Quantum intuition for everyone, or: why Einstein thinks reality isn't reasonable
Science tells us some weird and unintuitive things, like that the average puffy cumulus cloud weighs as much as five blue whales.
Most facts like this are inconsequential for our daily lives. Interesting but not existential.
So letâs talk about something weirder.
Itâs going to sound like a spiritual joke. But this is an extremely-consensus, rigorously-proven scientific fact.
The three impossible choices
Said more casually, modern science suggests that we live in either:
- A sort of dream-world, where properties solidify only when looked at
- Something like a holodeck, where far-flung pieces of the universe update in perfect sync without âcommunicatingâ by conventional means, or
- A puppet show, where our actions are wired into the hidden startup settings of the universe
I know this sounds a little bit woo. And, in fact, on this topic, Albert Einstein and two collaborators wrote a paper in 1935 that concluded:
No reasonable definition of reality could be expected to permit this.
â Albert Einstein, Boris Podolsky, and Nathan Rosen, 1935
But it turns out, 90 years and a dozen or so Nobel Prizes in Physics later, we are quite confident it is true, and Einstein, Podolsky, and Rosen â henceforth âEPRâ â would be pretty upset.
And quantum mechanics is to blame.
Einsteinâs wager
The Einstein, Podolsky, and Rosen paper (âEPR paperâ) in 1935 lays out a very interesting thought experiment (weâll get to it in a minute), and argues on its basis that quantum mechanics couldnât be quite right â or, at minimum, is missing something.
EPRâs ultimate conclusion was that in order to believe that quantum mechanics was correct and complete, youâd have to accept something â and that something simply could not be permitted in any âreasonable definition of reality.â
Unfortunately, it turns out that todayâs scientists are very certain that quantum mechanics is, in fact, correct and complete. Weâre about as certain of this as we are of any scientific theory ever.
Science has since proven that reality is âunreasonableâ in exactly the way I outlined above: at least one of those three realities (the dream-world, the holodeck, or the puppet show) is true. This is what bothered the EPR trio.
So letâs try to understand how this came to be, and what it means for us.
A note: this article is meant for people with no physics or quantum mechanics background. My goal is to give you some âholy shitâ realizations and intuition for the âmysteriesâThey arenât really mysteries â they just donât fit our classical intuitions. of quantum mechanics that typically require a reasonable amount of hard science and math â but in a friendlier format.
That may mean I offend some physicists with imprecise language. Please accept my apologies, and do email me if I get anything wrong or you have ideas for improvement!
What is quantum mechanics, anyway?
Before getting into this, we should understand: what is this whole âquantum mechanicsâ thing? Youâve certainly heard the words before, but theyâre often surrounded by either impenetrable math () or hand-wavy pop spirituality (âquantum healing!â).
When we do science, we create models and mechanisms. Models are descriptions of what we expect to observe (e.g. Keplerâs laws of planetary motion describing the trajectory of planets), and mechanisms are the prescriptive force that causes those outcomes to happen (e.g. planets move thanks to gravitational pulls from other objects).
Models are the map; mechanisms are the territory.
Quantum mechanics â today, at least â is chief among the first category: a model. We know the model cold â the equations work â but we donât know why. Itâs a seemingly-perfect map for a territory we havenât yet glimpsed.
Now, what is quantum mechanics actually a model for?
Think of reality like a high-resolution photo on a screen. When zoomed out and unified, everything looks smooth and intuitive, and âclassicalâ physics (the physics you learned in high school â force equals mass times acceleration, etc.) works well. But if you zoom really far in and begin to isolate quanta separated from a broader environment, discrete pixels appear. And whenever your experiment or interaction with reality bumps into those pixels, classical equations start glitching and quantum mechanics takes over.
In practice, that means quantum mechanics covers things like how atoms work, how light and color works, the architecture of chemistry, nuclear effects, semiconductors, and more.
In that zoomed-out, high-resolution photo, you could use the rules of âpixel worldâ to create the overall image â and in much the same way, you can generate classical physics from quantum physics. But most of the time, when you arenât at quantum scale, itâs just easier to look at the zoomed-out / classical view for simplicity.
In the early 1900s, scientists started to run into these sorts of âzoomed-inâ issues for the first time, where the classical physics that had worked so well for everything else started to fail.
