Sensory Expansion

Frontier The prize is a sense the species never had, delivered at the acuity of the organs it already owns and processed as perception rather than read off a display — but only if the new input is integrated rather than merely detected, and that condition is the whole brief. If a novel channel participates in the machinery of perception the way the native senses do, the reward is an added dimension of experience at no cost to the existing sensorium; if it does not, the reward is a gadget that reports numbers through the eyes. The bottleneck list in section 5 and the test in section 7 decide which.

Established Three 2019–2025 results show the channel can at least be installed and detected. The “Oz” system stimulated on the order of 1,000 individual M cones in five human subjects and produced “olo,” a blue-green of a saturation no natural or artificial light can match — a colour outside the natural gamut, corresponding to the receptor input (0, 1, 0) that daylight never delivers. A single injection of upconversion nanoparticles gave mice image-forming near-infrared vision, converting 980 nm to a 535 nm percept for over 10 weeks alongside intact visible vision. Upconversion contact lenses let human subjects read flickering near-infrared signals in the 800–1600 nm band, better with their eyes closed.

Established Measured against the existing route, that is the specific gain. Sensory substitution — the tongue display, the vOICe soundscape — reroutes new information through an already-used channel, needs weeks of training, and is abandoned at high rates outside the laboratory. Sensory expansion adds a channel that was not there, and the near-infrared work adds it without spending the visible one: both modalities operate at once.

Speculative Who collects it is narrow, and later than the headlines imply. Five people have seen olo, inside a 9-by-13-foot room, biting a fixation bar; the near-infrared lenses resolve no fine detail and need an infrared source to point at. The first beneficiaries are vision scientists, then perhaps night-operations and colour-critical specialists, not the public this decade. Frontier The upside is bounded by a quantity no study has yet reported: whether a genuinely novel channel adapts, produces aftereffects, and enters cross-modal illusions — the mark of a sense rather than a signal.

Established Sensory expansion means installing a perceptual channel the species never possessed, which is a different problem from restoring one that was lost. Sensory Restoration repairs a broken link in an existing pathway — a missing protein in a surviving hair cell, a dead photoreceptor replaced by an electrode — and can lean on the wiring evolution already built. Expansion has no template: the input is one the visual, auditory or somatosensory system was never selected to interpret, so the load-bearing question is not whether the signal arrives but whether the brain treats it as perception.

Established Four routes are on the table and they differ in where the new information enters. The oldest is sensory augmentation or substitution: an external transducer feeds a novel quantity — magnetic north, ultrasound range — into an intact channel such as touch or hearing. The second converts an out-of-band stimulus at the receptor itself, as upconversion nanoparticles turn near-infrared into visible green on the retina. The third drives receptor subpopulations directly, as the Oz system stimulates M cones in isolation. The fourth is genetic: add a photopigment and let the retina carry a signal it never had, as in opsin gene therapy.

Frontier Across all four, the field divides on one axis: detection versus integration. Detection is easy to demonstrate and has been, repeatedly — a subject reports a percept, discriminates a shape, presses the correct button. Integration is the harder and mostly unmeasured property: does the new channel adapt over time, produce aftereffects, bias the other senses, and get used automatically without deliberate decoding. The natural senses do all of this; whether an installed one does is the question every section below returns to.

Established Scope, because adjacent briefs own neighbouring ground. Sensory Restoration owns the electrode, photodiode and gene-therapy hardware this brief borrows; Neural Interfaces owns cortical writing and channel counts; Genetic Engineering owns opsin delivery and capsid biology; Neuroplasticity Engineering owns the adult-plasticity substrate that integration depends on. What this brief takes is the joint question none of them owns: when you add a channel the organism never had, does it become a sense?

Established A note on sourcing. This brief was commissioned in September 2026. Where a primary paper sat behind a publisher that refused retrieval, it is cited by name and the claim is carried by a fetched secondary report or by the authors’ own institutional account; figures with argumentative weight are dated no later than mid-2025.

