38] away a free bottle of omega-3 coenzyme [21:41] Q10 with your first subscription. Again, [21:43] go to drinkag1.com/huberman [21:46] to get a free bottle of omega-3 coenzyme [21:48] Q10 with your first AG1 subscription. [21:51] What is the relationship between odors [21:54] and hormones, and in particular [21:57] crying? Tears are a bodily liquid, [22:01] emotional tears, [22:03] that that we emit in emotional [22:05] situations where where these are [22:07] situations where nonverbal communication [22:09] is is critical and key. The most [22:11] influential text I think till this day [22:14] in in in emotion research is is Darwin's [22:17] book The Expression of the Emotions in [22:18] Man and Animals, I think is the full [22:20] name of the book. [22:23] >> six [22:24] is devoted to tears. So, Darwin [22:26] systematically went through the [22:27] expressions of emotions and for each one [22:29] went to their functional antecedent and [22:31] explained everything very nicely. [22:33] And then he got stuck with tears, right? [22:35] Because tears are an obviously emotional [22:36] expression. [22:38] And he could not find [22:40] a functional [22:41] uh antecedent. [22:43] So, he ended up saying, "You know, this [22:44] is an epiphenomenon." [22:45] So, we thought, "Well, maybe the [22:46] function is is a chemical signal." So, [22:49] with that in mind, we we harvested [22:51] emotional tears. So, we collected uh [22:53] tears and we exposed uh participants to [22:56] these tears. First of all, the tears are [22:59] completely odorless. You cannot detect [23:01] them at all. Completely odorless. [23:04] Um [23:06] and yet [23:07] when you sniff them [23:09] you have [23:10] a pronounced uh reduction in [23:13] testosterone [23:14] uh within about 20 minutes, half an [23:17] hour. [23:18] >> This is men and women smelling women's [23:20] tears. Just men smelling women's tears, [23:22] but not perceiving any odor. [23:24] >> Nothing. Just sniffing them. [23:26] And you have uh about a 14% drop in free [23:32] testosterone. And people who study [23:34] testosterone, which is not me, but they [23:35] tell me this is a really strong effect. [23:37] Like it's it's hard to even [23:39] pharmacologically get an effect like [23:40] that that fast. I mean, no in [23:42] pharmacology. [23:43] >> Yeah, years ago I spent time studying [23:44] endocrine effects of this sort um and [23:46] it's a a tremendous re-size effect. [23:49] >> Uh and so, here I'll point out in [23:51] passing [23:52] uh that uh what one of the concerns we [23:54] had because of the uh [23:57] effort to run this study is that nobody [24:00] would ever try to replicate it. And to [24:03] our joy, about 2 years later, an [24:05] independent group from uh [24:08] South Korea [24:09] uh Oh et al., who I don't know at all, [24:12] uh [24:12] replicated the testosterone effect to a [24:15] T. I mean, like same numbers. [24:17] So, [snorts] so um it [24:20] it lowers testosterone. And and we then [24:22] also looked uh using MR at the at the uh [24:27] effect [24:29] on brain activity and saw pronounced [24:32] effect on activity, a dampening, a [24:34] lowering of activity [24:36] under under an arousing state, a [24:39] lowering of activity [24:41] both in the hypothalamus [24:43] um, and in the fusiform gyrus for [24:45] whatever reason, I don't know. [24:47] >> area. [24:47] >> Amongst other things, yes. [24:50] Um, and we don't know why. [24:52] Um, but pronounced. So, usually in our [24:54] chemo signaling work, we look at what [24:55] happens in rodents and we see if the [24:57] same thing is happening in humans. This [24:59] was a rare case where after we did this [25:01] work, more or less identical effects [25:02] were discovered in rodents. [25:04] So, a paper published in Nature 2 years [25:07] later found that mouse tears [25:10] uh, mouse pup tears lower aggression in [25:13] in um, male adult mice towards them. [25:16] >> In a smell dependent way. [25:18] >> Yeah, yeah, yeah. And they also actually [25:20] found the actual component in tears that [25:22] so, the tear pheromone that lowers [25:24] aggression. Right? So, so, you know, [25:26] this has us thinking of a gret is tears [25:28] is is is you can think of tears is like [25:30] a chemical blanket in a way that that [25:33] you're covering yourself up again with, [25:34] you know, to protect against aggression, [25:36] right? [25:37] >> Smell being such a rich source of [25:40] behavioral and hormonal and other sorts [25:43] of deep deep information, [snorts] [25:46] what is your experience with generating [25:48] smells in silico in computers? And here [25:51] folks, if for those of you that aren't [25:52] catching on to this and and I don't [25:53] expect that everyone would because what [25:55] we're really alluding to here [25:57] um, is the idea that you'll look at [26:00] you'll put into a search engine um, [26:03] blueberry pancakes recipe [26:05] and that not only will you get photos of [26:08] those blueberry pancakes and a recipe, [26:10] but