As a starting point, most people point to Max Planckâs observations about how heated objects emit light. Classical physicsâ predictions for the behavior of this light was simply wrong, and it took Planckâs creation of the first quantum theory â for which he won the 1918 Nobel Prize in Physics â to resolve it.
Thatâs what brought about this âpixel-levelâ view of physics. And from 1900 through 1930, we laid the foundations of quantum mechanics, by building out a model â a series of key equations that encapsulated everything we understood about the topic. We didnât know why these were true â we didnât know the mechanism â but we had good math.
These mathematical equations then formed the foundation of the next century of experimentation. We would use the equations to predict what would happen in a certain experiment, then run the experiment, and check if the equations were right.
The more often the equations were right, in weirder and weirder situations and experiments, the more confident we became that the overall model of quantum mechanics was correct.
And 100 years later, thereâs never been a quantum mechanical experiment where the model and the results disagreed.
This is great! Itâs how science is supposed to work. We develop theories, we test them rigorously, and through such experimentation, the theory is confirmed. Our understanding of the world has grown, and we can then use this to improve humanity.
And indeed, this would be all well and good â and I probably wouldnât be writing this piece â if not for two issues:
- The results predicted by quantum mechanics just seem really, really unreasonable to us (and to Einstein!)
- We still donât know the mechanism
Weâve got a map, and every single time weâve used it, itâs brought us to the right place. But the map is super weird and very unlike any other map weâve ever seen, and we still donât understand the territory.
Interpretations abound
Ever since the model of quantum mechanics started coming together, people have put forth theories of the mechanism â that unknown territory underneath the map. These are generally referred to as âinterpretationsâ of quantum mechanics, and Wikipedia lists more than a dozen of them.
But this brings us back to the three âunreasonableâ possibilities about reality I originally mentioned: the dream-world, the holodeck, and the puppet show.
At their core, each interpretation of quantum mechanics effectively accepts one of those possibilities and then builds a view of reality on top of that.In a later article, Iâll share some thoughts on the landscape of quantum interpretations and where I lean personally.
It all began with that 1935 EPR paper I mentioned above. The heart of that paper is a thought experiment, where Einstein and co. outlined an idea at the âpixelâ scale of reality governed by quantum mechanics. They described what outcome quantum mechanics would predict, and then basically said âbut that canât possibly be true.â
In the original paper, they write it pretty densely and abstractly. The person who made a clear enough version for us to walk through here was David Bohm in the early 1950s. He took the âEPR paradoxâ thought experiment and created a version that could actually be experimentally tested.
Letâs try to build some intuition for this experiment, and perhaps weâll get the same uncomfortable feeling Einstein and his colleagues did.
A thought experiment
I want you to imagine an experimental setupWe owe David Mermin gratitude for explaining this experiment in such a clear-eyed way in his articles âIs the moon there when nobody looks?â and âQuantum mysteries for anybodyâ. Weâre going to stretch the limits of practicality a bit to make the point, and start with classical intuitions before shifting to quantum.
Thereâs a machine that shoots out two âparticles,â one in each direction. Far away from the machine on each side, there is a detector. Each detector receives one of the particles.

The detectors are each set to one of three measurement settings: shape, size, or color. That is: on a given test, detector 1 might be measuring shape, while detector 2 measures size. Or maybe theyâre both measuring size. But they each only measure one attribute.
When a particle hits one of the detectors, each of those measurement settings can only cause that detector to output âyesâ or âno.â So really, the settings are âis it round?â (shape), âis it big?â (size), and âis it purple?â (color).
Okay, so, to recap:
- machine shoots two particles, one at each detector
- each detector is set to one of three measurement settings
- when the particle hits the detector, the detector outputs âyesâ or ânoâ based on which measurement setting it is on, and whether the particle meets those criteria
An example:
- Detector 1 is set to âsizeâ
- Detector 2 is set to âshapeâ
- Two big, square, purple particles are shot out
- So detector 1 outputs âyesâ (the particle that it got was big)
- And detector 2 outputs ânoâ (the particle that it got was square)
Easy, right?
Okay. So now letâs add a constraint. We observe that whenever the detectors are set to the same measurement setting, the same yes/no output happens.
This is key for the experiment.
If theyâre both set to âsize,â either they both say âyesâ (big particles) or they both say ânoâ (small particles).