Established The strongest expansion result to date is the Oz system and the colour olo, reported by a University of California, Berkeley group in Science Advances in April 2025. An adaptive-optics rig scanned each subject’s retina, mapped individual cones, and delivered 543 nm laser microdoses — ordinarily seen as green — at roughly 10⁵ pulses per second to a population of about 10³ cones, stimulating M cones while sparing the L and S cones beside them. Established Because natural light never excites M cones in isolation, the brain received the receptor combination (0, 1, 0) it has no daylight precedent for, and five subjects reported a blue-green of, in the group’s words, unprecedented saturation. Established The validation was a colour match: subjects could reach the nearest monochromatic light only after desaturating olo with added white, placing it outside the natural gamut as a measured fact rather than a description.

Established Oz is a general programmable platform, not a one-colour trick. Beyond olo the group rendered a range of hues — orange, yellow, green, blue-green — and moving imagery such as red lines and rotating dots on an olo background, which is why the senior author calls it ready to set the activation of thousands of photoreceptors at will.

Frontier What Oz demonstrates is a stable, matchable percept from a never-occurring input; what it does not demonstrate is integration. The senior author states the platform is “ready for broad scientific work” and can programmably set the activation of thousands of photoreceptors, which is a claim about control, not about whether olo behaves like a colour — whether it adapts, produces an aftereffect, or shifts adjacent hues. Frontier Those experiments have not been reported, and until they are, olo is a demonstrated novel detection rather than a demonstrated new sense.

Established The near-infrared line is the clearest case of a channel added to an intact sensorium. In 2019 a group at the University of Science and Technology of China and the University of Massachusetts (Ma, Bao and colleagues, with Xue and Han) injected core-shell upconversion nanoparticles — roughly 38 nm β-NaYF₄:Yb,Er conjugated to a lectin that binds photoreceptor glycoproteins — into mouse eyes. Established The particles converted 980 nm near-infrared to a 535 nm emission the retina reads as green; mice showed pupillary constriction to infrared, cortical visually-evoked potentials, and learned to discriminate infrared shapes such as triangles from circles in a water maze, with pattern vision lasting over 10 weeks and normal daylight vision preserved throughout. Frontier The reported side effects were minimal, but the sample is a laboratory rodent and the safety margin in a human eye is unestablished.

Established The 2025 human extension moved the near-infrared channel out of the retina and onto the cornea. A group led by Tian Xue reported in Cell that upconversion nanoparticles embedded in the flexible, non-toxic polymer of a standard soft contact lens let human subjects detect near-infrared in the 800–1600 nm band: they read flickering, Morse-code-like temporal signals and discriminated the direction of an incoming infrared beam. Established Detection improved with the eyes closed, because near-infrared penetrates the eyelid better than visible light does — a clean demonstration that the percept rides the new physics rather than stray visible leakage. Established A three-band version mapped 980 nm to blue, 808 nm to green and 1,532 nm to red, giving a coarse false-colour infrared. Frontier The lens resolves no fine spatial detail — the converted light scatters over the short distance to the retina — and works only against a projected infrared source; a spectacle-mounted version reached higher resolution.

Established The augmentation route has the longest record and the most direct evidence on integration, and it points both ways. The feelSpace belt — sixteen waist-mounted vibromotors signalling magnetic north — was worn by nine subjects for seven weeks at about 1.57 hours per day in a 2016 PLOS One study; eight of nine reported a new, spatial, increasingly automatic sense (“space is getting wider and deeper”), functional-MRI showed training-related change in parietal, hippocampal and caudate regions, and REM sleep rose transiently as in procedural learning. Frontier Yet a formal homing task showed no significant improvement (p > 0.05), so the strongest evidence was subjective report and neural correlate rather than a hard behavioural gain.

