Daniel sent us this one — and it's basically a two-part challenge. First, he's been following the research shift on blue light, and he's noticed something interesting: the growing consensus that light intensity matters more for melatonin suppression than the specific wavelength. Which means the practical advice flips from "buy expensive amber glasses" to something much simpler — get outside in the morning, and at night, just turn the lights way down. But here's the catch. He's tried that. He's dimmed his phone to minimum. Still too bright. Same with monitors, same with TVs, same with smart bulbs. So the real question is: why is minimum brightness still so bright? What's the electrical engineering bottleneck? And for someone who wants genuinely low light without hacking firmware, which manufacturers are actually solving this?
This is one of those questions where the answer is way more interesting than people expect. Because the short version is: LEDs are terrible at being dim. They're fundamentally not designed for it. And the workarounds create problems that most manufacturers would rather you not notice.
Minimum brightness being too bright isn't a design oversight. It's a physics problem wearing a trench coat.
And the trench coat is PWM. Pulse width modulation. Which is the dominant method for dimming LEDs, and it's the source of basically everything we're about to talk about.
Walk me through it. I know PWM is pulsing the light on and off really fast, but why is that the go-to method?
Because LEDs don't dim the way incandescent bulbs do. An incandescent filament is gloriously simple — you reduce the current, the filament gets cooler, it emits less light, and the color shifts toward warm amber. It's a continuous, analog process. The light is always on, just at different intensities. LEDs are completely different. An LED is a diode — it has a threshold voltage. Below that threshold, it doesn't emit light at all. Above it, it emits light at full efficiency. There's a tiny sweet spot where you can vary the current and get variable brightness, but it's narrow, it's nonlinear, and it causes color shift. The LED's spectral output literally changes depending on how much current you're feeding it.
You can't just turn the knob down. It's either on or off, basically.
And "basically" is doing a lot of work there, but that's the fundamental challenge. So the industry settled on PWM as the solution. Instead of reducing current, you switch the LED on and off thousands of times per second. If it's on for fifty percent of the time and off for fifty percent, your eye perceives it as half brightness. If it's on ten percent of the time, it looks like ten percent brightness. The LED itself is always running at full power during those on-pulses, so the color stays consistent. Problem solved, right?
Except I'm guessing the problem isn't solved, because Daniel's phone at minimum brightness is still lighting up his bedroom like a landing strip.
First, at very low duty cycles — say, one percent or below — the pulses get so short that the electronics struggle to deliver them cleanly. You get jitter, you get inconsistent pulses, you get flicker that some people can perceive and others can't. Second, and this is the part that connects directly to the prompt — even at one percent duty cycle, a modern LED is still really bright. These things are efficient. A phone screen LED at one percent PWM might still be pumping out several nits of brightness, which in a completely dark room is absolutely noticeable.
That's the unit for screen brightness?
Yeah, nits — candelas per square meter. A modern phone screen at full brightness might hit eight hundred to a thousand nits outdoors. Minimum brightness on most phones is around two to four nits. Which sounds low, but in a dark-adapted bedroom at midnight, two nits is a lighthouse. Your fully dark-adapted eye can perceive fractions of a nit.
The floor is still too high. And Daniel mentioned an app called Darker Pro that somehow goes below that floor. How does that even work if the hardware can't do it?
This is clever. Darker Pro and similar apps use a software overlay trick. They don't actually change the backlight brightness at all. What they do is draw a translucent black overlay on top of everything — it's literally a screen filter. The backlight stays at whatever the hardware minimum is, but the pixels themselves are told to display a darker image. So the effective brightness reaching your eyes is lower than what the hardware alone can achieve.
That's a hack, not a solution. You're losing contrast, you're probably messing with color accuracy.
And on some display technologies, particularly OLED, it can introduce weird artifacts because of how OLED handles near-black pixels. You get something called black crush, where subtle dark grays all collapse into pure black and you lose shadow detail. It's the visual equivalent of an audio compressor set way too aggressively.
New term for me. So the software overlay is a band-aid, and the real fix has to come from the hardware side. Which brings us to the engineering challenge. What's actually hard about making an LED go dim?