If theyâre both set to âcolor,â either they both say âyesâ (purple particles) or they both say ânoâ (non-purple particles).
Now, letâs say the measurement settings on each detector are set randomly, and the machine shooting the particles out doesnât know what measurement settings are set (in fact, perhaps the measurement settings arenât even set until the particles are on the way).
What does that mean about the particles?
Itâs a pretty simple conclusion: the particles must be the same across all three attributes. If one is small, square, and purple, so too must the other be â if weâre saying that the detectors must show the same output if set to the same measurement setting.
Make sense? If not, give that another quick read. Weâre saying that we see when the measurement settings are the same, the outputs are the the same. This means that measured attribute of the particles must be the same. But if we donât know which measurement setting is going to be picked, then that means all the attributes must be the same, to account for any possible selection.
Okay. This seems fine. But now we get to the weird bit.
The classical prediction
Letâs now look at the case where the measurement settings are different â i.e. detector 1 is looking at size, detector 2 is looking at shape.
There are six possible cases where settings are different:
Detector 1 | Detector 2 |
---|---|
Size | Shape |
Size | Color |
Shape | Size |
Shape | Color |
Color | Size |
Color | Shape |
If we keep in mind our assumption that the two particles must have the all same attributes according to the logic above, weâll end up with this conclusion:
The detectors will show the same result (both âyesâ or both ânoâ) at least one-third of the time, and they will show a different result no more than two-thirds of the time. Hereâs an example:
With particles that are: big, square, purple
Detector 1 | Detector 2 | Same result? |
---|---|---|
Size = yes | Shape = no | No |
Size = yes | Color = yes | Yes |
Shape = no | Size = yes | No |
Shape = no | Color = yes | No |
Color = yes | Size = yes | Yes |
Color = yes | Shape = no | No |
1/3 same result; 2/3 different result.
Hereâs another:
With particles that are: small, square, green
Detector 1 | Detector 2 | Same result? |
---|---|---|
Size = no | Shape = no | Yes |
Size = no | Color = no | Yes |
Shape = no | Size = no | Yes |
Shape = no | Color = no | Yes |
Color = no | Size = no | Yes |
Color = no | Shape = no | Yes |
100% same result; 0% different result.
In any configuration of particle attributes, with the six different detector settings, youâll never end up with fewer than 1/3 of the outcomes being the same. Try it for yourself: see if you can get âPercentage Sameâ to less than 33% in the demo below (hint: you canât).
The quantum prediction
So this is all well and good, right?
Well â this is the awkward bit.
Everything we just said is true for âclassicalâ experiments â like if you were measuring classical-scale properties of particles like size, shape, or color. But if we zoom in to the quantum level and instead look at quantum properties like polarization angles or spin measurements⌠despite all the logic feeling like it should be the same⌠itâs just not what the quantum math says.
The quantum math ends up saying that 25% of the time, youâll get the same result (both âyesâ or both ânoâ), and 75% of the time, youâll get a different result. 25% same â but our intuition told us that you can never get less than 33%!
In fact, the quantum math says that each of those six configurations gives the same result 25% of the time, and different results 75% of the time.
To give a concrete example:
- if both the settings are on âvertical polarization,â they will always say âyesâ together, or ânoâ together â meaning both particles are measured as vertically polarized, or both as not vertically polarized
- if both settings are on âhorizontal polarization,â they will always say âyesâ together, or ânoâ together â meaning both particles are measured as horizontally polarized, or both as not horizontally polarized
- but if one setting is on âvertical polarizationâ and one is on âhorizontal polarization,â they will say the same thing 25% of the time and different things 75% of the time â meaning that the particles have different attributes part of the time
How is this possible?
This is pretty much exactly what Einstein ran into. He thought this through and said âwell this just doesnât make any sense.â
Bellâs coup
In 1964, John Bell came up with Bellâs Theorem and its associated âBell inequalitiesâ â these are effectively the formalization of the intuitively-broken math framed above. Over the next couple decades, other scientists added to the structure of the problem. And where they collectively landed was effectively the following:
The only way those quantum-theorized results could be true is if one of three assumptions was incorrect:
- âRealismâ â the idea that the two particles already carry definite, pre-existing properties for every possible measurement, independent of whether we observe them. If they donât, then at the time of measurement, they could swap to hit the quantum odds rather than obeying their âdefinite, pre-existing property.â
- âLocalityâ â the idea that the two particles cannot communicate with each other instantly from a distance. If they could, then they could âcoordinateâ on a âtargetâ quantum outcome from afar (faster than the speed of light!), instead of needing to stick with whatever settings were either coordinated at the origin or randomly determined along the way.