Established The single most important integration result is a 2017 Scientific Reports study that made an augmented signal recalibrate an existing sense. Twenty-seven blindfolded subjects were rotated in a chair while an auditory compass signal, gain-scaled, accompanied the motion; afterwards, with the signal switched off, their felt magnitude of passive self-rotation had shifted — a slope change of −0.21 under gain 0.5 and an intercept increase of 27.5° under gain 2 (both p < 0.001), with the recalibration persisting to a second day. Frontier Because the signal was absent during the test, the effect cannot be a conscious counting strategy; it is a genuine perceptual aftereffect of a novel channel. Frontier It is also the closest thing on record to the decisive test in section 7, and it used an augmented signal fed through hearing, not a novel receptor channel.

Established The genetic route has an animal proof and no human novel-channel result. In 2009 Mancuso and colleagues delivered a human long-wavelength opsin gene by adeno-associated virus to adult dichromatic squirrel monkeys; over about five months the treated animals gained trichromatic colour-discrimination behaviour on an animal Cambridge Colour Test, and the effect persisted for at least two years. Frontier The accompanying Nature commentary called it a new dimension of colour vision and evidence of extreme adult plasticity, while conceding that how the new photoreceptor signal becomes colour perception “is a fascinating question” — that is, behavioural discrimination was shown and a new percept was inferred, not proven. Established Earlier knock-in mice expressing a human cone pigment (Jacobs and Nathans, 2007) likewise gained a new colour-discrimination dimension, establishing that a mammalian visual system can extract information from an added photopigment it never evolved with.

Established Human biology already runs the control experiment, and it argues that having the receptor is not enough. Roughly half of women are heterozygous carriers of an L-cone polymorphism and so carry the genetic basis for a fourth cone class; a UK estimate put carriers of hybrid opsin genes near 3.9 million but those with the optimal 12 nm spectral separation at about 48,585. Established When Jordan and colleagues tested carriers directly in 2010, only one subject (cDa29) of 24 deuteranomaly carriers showed four-dimensional colour vision across all tasks; the rest, with the extra receptor present, discriminated colour in three dimensions like everyone else. Frontier The lesson the expansion field cannot dodge is that an added channel is usually not exploited — integration, not installation, is the rate-limiting step.

Frontier Does a novel channel become perceptual, or stay cognitive? A subject who presses a button when olo appears has detected it; a subject in whom olo produces a colour aftereffect, tilts an adjacent hue, or is mistaken for a similar colour under distraction has perceived it. No published experiment has run those tests on olo or on near-infrared vision, so the field’s central claim — that these are new senses — rests on a property nobody has measured.

Frontier What determines whether the brain exploits an extra channel? The tetrachromacy natural experiment says spectral separation, receptor numerosity and the weighting of chromatic signals to cortex all matter, and that most carriers of a fourth cone never build the fourth dimension. Whether targeted training, developmental timing or a stronger signal can move a latent channel into use is open, and it is the difference between a curiosity and a capability.

Frontier Is there a critical period for a genuinely new modality? Adult squirrel monkeys and adult-relevant mouse work suggest the visual system can use an added photopigment without a developmental window, which is surprising and encouraging. It does not follow that a modality with no evolutionary precedent at all — a magnetic sense, an electric sense — will find a cortical home as readily as a variant of colour, which sits on machinery already built for wavelength.

Speculative Is olo a new experience or a relabelled old one? Pure M-cone stimulation is a point the natural gamut cannot reach, but the percept is still built from the same opponent channels; whether that counts as a new quale or an extreme setting of an existing one is a question Consciousness Research owns and this brief cannot settle. It matters because “a colour you have never seen” and “the most saturated teal possible” are different claims.

Frontier How much information can an installed channel actually carry? The near-infrared lens delivers direction and flicker but no fine form; the magnetic belt delivers one bearing. A channel with a handful of distinguishable states is a sense in name; whether upconversion or targeted stimulation can reach the bandwidth that would make a new modality worth living with is unquantified.

Frontier The deepest bottleneck is that nobody can predict when a new input becomes a percept. There is no model that takes a signal’s statistics, a training regime and a target channel and says whether the result will adapt, produce aftereffects and integrate, or remain a decoded readout. Without it, every expansion project is an empirical gamble, and the tetrachromacy record warns that the default outcome is non-integration.