Let me give you the three specific bottlenecks. First is the PWM resolution problem. PWM is controlled by a counter in the driver chip. If you want to dim to one percent, you need the counter to be able to distinguish one part in a hundred. That's easy. But if you want to dim to zero point one percent, you need one part in a thousand resolution. Zero point zero one percent, one part in ten thousand. The driver chips in most consumer electronics use eight-bit or ten-bit PWM, which gives you two hundred fifty-six or one thousand twenty-four steps. That sounds like a lot, but when you're trying to cover the range from eight hundred nits down to zero point one nits, you need more like fourteen-bit or sixteen-bit resolution. That requires more expensive driver chips, more careful board layout, and better firmware.
Nobody's putting a sixteen-bit PWM controller in a two-hundred-dollar monitor because the market doesn't demand it.
And the market doesn't demand it because most people don't use their monitors in pitch-black rooms at midnight. But the second bottleneck is even more fundamental. It's called LED quantum efficiency droop at low currents. Even if you do decide to use current control instead of PWM — which some high-end systems do — LEDs become less efficient at very low currents. The light output doesn't scale linearly. You get less light per milliamp than you'd expect, and the color shifts because the phosphor coating on white LEDs doesn't get excited uniformly. So you're fighting physics at both ends.
Wait, I want to pause on that. The phosphor coating — that's how white LEDs work, right? They're actually blue LEDs with a yellow phosphor on top?
A white LED is a blue LED at heart. The blue light excites a phosphor layer that re-emits a broad spectrum of yellow and red light, and the combination looks white to our eyes. But at very low currents, the blue LED is barely emitting, so the phosphor is barely getting excited. The result is that the color temperature shifts, often toward blue, which is exactly what you don't want for a bedtime display. So even if you solve the dimming problem, you might end up with light that's dimmer but proportionally more blue.
Which undercuts the whole point of dimming for circadian health. You're trying to reduce melanopic lux, but if the spectrum is shifting blue, you're partially undoing the benefit.
Melanopic lux — that's the metric that weights light by how much it stimulates the melanopsin-containing retinal ganglion cells that drive circadian rhythm. And you've just named the exact tension at the heart of modern circadian lighting research. For years, the story was "blue light is the enemy." Block blue light, protect your melatonin. That was the entire premise of Night Shift and f.lux and every blue-light-filtering app.
The blue-light-blocking glasses industry, which I'm sure is thrilled to hear that intensity might matter more than spectrum.
The evidence has been shifting. There was a major review published in Nature Reviews Neuroscience in late twenty twenty-three that really crystallized this. The researchers found that while blue light does suppress melatonin more than other wavelengths at equal intensity, the effect size is smaller than early studies suggested. And crucially, bright light of any color suppresses melatonin. A bright warm light will still mess with your circadian rhythm. A dim blue light might not.
The practical advice inverts. Instead of "filter out the blue," it's "just make it dimmer.
The corollary, which the prompt touched on: morning light exposure is non-negotiable. Getting outside within an hour of waking, even for ten or fifteen minutes, does more for your circadian health than any amount of evening blue-light avoidance. The morning light signal is the anchor. It sets the timer for melatonin release about fourteen to sixteen hours later. If you get that morning anchor right, your evening light sensitivity is somewhat buffered.
This connects to something I've noticed anecdotally. People who obsess about blue light blocking often have terrible sleep anyway because they're not going outside in the morning. They're filtering their screens but they're not getting the primary zeitgeber.
Zeitgeber — time-giver. The German term for the environmental cue that entrains the circadian clock. Light is the primary zeitgeber. And morning light is the most potent zeitgeber of all. There's a reason the clinical guidelines for circadian rhythm disorders always start with timed morning light exposure, not evening light restriction.
Daniel's instinct is right. Turn the lights down, don't stress about the spectrum. But he's still stuck with hardware that won't turn down far enough. You mentioned three bottlenecks — PWM resolution, LED efficiency droop, what's the third?
The third is the one that most people never think about: leakage current. Every transistor in the driver circuit has a tiny amount of current that flows even when it's supposed to be off. At normal brightness levels, this is negligible. But when you're trying to dim to zero point zero one percent, that leakage current can be enough to keep the LED glowing visibly. It's like a faucet that won't quite shut off. The drips might be small, but in a dark room, you see them.
You're fighting the physics of semiconductors themselves. The thing is fundamentally leaky.
Fixing it requires more sophisticated driver topologies — things like active discharge circuits, or hybrid dimming that combines PWM with analog current control, or what's called "PWM with controlled current amplitude" where you vary both the pulse width and the pulse height simultaneously. These techniques exist, but they add cost and complexity. A basic PWM driver might cost twenty cents in volume. A hybrid dimming controller with sixteen-bit resolution and active leakage suppression might cost two or three dollars.