- âFree choiceâ â the idea that the detector settings are truly random, not pre-determined or correlated by the universe with the particlesâ properties. If it was pre-determined, then the ârandomâ measurement settings of the detectors could be âset upâ to land in the target percentages for those particles â nature playing a trick on us, making us think weâre controlling the experiment when we really arenât.
The violation of any of those assumptions â now formalized â was what EPR didnât like, way back in 1935. If the quantum theory was correct, that would mean one of those assumptions is wrong, and âno reasonable definition of reality could be expected to permitâ that.
This whole time, weâve been carrying those implicit assumptions â that particles actually have properties independent of being measured; that they canât instantly coordinate across space and time; and that itâs possible to have truly random, not pre-ordained settings. Those are the same assumptions Einstein had about the world.
And, in fact, a violation of these assumptions is the only way you could get the theorized results from the system (i.e. those 25% odds). Try to come up with another way!
But up until now, this was all just a theory. Quantum mechanics seemed to be correct, but this experiment in question had never been run in the real world.
What the labs saw
That all changed starting in 1972, when physicists began running versions of this experiment. It began with Freedman and Clauserâs 1972 entangled calcium-cascade photon experiment, then Aspectâs experiments starting in 1981, then Zeilingerâs from 1998 through the 2010s.
These experiments are very tricky to run, and what the physicists were doing for these five decades was progressively closing âloopholes.â That is, there are ways that any of those three assumptions could âsneak inâ based on issues with experimental design. But from the beginning, it was evident that the quantum conclusions were likely to prevail.
In 2015, âloophole-freeâ tests were run in Vienna and the US. The public writeups from them (Austria, US) are brief and readable.
What did the labs actually see? Exactly the 25 / 75 splits, not the 33% we would intuitively expect.
And in 2022, the Nobel Prize in Physics was awarded to Clauser, Aspect, and Zeilinger for âexperiments with entangled photons, establishing the violation of bell inequalities and pioneering quantum information science.â
In plain English?
The quantum predictions were right the whole time, at least one of the assumptions is wrong (realism, locality, or free choice), and 1935 Albert Einstein would say that must mean we do not live in a reasonable reality.
The science says: we either live in a dream-world, a holodeck, or a puppet show (or perhaps more than one of those).
The Nobel laureates weigh in
In his 2022 Nobel Prize lecture, John Clauser talks about this idea, saying that âlocal realismâ is clearly wrong â that is, in my words, we either live in a dream world or a holodeck. To drive home how weird this is, he puts on a slide (emphasis mine):
What does Local Realism assume?
Local Realism is extremely simple and reasonable. Anton Zeilinger once noted that if⌠Local Realism [had been] formulated prior to the discovery of Quantum Mechanics, [it] would have been adopted as new âlaws of nature.â
Local Realism assumes that nature consists of stuff, of objectively real objects, i.e., stuff that you can put inside a box. Local Realism further assumes that objects exist whether or not we observe them.
We may not know what the stuff is, but we assume that it exists and that it is distributed through space. EPR called stuff âelements of realityâ.
âŚ
Local Realism assumes that the stuff within a box has intrinsic properties, and that when someone performs an experiment within the box, the probability of any result⌠is somehow influenced by the properties of the stuff within that box. If one performs say a different experiment⌠then presumably a different result obtains.
Nobody had ever really thought to question this concept â itâs been a foundational assumption for, as best I can tell, all of time. It was only formalized as a possible theory in 1974 in light of quantum mechanics putting it into doubt.
But, as Clauser continues:
Local Realism was a very short lived viable theory. It was already refuted by existing experiments before it was fully formulated
Even though Local Realism was dead-on arrival, that fact wasnât good enough for many people. It thus fostered very many additional subsequent experimental tests to totally destroy itâŚ
This is really something. Clauser â in his Nobel Prize lecture on this very topic â is telling us in no uncertain language that our belief that ânature consists of stuff, of objectively real objectsâ is wrong.