you will get the [26:12] hopefully validated odor of those [26:14] pancakes and that recipe coming at you [26:17] in real time through the computer. [26:20] >> So, Google had an April Fools spoof. [26:24] >> All right. [26:24] >> And they put out this video [26:26] of Google smell. [26:28] Okay? And and it had all these like [26:31] classic like sales images of you know, [26:34] holding up your phone to a rose and it [26:36] generating rose and and and all these [26:39] things, right? [26:41] So, Google is now trying to do that, you [26:43] know, to digitize uh smell. And there [26:45] are other companies that are trying to [26:47] do this as well. And recently in our lab [26:50] we've had and and I hope this doesn't [26:51] come across as overly arrogant, but we [26:53] we've had a sort of mini breakthrough on [26:56] this front. [26:56] >> To call something a mini breakthrough is [26:58] far from arrogant. [26:59] >> [laughter] [26:59] >> And and and and this is a paper uh led [27:02] by Aaron Raviv uh [27:05] uh from our lab and Kobe Snitz also a [27:07] major contributor there. A paper [27:09] published in nature. And in this paper [27:12] we develop an algorithmic framework [27:14] uh where we can [27:17] predict the perceptual similarity [27:21] of any two molecular mixtures [27:24] with very very high accuracy. So, if you [27:27] give me two molecular mixtures, I can [27:28] predict how similar you you will smell [27:30] them to be. Okay? Not only could we [27:33] predict that, but we could design it. [27:34] So, we can generate um [27:37] uh [27:38] mixtures with known similarities. The [27:40] result was highlighted and and and [27:42] you'll appreciate this coming from [27:43] vision [27:44] uh is that [27:46] using our algorithmic solution, we [27:48] generated olfactory metamers. So, we [27:51] measured mixtures [27:53] completely non-overlapping in their [27:55] molecular structure [27:57] but they smell exactly the same. Okay? [27:59] Now, if you would come to a classic [28:00] perfumer or most classic perfumers and [28:03] tell them that you can generate two [28:05] mixtures with zero molecules in common, [28:08] but smell exactly the same, [28:11] they would tell you no. [28:13] And yet, we did and and anybody can [28:15] recreate them. This is simple, actually. [28:17] And in the paper we do a few things like [28:19] we generate a metamer for uh Chanel No. [28:21] 5 with no component from Chanel No. 5 in [28:24] it. [28:25] >> I'd like to take a quick break and [28:27] acknowledge our sponsor, Eight Sleep. [28:29] Eight Sleep makes smart mattress covers [28:30] with cooling, heating, and sleep [28:32] tracking capacity. Eight Sleep just [28:34] launched the Pod 6, which is their most [28:36] advanced sleep system. The Pod 6 is half [28:39] the size of their previous models. It's [28:40] 20% faster with its cooling and heating, [28:43] and it has nine times as many sensors [28:45] for tracking your sleep. The reason why [28:47] Eight Sleep is so effective at improving [28:49] your sleep quality and duration is [28:51] because it helps control your core body [28:53] temperature. And the reason for that is [28:55] that in order to fall and stay deeply [28:57] asleep, your core body temperature has [28:59] to drop by about 1 to 3°. And to wake up [29:02] feeling completely refreshed and [29:03] energized, you actually want your body [29:05] temperature to be increasing in the [29:07] final hours of your sleep. Eight Sleep [29:09] automatically regulates the temperature [29:11] of your bed throughout the night and is [29:13] monitoring your sleep and dynamically [29:15] adjusting the temperature of your sleep [29:16] environment so you get the maximum [29:18] amount of deep sleep and rapid eye [29:20] movement sleep. I've been sleeping on an [29:22] Eight Sleep mattress cover for nearly 5 [29:24] years now, and with each new iteration [29:26] of Eight Sleep, it gets better and [29:28] better at figuring out what my body [29:29] needs in order to get the best possible [29:31] night's sleep. If you would like to try [29:33] Eight Sleep, go to [29:34] eightsleep.com/huberman [29:36] to get up to $200 off their new Pod 6. [29:39] Eight Sleep ships to many countries [29:41] worldwide, including Mexico and the UAE. [29:44] Again, that's eightsleep.com/huberman [29:46] to save up to $200. [29:49] >> In a way, we've generated the the the [29:52] infrastructure for digitizing smell [29:54] because, again, what our algorithm [29:56] predicts our our framework predicts is [29:58] similarity. [29:59] But in a way, that's enough for you. Why [30:00] is that enough? We have a map [30:03] of 4,000 molecules. [30:07] Uh for each one, we know their perceived [30:09] smell. [30:10] Now, you can make up any mixture you [30:12] want for me. I can project it into that [30:14] map [30:15] and measure its pairwise distance from [30:17] all the points in the map. If it falls [30:20] on lemon, then what you generated smells [30:22] like lemon. And if it falls, you know, [30:24] on tomato, then