Which in consumer electronics is a chasm. Nobody's spending an extra two dollars and fifty cents per driver so that the point-one-percent-of-users who want sub-nit brightness are happy.
This is the economic reality of the thing. The market for "I want my screen to be almost completely dark" is small. Most users complain about screens not being bright enough in sunlight, not about them being too bright in darkness. The entire engineering effort of the display industry for the last decade has been about peak brightness. HDR, Dolby Vision, a thousand nits, two thousand nits. The bottom end of the range has been neglected because nobody markets a phone based on how dim it gets.
Although that's exactly what Daniel is asking for. A phone that wins on minimum brightness. Are there any?
There are a few manufacturers that have paid attention to this. OnePlus has been fairly good about low minimum brightness on some of their phones — they've used higher-resolution PWM controllers and their OxygenOS has historically allowed lower minimums than stock Android. Google's Pixel phones, interestingly, have gotten better at this in recent generations. The Pixel nine series, which came out in late twenty twenty-four, can go down to about one nit according to measurements from Notebookcheck. That's still not sub-nit, but it's better than the two-to-four nit floor on most phones.
One nit is still a lighthouse in a dark room, but it's a smaller lighthouse.
The real standout in the phone space has been some of the flagship Samsung Galaxy devices. Their AMOLED panels can achieve lower minimum brightness than most LCDs because OLED pixels are self-emissive — each pixel is its own tiny LED, and you can control them individually without a separate backlight. The Galaxy S twenty-five Ultra, measured by DXOMARK earlier this year, hits about zero point eight nits at minimum brightness in their testing. That's low.
Still not sub-nit though. Zero point eight is under one, but we're talking about thresholds where zero point one would be ideal.
This is where the prompt's mention of Philips Hue is really astute. Because Philips cracked this problem in the smart bulb space years ago. Their Hue White and Color Ambiance bulbs, especially the newer generations, can dim down to what Philips claims is about zero point one percent of full brightness. In practical terms, that's a glow so faint you can barely tell the bulb is on.
I've seen this. A Hue bulb at minimum brightness in a dark room is like a dying ember. It's actually useful as a nightlight.
The way they achieve it is through what's called "hybrid dimming" or "deep dimming" depending on which white paper you're reading. They use a combination of PWM and what's essentially a current-limiting circuit that reduces the amplitude of each pulse, not just the width. At higher brightness levels, they use standard PWM. But when you go below about five percent, the driver switches to a mode where it reduces the current itself, and then applies PWM on top of that reduced current. So you get the fine granularity of PWM multiplied by the amplitude reduction. It's like having a coarse dimmer and a fine dimmer in series.
That's elegant. Why doesn't everyone do that?
Cost, complexity, and patent issues. Philips has a lot of intellectual property around their deep dimming implementation. And the driver circuitry is more expensive — you need better components, tighter tolerances, more calibration in firmware. A basic smart bulb might have a five-dollar bill of materials. A Hue bulb with deep dimming might have a bill of materials closer to eight or nine dollars. In the smart home space, where people are buying dozens of bulbs, that difference adds up fast.
For the bedroom, where you want exactly this capability, it's worth it. Daniel specifically called out that Philips has differentiated on this. And I think he's right — it's one of those features that doesn't show up on a spec sheet comparison but makes a huge difference in daily use.
There are other players in the smart bulb space that do decent low-end dimming. LIFX has some bulbs that go quite low. IKEA's Tradfri line has improved with their newer hub firmware. But Philips is still the benchmark for "can this bulb be almost off but not quite?" in a way that feels smooth and intentional rather than flickery and unstable.
What about monitors? Daniel mentioned finding the same problem on computer monitors. Are there displays that natively support very low brightness?
This is actually harder than phones or bulbs, because monitors have different constraints. A monitor needs to maintain color accuracy and uniformity across the entire brightness range. When you dim a monitor's backlight very low, you often get what's called "backlight bleed" becoming visible — the edges of the panel look brighter than the center because the LED strips around the bezel can't distribute light perfectly uniformly at very low output. It's the same leakage problem but at the optical level.
Even if the electronics can do it, the physical light guide in the panel might not cooperate.