More formally, heâs telling us that either local realism is wrong, or free choice is (the puppet show hypothesis), but:
[free choice being false would] essentially dismiss all results of scientific experimentation. Unless we proceed under the assumption that hidden conspiracies of this sort do not occur, we have abandoned in advance the whole enterprise of discovering the laws of nature by experimentation
â Shimony A, Horne M A and Clauser J F, âComment on the theory of local beablesâ, Epistemological Letters (1976)
Or as Zeilinger, his fellow laureate said:
[W]e always implicitly assume the freedom of the experimentalist⌠This fundamental assumption is essential to doing science. If this were not true, then, I suggest, it would make no sense at all to ask nature questions in an experiment, since then nature could determine what our questions are, and that could guide our questions such that we arrive at a false picture of nature.
â A. Zeilinger, Dance of the Photons
So what?
As Iâve been writing this piece, Iâve been wondering: why am I dwelling on the concept so much? Why do I feel the need to shake people by the shoulders and tell them about it?
I think the heart of it is that it is both deeply kooky-sounding and so roundly affirmed by science.
Whenever you run into a new piece of information outside your current worldview, you could ask yourself: how much does this piece of information change my perspective?
If the information is a huge departure from what you understand about the world, but is a non-validated, fringe idea accepted by few people â it probably doesnât change your perspective too much. Maybe you file it away, waiting for credible (to you) sources to validate it. But without much validation, even though the idea is impactful, you donât meaningfully update your worldview.
On the other hand, if the information is really strongly validated, but doesnât differ too much from your existing worldview, it also probably doesnât change your perspective significantly.
Maybe you find out for the first time that, actually, water doesnât always boil at 100 degrees Celsius â it depends on the altitude and corresponding air pressure. But in practice, on Earth, itâs pretty much within 10% of 100 degrees Celsius. This is a well-validated minor departure from what you already know â so, sure, you integrate it, but it doesnât change your life.
But whatâs much rarer is an idea that is a huge departure from your conventional intuition and is strongly validated. When you run into one of these, it requires a major update.
Thatâs what these quantum implications on reality are, at least for me. We just quoted a Nobel Prize lecture where the laureate said that the theory âthat nature consists of stuff, of objectively real objectsâ was âdead on arrivalâ and has been âtotally destroy[ed].â
Major departure from conventional scientific worldview? Yep. Strongly affirmed by effectively all scientific evidence? Also yes.
Itâs easy to look at historical step-changes in our understanding of the world and reality and think about how odd it was that anyone believed anything else, like heliocentrism, when we realized the Earth was not the center of the universe. Or evolution, when we understood that species evolve over time and from one another. Or relativity, when we found that time and space arenât fixed but rather are subjective.
And weâre in the middle of one of those right now. A real, live step-change in our understanding of reality, one that is deeply experimentally-validated. One where itâs almost certain that in 100 years, people will look back at us and laugh, saying, âhow did they go about their lives not understanding how our universe works?â
Isidor Isaac Rabi, another Nobel laureate, said in 1985 about quantum mechanics:
I think we are missing a very basic point. The next generation, when they find it, will knock on their heads and say, âHow could they have missed that?â
Richard Feynman, yet another laureate, said:
We have always had a great deal of difficulty in understanding the world view that quantum mechanics represents⌠I still get nervous with it⌠you know how it always is, every new idea, it takes a generation or two until it becomes obvious that thereâs no real problem.
I cannot define the real problem [with quantum mechanics], therefore I suspect thereâs no real problem, but Iâm not sure thereâs no real problem.
So what?
At the end of the day, how does all this actually affect you? How might you want to update your worldview?
Well, Iâd note three things:
First, even though we donât know the mechanism, these principles are already in use. The GPS on your phone uses quantum mechanics (and wouldnât work without it!), MRI machines leverage quantum physics, and more. Applications will continue to develop, from quantum communication (ultra-secure encryption) to quantum computing, and more.
Second, there are people out there working every day on trying to discern the mechanism â the actual reasons reality behaves in this way. One day, I imagine, weâll settle on one. And you should keep an eye out for that.
But third: even in the meantime, Iâd encourage you to take this all as evidence that itâs worth being open to âweirdâ interpretations of reality.
We â as in all of human history â often sit comfortable in our âconventionalâ understanding of the world, whatever that may be at the time. But, as science â the pursuit of understanding Nature â has shown over and over again, the conventional understanding is never quite right or complete.