Established The delivery hardware is where the demonstrations become fragile. Oz needs a room-sized adaptive-optics bench, per-cone retinal tracking and a bitebar to hold the eye still, and it addresses a retinal patch far smaller than the visual field; it is an instrument, not a wearable. The near-infrared contact lens is wearable but resolves no fine detail because the converted light scatters over the millimetres to the retina, and it needs a bright, narrowband infrared source in view.

Frontier Ambient-light upconversion is the materials problem gating a useful infrared sense. Current nanoparticles respond efficiently to concentrated near-infrared such as an LED, not to the faint broadband infrared of a night scene, and pushing quantum yield up while widening the absorption band without heating the eye is unsolved. Until it is, “infrared vision” means “vision of infrared beacons.”

Frontier The genetic route inherits every constraint of ocular gene therapy plus one of its own. Adding a photopigment needs a capsid that reaches the target cones, a coding sequence within packaging limits, and a spectral class separated enough from the natives to carry independent information — the tetrachromacy data suggest roughly a 12 nm separation is near the useful floor. Genetic Engineering owns the vector; the spectral-tuning and retinal-tiling problem is specific to expansion.

Speculative Bandwidth is a quieter ceiling. A channel that resolves a few states cannot become a rich modality however well it integrates; the engineering that raises the number of distinguishable percepts — more cone subtypes addressed, finer temporal or spatial structure — is barely begun, and it is not obvious the cortex would carve out the representational room for it.

Frontier Everything here depends on adult cortical plasticity, and the amount available is unsettled. The monkey and mouse opsin results say a visual system can use an added pigment in adulthood; the tetrachromacy record says the same system usually does not exploit a channel it has carried since birth. Neuroplasticity Engineering owns the mechanisms; expansion needs to know the dose-response of plasticity to a novel input, which nobody has mapped.

Established The genetic route depends on vector delivery it does not control. Capsid tropism, packaging limits, pre-existing immunity and the class safety record are all set by Genetic Engineering and imported wholesale; an opsin-addition therapy is only as safe and deliverable as the ocular gene-therapy platform underneath it.

Established The optical routes depend on nanomaterials and adaptive optics from elsewhere. Upconversion efficiency and biocompatibility come from the nanoparticle field that Nanomedicine tracks; per-cone stimulation depends on the adaptive-optics and retinal-tracking stack built for imaging and for Sensory Restoration. Neither was built for expansion, and both cap what it can attempt.

Frontier The most orphaned dependency is the psychophysics of integration. Aftereffect and cross-modal-illusion protocols are standard tools in vision science, but nobody funds them as the acceptance test for a new sense, so the one measurement that would tell an expansion project whether it succeeded is the one no programme is resourced to make.

Frontier The decisive experiment is not another demonstration that a novel channel can be detected; it is whether that channel participates in the machinery of perception. The battery is standard and specific: does olo produce a colour aftereffect or a McCollough-style contingent aftereffect; does near-infrared input feed a cross-modal illusion such as the sound-induced flash; does either channel adapt so that prolonged exposure shifts the appearance of the senses around it. A channel that does these things is integrated; one that does not is a readout the subject learns to interpret.

Frontier This experiment needs no new hardware and has not been run on a novel receptor channel. The Oz bench can hold olo steady long enough for an adaptation protocol, and the upconversion contact lens is a wearable infrared source; what is missing is the psychophysics, not the apparatus. Its absence is why the field’s new-sense claims currently rest on detection alone.

Established The nearest completed version already returned a positive result, in a different modality. The 2017 auditory-compass study made an augmented signal recalibrate felt self-rotation, an aftereffect that persisted with the signal switched off. Frontier It proves an installed channel can bias an existing sense; it leaves open whether a channel entering through a novel receptor, rather than through hearing, does the same.

Frontier A longitudinal training study is the honest second experiment. Take subjects with the near-infrared lens or repeated Oz access, train them for weeks as the feelSpace protocols do, and measure automaticity, aftereffects and cross-modal transfer over time rather than at a single session. The feelSpace three-stage work reports subjective integration but, by its own admission, collected no aftereffect or cross-modal data; supplying that is the missing half.