That said, there are some monitors that handle low brightness well. Dell's UltraSharp line, particularly the models aimed at graphic designers, often have better low-end brightness control because they use higher-quality backlight drivers and better panel calibration. The Dell U series monitors can typically go down to about thirty to forty nits at minimum, which is lower than most consumer monitors that bottom out at fifty to sixty nits.
Thirty nits is still not what Daniel's looking for. He wants sub-nit.
For sub-nit on a monitor, you're mostly looking at OLED monitors, which have started appearing in the last couple of years. LG's OLED monitors, the ones based on their TV panels, can achieve low black levels because OLED pixels can be turned almost completely off. The LG twenty-seven-inch OLED monitor, the model they released in late twenty twenty-four, can achieve effectively zero nits on pure black content. But for an all-white screen at minimum brightness, you're still looking at maybe five to ten nits.
Because you've got all those pixels emitting, even at minimum drive current.
And the monitor has to maintain white balance across the whole screen, so it can't just turn everything down arbitrarily. There's a floor imposed by the need for uniformity. This is an area where micro-LED technology might eventually help — each pixel is an independent LED with its own driver, so in theory you could achieve much finer per-pixel dimming. But micro-LED monitors are still mostly prototypes and demo units. Samsung showed a micro-LED monitor at CES in January, but it's not shipping to consumers yet, and the expected price is somewhere in the five-figure range.
The practical advice for someone who wants a dim monitor right now is either an OLED panel or software overlays like Darker Pro. Neither is perfect.
There's a third option that's a bit more involved but works well: bias lighting. Putting a neutral light behind the monitor, around sixty-five hundred Kelvin, at a low intensity. The idea is that by raising the ambient light level slightly behind the display, your eyes don't need to adapt as fully to darkness, so the monitor's minimum brightness feels lower relative to the surroundings. It sounds counterintuitive — adding light to make the screen feel dimmer — but it works because of how human brightness perception is relative rather than absolute.
That's a clever perceptual hack. You're not actually reducing the screen brightness, you're changing the reference point your visual system uses.
It has the side benefit of reducing eye strain, because your pupils aren't constantly adjusting between a bright screen and a dark wall. The bias light keeps your iris at a more consistent aperture.
I want to circle back to something you mentioned earlier about the Nature Reviews Neuroscience paper. The shift from "blue light is the enemy" to "intensity is what matters" — how settled is this? Because I can imagine the blue-light-blocking glasses industry pushing back hard.
It's not completely settled, and I should be careful not to overstate the consensus. What's happened is that the effect size for blue light specifically has been revised downward. Early studies, some of them from the early two thousands, found that blue light at four hundred eighty nanometers suppressed melatonin by roughly twice as much as green light at five hundred fifty-five nanometers at the same intensity. Those numbers have held up reasonably well. What's changed is our understanding of the dose-response curve.
Meaning how much light you need before any of this matters.
The early studies used fairly bright light sources — hundreds of lux. When you test at lower light levels, the difference between blue and other wavelengths shrinks. Below about thirty lux, which is roughly the light level of a dimly lit living room in the evening, the spectral composition matters much less than the total intensity. And here's the kicker: most people's evening light exposure is in this range. Your phone at minimum brightness is maybe ten to twenty lux at viewing distance. At that level, whether it's blue-filtered or not makes a negligible difference to your melatonin. What matters is that it's light at all.
The Night Shift mode on iPhones — which shifts the display warmer — is doing almost nothing at low brightness. The real win would be making the phone dimmer, not warmer.
This is exactly the conclusion that the circadian research community has been inching toward. The practical intervention that actually works is dimming, not filtering. And the most potent intervention of all is morning light. There was a really striking study from the University of Colorado Boulder in twenty twenty-four that took a group of participants camping for a week with no artificial light at all. Their melatonin onset shifted earlier by nearly two hours, and their sleep quality improved dramatically. But the researchers noted that the biggest factor wasn't the absence of evening blue light — it was the massive dose of morning sunlight they were getting.
Camping as circadian therapy. Daniel's prompt didn't mention camping, but the logic leads there. If the morning light anchor is so powerful, the ideal intervention might be "go camping for a weekend" rather than "buy these glasses.
Or just go outside in the morning. Even on a cloudy day, outdoor light is thousands of lux. Indoor light, even in a brightly lit office, is maybe three to five hundred lux. The difference is an order of magnitude. Your circadian system evolved for outdoor light levels, and indoor lighting is, from an evolutionary perspective, perpetual twilight.