what you generated [30:25] smells like tomato. [30:27] So, [30:28] we we now solved that problem. We can [30:31] predict the odor of any molecular [30:33] mixture. We we say how it's going to [30:34] smell. What we can do is then find a set [30:37] of components, which we call odor [30:39] primaries, [30:40] that can be used to mix any odor that [30:43] you can perceive. And that's what we're [30:45] working on now. So, this is in [30:47] collaboration with the lab of a Jonathan [30:49] Williams and Max Planck in Munich. [30:51] Um Jonathan Williams is an [30:54] atmospheric chemist, [30:56] uh [30:56] >> [snorts] [30:56] >> but he's really good at at using GCMS, [30:59] these tools that measure molecules. So, [31:03] uh Jonathan Williams measured odorants [31:05] in Germany, [31:06] uh transmitted the information to us [31:08] over IP. [31:10] Uh we fed that into our algorithmic [31:12] framework and recreated it from a device [31:15] that mixes primaries. And we tried to do [31:17] four different odorants [31:19] uh in our our proof-of-concept test. [31:22] Uh one of them uh was rose, and we [31:26] failed at recreating rose. [31:28] It we in fact recreated something that [31:30] had a percept, but most people perceived [31:32] it as bubblegum. [31:34] Uh the second one we tried to do was [31:35] anise, and we failed at recreating [31:38] anise. [31:39] Um and most people said it was cherry, [31:42] which is not very far, but it failed. [31:45] The third was gasoline, and we were [31:47] slightly but significantly better than [31:49] chance at recreating gasoline. And the [31:51] fourth was violets. [31:53] And 15 of 16 people said violets. [31:57] So, the first odor ever transmitted over [31:59] IP is violets. You know, we're don't [32:01] hold your breath for this to be on your [32:03] table tomorrow. You know, all we have in [32:05] hand is this very initial [32:07] proof-of-concept, but we're on we're on [32:09] the path. We're talking now of trying to [32:13] achieve [32:15] uh the olfactory equivalent of circa [32:17] 1956 black-and-white TV, okay? [32:20] Basically, right? And you know, I'm not [32:23] dreaming, let's say, of being able to [32:24] transmit to you the difference between a [32:26] cabernet or a merlot, right? But if I [32:28] can generate something that's vaguely [32:30] wine, that will be an amazing success [32:32] from my perspective, right? [32:34] But jump ahead in your imagination [32:37] to to 4K odor transmission then medical [32:40] diagnostics is where you want to be [32:42] talking about because this is uh uh [32:45] over uh extension but [32:48] you can almost say that every disease [32:50] will have an odor. I mean every disease [32:52] is a specific metabolic process. [32:54] Metabolic process have metabolites, [32:56] metabolites have a smell. Olfaction once [32:59] it's digitized [33:00] and and high resolution which again it [33:04] in our hands it's not going to be I mean [33:05] we're talking you know in my retirement [33:07] maybe I'll read about this one day if [33:09] I'll still have vision. I mean this is [33:11] not close. [33:13] But [33:14] when when olfaction digitization is [33:17] brought to the equivalent of 4K uh uh [33:20] vision and audition that you have now [33:22] then it will be in medical diagnosis. [33:24] Excuse me for the uh imagery but you [33:27] will have an electronic nose in your [33:28] bathroom. [33:29] Uh each one of us will have in the [33:31] toilet. And it will be doing diagnostics [33:34] all the time. And and uh that's that's [33:37] where it's going to go. But again not [33:40] anywhere in the very close future. [33:43] >> Now my I want to say thank you for your [33:46] time today. First of all uh this was a [33:49] tremendously interesting conversation. I [33:52] mean we touched on so many things [33:53] hormones, smells, the architecture of [33:55] the olfactory system. I know that people [33:57] listening to this are realizing but I'm [33:59] going to say it anyway uh what an [34:01] incredible gift you've given us in um as [34:04] a as a [34:06] expert in this field giving us this tour [34:08] of of the work that you and others uh [34:11] who you credit so generously um have [34:14] done and the spirit that you bring to [34:15] it. Whatever drives that spirit as the [34:18] great late Ben Barres used to say keep [34:20] going because we are all benefiting [34:22] tremendously. And I also just want to [34:24] say that you know for people listening [34:26] to this the the spirit of [34:28] science is one of as you mentioned [34:30] there's complex politics and all these [34:32] things but it's absolutely clear that [34:35] you delight in the work you do. And um [34:37] so I delight in it. I'm grateful for it. [34:39] I'm grateful for your time today. And so [34:41] on behalf of me and and the many, many [34:43] people listening to this, I just want to [34:44] extend a huge debt of gratitude. Thank [34:46] you so much. [34:48] >> Thank you so much for [34:50] very [music] warm words. [34:55] >> [music]