Established The tetrachromacy natural experiment is already running and should be read as a control. Carriers of a fourth cone class are, in effect, humans with a pre-installed extra channel; the finding that only one of 24 tested exploited it sets the base rate an engineered channel must beat and identifies spectral separation and cortical weighting as the variables to manipulate.

Speculative The animal electrophysiology that would close the loop is a targeted recording study. In an opsin-added monkey or mouse, does the novel photopigment’s signal acquire its own cortical representation, or is it read out through the existing channels? Frontier That measurement would tell the human genetic route whether it is building a new dimension or a louder version of an old one, and it has not been reported.

Established The Oz route is an adaptive-optics engineering problem before it is anything else. Cell-by-cell stimulation requires mapping the cone mosaic, tracking eye motion at the scale of single cones, and gating a 543 nm beam onto chosen cells at roughly 10⁵ pulses per second; today that means a fixed bench, a bitebar and a small retinal field. Shrinking any of it toward a deployable device is the entire engineering roadmap, and none of it is close.

Established The upconversion route trades power against wavelength against safety. Nanoparticle emission must land near the eye’s 550 nm sensitivity peak, absorb in a band a real scene supplies, and do so without heating or clouding the tissue; the contact-lens form factor solves biocompatibility and preserves visible vision but sacrifices resolution to scatter. The spectacle-mounted variant recovers resolution at the cost of being a headset rather than a lens.

Frontier The genetic route is a spectral-design problem layered on ocular gene therapy. Beyond the vector, the added opsin must be tuned to a peak far enough from the natives to carry independent colour information and expressed in a retinal subpopulation dense enough to sample the scene; this is the same delivery hardware Sensory Restoration and Genetic Engineering describe, aimed at addition rather than repair.

Frontier Every route eventually meets the encoding question the electrode fields already hit. If expansion moves upstream to cortical writing, it inherits the problem Neural Interfaces owns: nobody can specify the stimulation pattern that yields a target percept. Staying at the receptor, as olo and upconversion do, is precisely how these results sidestepped an unsolved cortical code — and it is why they cannot yet reach an arbitrary new modality.

Established Sensory Restoration is the mirror image and the hardware donor. It restores channels that existed and can reuse evolved wiring; expansion installs channels that did not and has no wiring to reuse. The photodiode arrays, adaptive optics and gene-therapy vectors are shared, which is why expansion advanced fastest exactly where restoration had already built the tools — at the retina.

Frontier Sensory substitution is the adjacent field most often mistaken for this one. Devices such as the vOICe soundscape and tongue displays route new information through an intact sense; a 2015 study found naive subjects match some soundscapes to images above chance without training, and a 2025 study found arbitrary image-to-sound mappings learn about as well as intuitive ones — evidence the process is partly learned association, not automatically perception. Substitution reaches perception-like automaticity in some hands and is abandoned by many; it is the cautionary baseline expansion should be measured against, not conflated with.

Established Genetic Engineering supplies the only route that needs no external device. An added opsin, once delivered, is carried by the organism; the animal results live or die on that field’s vector safety and expression control, and any human novel-channel gene therapy will arrive from there rather than from optics.

Frontier Neuroplasticity Engineering and Consciousness Research bracket the open question. The first asks whether the adult brain can be induced to allocate representational capacity to a novel input; the second asks whether the result is a new experience at all. Expansion sits between them and borrows from both.

Speculative The enhancement framing connects it to Biological Enhancement and Human-Machine Symbiosis. A successful new sense is, unlike restoration, elective augmentation of a healthy person, which places it in those briefs’ territory for questions of who chooses it and what living with it means.

Established There is no regulatory category for elective sensory augmentation, because every existing pathway assumes a disease. Device and gene-therapy approval turn on treating or diagnosing a condition; a healthy person who wants an added channel fits no indication, so the near-term legal home for expansion is either an off-label stretch of a therapeutic approval or the unregulated consumer and body-modification market. Neither is built for an irreversible change to a healthy sensorium.