Which brings us back to the hardware question with a different framing. Maybe the problem isn't just that minimum brightness is too bright. Maybe it's that our entire indoor light environment is badly calibrated for human biology. We're living in a world where the dim setting is still too bright, and the bright setting is still too dim relative to outdoors.
That's the deeper insight, and I think it's where the prompt was pointing even if it wasn't stated explicitly. Daniel's frustration with minimum brightness is a specific instance of a broader mismatch between our technology and our biology. The LED revolution gave us efficiency and longevity, which is great — an LED bulb uses about eighty-five percent less energy than an incandescent and lasts twenty-five times longer. But we lost something in the transition. We lost the smooth, analog dimmability of a hot wire. We lost the natural warm-shift that happens when you dim an incandescent. We replaced a technology that was accidentally good for circadian health at low levels with one that's accidentally bad.
Accidentally good is a great phrase. The incandescent bulb dimmed by reducing current through a filament, which naturally shifted the spectrum toward red, which is exactly what you want at night. Nobody designed it for circadian health. It just happened to align.
LEDs, which are blue at heart with a phosphor coating, are the opposite. At low currents they can shift blue, as we discussed. The technology that's more efficient in energy terms is less aligned with our biology. It's a genuine tension.
Where does this leave someone who wants to actually solve the problem today? Daniel asked specifically about manufacturers to watch. You mentioned OnePlus and Google for phones, Samsung's AMOLED panels, Dell UltraSharp and LG OLED for monitors, Philips Hue for smart bulbs. Any others worth naming?
In the smart bulb space, I should mention that there's a company called Ketra, which was acquired by Lutron a few years ago, that makes extremely high-end lighting systems with what they call "natural dimming." Their bulbs can go incredibly low and they maintain color accuracy across the entire dimming range. But they're not consumer products — we're talking hundreds of dollars per bulb plus a control system. It's the kind of thing that goes into luxury hotels and high-end residential installations.
Not exactly a practical recommendation for someone who just wants their bedside lamp to be less aggressive.
For the practical recommendation, Philips Hue is still the sweet spot. The White Ambiance bulbs, which do tunable white but not full color, are about twenty to twenty-five dollars each and they support deep dimming. The full color bulbs are a bit more expensive but use the same deep dimming driver. And the Hue ecosystem has good software control — you can set precise brightness levels in the app, create schedules, and integrate with Home Assistant or other smart home platforms for more advanced automations.
What about TVs? Daniel mentioned those too.
TVs are interesting because OLED TVs have been the best at low black levels for years, but that's for dark scenes in content, not for overall brightness. If you're using a TV as a light source — say, playing a YouTube ambient video for sleep — the minimum brightness on most TVs is still surprisingly high. LG's OLED TVs can go quite low, and Sony's OLEDs are similarly good. But the real trick with TVs is to use the built-in picture settings. Most TVs have a "movie" or "cinema" mode that's calibrated for dark-room viewing and runs at lower brightness than the default "standard" or "vivid" modes. Some also have an "expert" or "ISF" mode that gives you finer brightness control.
You're saying people might just have their TV in the wrong mode.
A lot of the time, yes. The default settings on most TVs are designed to look good on a showroom floor under bright fluorescent lights. They're cranked up. Switching to cinema mode and then manually pulling the backlight slider down can often get you much closer to where you want to be for nighttime viewing.
I want to go back to something you said about the Colorado camping study. Two-hour shift in melatonin onset after a week without artificial light. That's a massive effect size. It suggests that a lot of what we're doing with technology interventions — the apps, the glasses, the smart bulbs — is compensating for a more fundamental problem: we don't get enough outdoor light during the day.
That's the uncomfortable truth at the bottom of all this. The most effective circadian intervention is free and requires no technology: go outside in the morning. Everything else is mitigation. The blue light glasses, the screen filters, the deep-dimming smart bulbs — they're all ways of reducing harm rather than providing benefit. They're harm reduction for an indoor lifestyle.
Which doesn't mean they're not useful. Harm reduction is a legitimate strategy. If you can't or won't change your morning routine, then yes, dimming your lights aggressively in the evening is a good idea. But it's worth being honest about what you're doing.