And so, all Iâm asking of you is to be open to the idea that things arenât always as they seem. The world is weird and we often donât know why.
One of my favorite articles of all time is Reality is Very Weird and You Need to be Prepared for That by the pseudonymous bloggers Slime Mold Time Mold.
I wonât spoil the story â itâs worth a read â but the essay contains this idea:
Real explanations will sometimes sound weird, crazy, or too complicated because reality itself is often weird, crazy, or too complicated.
Or, as Albert Einstein once said:
The most incomprehensible thing about the world is that it is comprehensible
But the remarkable thing is that, now that weâve glimpsed this currently-incomprehensible behavior of the universe, we are working towards making it comprehensible. And the smart money says that future science will hand us a true mechanism as alien to our current sensibilities as heliocentrism was to early humans, or complex numbers to Pythagoras.
But what about chairs?
Letâs close here on a practical note. We just discussed how, apparently, the universe is not full of âstuff,â or âobjectively real objects.â
âButâ â you may say â âChairs! Tables! Fingers! Computers! They all certainly seem objectively real to me! I donât care what those theoreticians say, the world is real!â
(If youâve read my recent post on Ways of Looking, some of this may sound familiar â even if it doesnât seem like it should be related.)

Your objection is fair, and is a result of what quantum physicists call âdecoherence.â
In short: the scienceâs view is that the universe typically follows non-locally-real weird quantum behavior, which is called âcoherentâ (hah!).
But, circumstantially, at a certain scale and under certain conditions, particles come together and âdecohere,â losing their quantum behavior. The table in front of you is ârealâ inasmuch as the particles that make it up are, under a certain set of observational conditions, decohered and behaving classically. The table only seems stable and locally real â itâs not a fact of the table itself.
Quantum âweirdnessâ requires isolation, in a sense. A single photon can âbeâ in multiple places simultaneously, but only if nothing â no air molecule, no stray electromagnetic field, no measurement apparatus â âlooksâ at it, entangling it. The moment the quantum system interacts with its environment, the multiple possibilities rapidly collapse into definite classical outcomes.
This isnât because large objects are fundamentally different. Itâs because theyâre constantly âmonitored byâ or âentangled withâ their surroundings. Every atom in your table is continuously bumping into others, absorbing and emitting photons, exchanging electrons. This relentless environmental surveillance forces what we understand as the table to decohere and behave classically â like it has definite properties at definite locations.
But isolation breaks this. Physicists routinely make objects containing many atoms behave quantum mechanically by cooling them dramatically and shielding them from outside influence. Itâs not so much that the quantum world is microscopic â itâs that it is quiet and unobserved.
Interestingly, much of the work happening in quantum mechanics these days is around getting larger and larger chunks of matter to stay âcoherentâ or âquantum,â and to stay that way for longer and longer. A core task of quantum computing efforts, in some sense, is to get systems to stay coherent and not lapse into stability and ârealnessâ too quickly for the computer to finish its computation.
If we go back to our picture and pixels analogy from earlier: itâs like there are rules for pixels, but most of the time, people only see those pixels in the context of a larger picture. And when that happens, the behavior of the pixels is superseded by the attributes of the overall picture. Our eyes focus on the picture, not the pixels.
Quantum mechanics suggests that we view reality in much the same way. At the scale of our interaction with the universe, by default, we see pictures, not pixels. We interact classically, and whatever it is that is making up the universe behaves nicely for us to do so.
But in this last century, weâve glimpsed the circumstantiality of all of that. Weâve concluded â with some liberties â that under the hood, we live in a dream-world, a holodeck, or a puppet show. And now, we continue to attempt to figure out which it is.
Despite decades of thinking, experimentation, and discussion, even the most quantum among us leave this whole situation a little bit confused. John Clauserâs Nobel Lecture on his prize-winning experiments is titled:
Nonlocal quantum entanglement is real!
Confirming experiments, Local Realism, and Why I still find Quantum Mechanics difficult to understand
So, if youâre shaking your head about all this â just know you are in good company.
Looking for more to read?
Want to hear about new essays? Subscribe to my roughly-monthly newsletter recapping my recent writing and things I'm enjoying:
And I'd love to hear from you directly: andy@andybromberg.com