Frontier Reimbursement and market structure have no answer for a non-therapeutic sensory device. Payers fund treatments, not upgrades; a one-time gene therapy that adds a colour dimension has no cost-offset story and no reimbursement path, so the first products will be self-funded, which sets who gets them before any equity question is asked.

Frontier Combination products sit across two regulators at once. An upconversion contact lens is a device carrying an active nanomaterial; an optogenetic or opsin approach is a biologic plus, often, a light source. The restoration field already found that combination products have no clean regulatory home, and expansion inherits that gap with the added complication of a non-therapeutic purpose.

Speculative An unregulated do-it-yourself community is already the field’s uncontrolled arm. Magnet implants and subdermal sensors are installed outside medical oversight today; as expansion tools cheapen, the gap between a lab that reports five carefully studied subjects and a marketplace that ships thousands unmonitored is an institutional risk, not a hypothetical.

Frontier Expansion is enhancement of the healthy, which changes the consent and justice questions. Sensory Restoration flags that the boundary between restoring a sense and extending one has not been forced by any product; expansion is where it gets forced. Nothing in the approval framework distinguishes giving sight back from giving a new sense, and the first case to reach a regulator will have to.

Established Irreversibility raises the stakes for the genetic route specifically. A gene therapy that adds a photopigment cannot be undone and, like the ocular platform it borrows, cannot readily be re-dosed; a device can be removed, an injected opsin cannot. Consenting a healthy adult to a permanent, poorly reversible change of perception is a different act from consenting a patient to a treatment for a deficit.

Frontier Access will stratify sharply because there is no reimbursement. A capability sold only to those who can pay, in a domain as fundamental as perception, is a cleaner engine of advantage than most enhancements; Cognitive Liberty owns the rights framing, and the specific worry here is a perceptual divide that tracks wealth.

Speculative There is a subtler harm if integration works too well. A sense you cannot switch off — an always-on magnetic pull, an infrared overlay — may be unwanted once novel, and the feelSpace reports of feeling “insecure without the belt” hint that removal is not costless. Designing for reversibility of the experience, not just the hardware, is an ethical requirement the field has not addressed.

Frontier The near-term civilizational effect is small and instrumental: a new tool for vision science. Oz is, for now, an instrument that lets researchers set receptor activation directly, which will teach more about colour and plasticity than it changes daily life. Treating that as the honest scope keeps the larger claims where they belong — conditional on integration.

Speculative If integration is solved, the effect is a widening of the human umwelt, adopted unevenly. The plausible first users are occupational — night operations reading infrared beacons, colour-critical work using an extra dimension — not the public, and a population split between people who share a sense and people who do not is a novel kind of social difference, closer to language than to eyewear.

Speculative Military and industrial adoption would likely lead, as it has for night vision. A wearable that adds infrared without spending visible vision has obvious operational value, and the incentives there do not wait on the perceptual-integration question the science cares about; a channel that is merely detectable is already useful to an operator.

Handwave The largest claims — senses that let humans perceive magnetic fields, radio, or whole new modalities as richly as sight — assume the integration problem and the bandwidth problem are both solved, and then describe the consequences. The assumption is the content of the claim, and section 5 is why it is not yet earned.

These horizons track measured percepts and integration evidence rather than device generations, because the demonstrations have outrun the science of whether they are senses.

• 10 yr: Frontier Aftereffect and cross-modal tests are finally run on olo and on near-infrared vision, and settle whether either integrates; Frontier upconversion eyewear for infrared beacons finds an occupational niche; Speculative a first human opsin-addition study aimed at a new dimension, rather than at restoring dichromats, is attempted.
• 25 yr: Speculative Ambient-light upconversion, if the materials problem yields, turns “infrared beacon vision” into “infrared scene vision”; Speculative a training protocol reliably moves a latent or installed channel into perceptual use, beating the tetrachromacy base rate; Frontier a regulatory category for elective augmentation exists, or the field stays in the consumer grey market.
• 50 yr: Speculative A genuinely non-visual, non-auditory modality — a magnetic or electric sense with real bandwidth — is integrated, or the evidence says the cortex will only extend senses it already has; Handwave population-scale adoption would require a safety and reversibility record nothing on this page yet supports.
• 100 / 250+ yr: Handwave Claims of arbitrary designer senses written directly to cortex assume the encoding problem Neural Interfaces flags as unsolved is solved; the assumption carries the entire prediction.