And I think that's actually what I appreciate most about Daniel's prompt. He's not asking "what's the perfect circadian optimization strategy?" He's asking a practical question: "given that I have a bedtime routine that involves a screen, how do I make that screen less disruptive?" And the answer involves understanding PWM controllers and LED phosphor physics, which is exactly the kind of niche technical depth that makes this fun.
That's the show in a nutshell. Practical question, and the answer requires a tour through semiconductor physics, display engineering, and evolutionary biology.
Should we talk about where this is heading? Because there are some interesting developments on the horizon that could change the landscape.
Go for it.
Micro-LED is the big one. I mentioned it briefly, but it's worth elaborating. Micro-LED displays are made of microscopic LEDs — each pixel is its own independent light source, like OLED, but using inorganic gallium nitride instead of organic compounds. The advantage for our purposes is that inorganic LEDs can be driven at much lower currents than OLEDs while maintaining stability. In theory, a micro-LED display could achieve true sub-nit brightness with perfect color accuracy and no PWM artifacts. Apple has been investing heavily in micro-LED for years, and while they've had setbacks — they reportedly canceled a micro-LED Apple Watch project in twenty twenty-four — the technology is still advancing.
We're still in the "any day now" phase with micro-LED.
For consumer products that are affordable, yes. But the trajectory is clear. The other development is in driver electronics. There's a new generation of LED driver chips using something called "hybrid log-gamma dimming" that combines logarithmic current control at the low end with PWM at the high end. These are starting to appear in higher-end monitors and some automotive displays. The company Analog Devices has a line of these drivers, and Texas Instruments has competing parts. They're not cheap yet, but they're filtering down.
On the software side? Are operating systems getting better at exposing fine brightness control?
Android has had a hidden "extra dim" feature since Android twelve that applies an additional dimming layer beyond the hardware minimum — it's essentially a system-level version of what Darker Pro does. Apple added something similar in iOS eighteen called "Dim Flashing Lights" but it's buried in accessibility settings and most people don't know it exists. The trend is toward more granular control, but it's still not a first-class feature that gets marketed.
Which is the recurring theme here. The capability exists, it's just not prioritized because the market incentives point toward brightness, not dimness.
That's unlikely to change until consumers start asking for it. The prompt we're discussing is, in a small way, part of that demand signal. Someone saying "I want my screen to be darker, and I'm willing to choose products based on that" is how the market eventually shifts.
To synthesize the practical takeaways: for phones, look at Samsung's AMOLED flagships or Pixel phones with extra dim enabled. For monitors, OLED panels from LG or Dell's UltraSharp line, plus bias lighting as a perceptual hack. For smart bulbs, Philips Hue is still the benchmark for deep dimming. For TVs, switch to cinema mode and pull the backlight down. And for everything, the software overlay option — Darker Pro on Android, similar tools — can bridge the gap where hardware falls short.
The meta-takeaway: the research increasingly says intensity trumps spectrum. Get morning light, dim everything aggressively in the evening, and don't stress about whether your screen looks slightly warm. The dimmer is the better intervention.
Dimming over filtering. Morning sun over evening gadgets. It's almost disappointingly simple.
The best advice usually is.
Now: Hilbert's daily fun fact.
Hilbert: In the nineteen hundreds, a Russian ethnographer named Vladimir Jochelson proposed that the indigenous Itelmen people of Kamchatka had once used a whale-rib xylophone for shamanic rituals. The theory was based on a single carved bone fragment he found in nineteen ten. He published extensively on it before admitting, decades later, that the bone was probably just a tool for scraping hides.
The whale-rib xylophone was a back-scratcher.
A hide-scratcher. Which is somehow even less mystical.
Before we wrap, I want to leave listeners with one thought experiment. If the research on intensity versus spectrum continues to firm up, we might look back on the blue-light panic of the twenty-tens the way we now look at the low-fat diet craze of the nineties. A plausible-sounding mechanistic story — blue light hits melanopsin, melanopsin suppresses melatonin — that turned out to be less important in practice than the simpler, less marketable truth: too much light at night is bad, and not enough light during the day is worse.
The low-fat diet of sleep science. I like that. It had a good lobby and a weak evidence base.
The real solution was mostly "eat less, move more." Or in this case, "go outside in the morning, turn the lights down at night.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. You can find us at myweirdprompts dot com. If you enjoyed this episode, leave us a review wherever you listen — it helps other people find the show.
Until next time.