• Depends on Neuroplasticity Engineering for the adult plasticity that would let a novel input claim representational room, Genetic Engineering for the vector and opsin delivery the genetic route needs, and Sensory Restoration for the retinal hardware and adaptive optics it borrows. No result on this map blocks the detection demonstrations — olo and near-infrared vision already exist; the blockers below govern whether they become senses.
• Requires (not on this map) six things in the order they bind. A perceptual-integration test for novel receptor channels — aftereffects, cross-modal illusions, adaptation — is the missing acceptance criterion without which every result is mere detection. A model predicting when a new input becomes perceptual would turn expansion from a gamble into engineering. Broadband ambient-light upconversion nanomaterials are what separate infrared-beacon vision from infrared-scene vision. A deployable single-cone stimulation device is what would take Oz off its bench. A regulatory category for elective sensory augmentation is the legal home no disease-based pathway provides. A market pathway for a non-therapeutic sensory device decides who gets one before equity is even discussed.
• Enables an added perceptual dimension at the acuity of the native organs; occupational infrared perception without spending visible vision; and, at the far end, modalities with no evolutionary precedent — each conditional on a constraint above rather than on a result here, so no typed enabling edge is claimed.
• Adjacent Sensory Restoration, Neural Interfaces, Genetic Engineering, Neuroplasticity Engineering, Nanomedicine, Consciousness Research, Biological Enhancement; and off the map, vision science, psychophysics and the body-modification community.

Frontier “Scientists let humans see a brand-new colour.” Five subjects, inside a room-sized adaptive-optics rig biting a fixation bar, saw olo when about a thousand of their M cones were driven in isolation. It is a real, measurable percept outside the natural gamut and it is a demonstration of detection; whether olo behaves like a colour — adapts, produces aftereffects — is untested.

Frontier “Infrared contact lenses give you night vision.” They let wearers detect near-infrared from an LED-class source, read its flicker and judge its direction, with no fine detail because the converted light scatters. That is beacon detection added alongside normal vision, not the passive scene vision the phrase “night vision” implies.

Frontier “Gene therapy can just add a new sense.” Adult monkeys gained trichromatic colour-discrimination behaviour from an added opsin, which is a strong result; the commentary that announced it conceded that how the signal becomes perception is unresolved. Behavioural discrimination was shown; a proven new percept was not.

Established “The brain will integrate whatever you feed it.” Roughly half of women carry the genetic basis for a fourth cone, yet in direct testing only one carrier of 24 showed four-dimensional colour vision. The extra channel is usually present and usually unused; integration, not installation, is the hard part, and assuming it away is the field’s most common error.

Frontier “Sensory substitution already gives people new senses.” Substitution devices route new information through an intact sense and reach automaticity in some trained users, but recent work shows arbitrary mappings learn about as well as intuitive ones, which points to learned association as much as to perception. It is a useful assistive baseline, not a settled case of a new sense, and it is not what expansion attempts.

Speculative “This is basically sensory restoration for people who never had the sense.” Restoration can reuse wiring evolution built; expansion has no template, which is why an installed channel so often stays a signal rather than becoming a sense. The two share hardware and almost nothing else about the hard problem.

Handwave “Animals have dozens of photoreceptors, so super-senses are easy for us.” A mantis shrimp carries many opsin classes and does not, on the evidence, out-discriminate colours the way the count implies; more receptors do not equal richer perception, and the inference from animal receptor counts to easy human super-senses skips every integration step this brief is about.