Every so often, a novel headline will fly across the world, claiming to revolutionize one or more of our most deeply held scientific ideas. The declarations are always sweeping and revolutionary, and include gems like:
- “The Big Bang never happened,”
- “This idea does away with dark matter and dark energy,”
- “Black holes aren’t real,” and the perennially popular,
- “Maybe this difficult-to-explain astronomical phenomenon is due to aliens.”
In 2026, with the rise in LLMs and sycophantic AI slop, physicist’s inboxes all over the world are filling up faster than ever with armchair versions of this as well. Despite occasionally glowing coverage of such novel proposals, they almost exclusively are doomed to languish in obscurity, attracting little mainstream attention other than a myriad of dismissals from professionals.
Commonly, it’s portrayed that scientists in such fields, like physics and astrophysics, are dogmatic, wedded to outdated ideas, and close-minded. This narrative is extremely popular among contrarian scientists or those who hold fringe beliefs for themselves, but it does not reflect the scientific truth. In reality, the evidence supporting the prevailing theories is overwhelming, and the new headline-grabbing proposals are typically no more compelling than the scientist’s equivalent of playing in the sandbox. There are four big flaws that commonly occur with new ideas, and you should be aware of them all. Once you are, you’ll understand why you’re likely to never hear about most such ideas again after they’re first put forth.
1.) When you work, every day, with “the real McCoy,” you can immediately spot an impostor’s shortcomings.
In science, we have accumulated an enormous body of knowledge — a set of experimental and observational data — that points to a set of theories as successfully providing a framework for accurately describing the governing rules of reality. Many of the results that we obtained were initially bizarre and counterintuitive, with multiple theoretical possibilities proposed to explain them. Over time, further experiments and observations winnowed them down, and the most successful theories with the greatest degrees of validity were the only ones that survived: because they remained consistent with the data.
Proposals that attempt to revolutionize one (or more) of our accepted theories have a large suite of hurdles to overcome. In particular, they must:
- reproduce all the successes that the prevailing theory has already produced,
- explain an already-existing phenomenon more successfully than the current theory can explain it,
- and make novel predictions, which can be tested, that differ from the predictions of the theory it’s attempting to supersede.
It’s very rare that all three of these criteria are met. In fact, the overwhelming majority of these grand proposals fail on even the first point: reproducing what’s already been successfully explained by the prevailing model.
Attempts to explain the Universe without a hot Big Bang, for example, fail to explain the existence and properties of the Cosmic Microwave Background: an omnidirectional pattern of radiation that’s been known for more than 60 years, with the most perfect blackbody spectrum of anything in the Universe. Contentions that gravitational wave detectors are seeing noise, rather than signals, ignore the large suite of evidence that links electromagnetically observed events with their gravitational wave counterparts: demonstrating that the gravitational wave signals are indeed real, physical signals that correspond to independently observable astronomical events. And the idea that gravity might emerge from another entity, like entropy, yield absurd results for the problem of dark matter, as they predict ratios of (what we perceive as) dark matter to normal matter that would change over time, rather than remain constant: required by cosmic observations.
It’s not enough, by scientific standards, to simply propose a wild idea that explains one property that the prevailing, presently accepted theory has difficulty with. One new observation can always be explained by one new “free parameter,” which is a kindhearted way to say “invoking something brand new.” If that new theoretical addition lacks the power to explain other phenomena as well, however, it’s not likely to gain serious traction of any type, and if you dare invoke an extra new free parameter atop the first one, you’ve just demolished your own credibility.
2.) Many “new ideas” are unoriginal: old, discredited ideas that have simply been repackaged, where nothing has changed to merit their reconsideration.
Most of us, if we have any sort of imagination at all, have played the “what if” game about some aspect of reality at some point. Perhaps you’ve wondered about this yourself, and have had ideas such as:
- What if you traveled in a straight line through the Universe for a long enough distance; would you ever come back to your starting point?
- What if the particles that we think of as fundamental today — quarks, electrons, photons, etc. — are actually composite particles made up of more fundamental components?
- What if there’s some sort of extra, new field in the Universe that permeates all of space, and that’s the explanation behind what we’re currently calling “dark matter” and “dark energy?”
All of these ideas are, in fact, actually good ideas: fertile ground for theorists to cultivate. There are many scientific papers from the 20th and 21st centuries that have ultimately sprung forth from these seeds, and some of those ideas have now been explored in great detail.
But each and every one of them has faced difficulties that ultimately led to them being abandoned, save for those who insist on working on such an idea until it can be 100% ruled out. (Which, to be honest, no idea ever can be.) In all of those cases, no new evidence has subsequently come in that would lead to us favoring them over the prevailing theories. For example:
- The idea that the Universe might have a non-trivial topology continues to be interesting, but if it does, the evidence demonstrates that whatever the “size” of the Universe is, it must be significantly larger (by at least a factor of hundreds) than the entire observable Universe, due to curvature constraints.
- Similarly, if any of our (presently thought-to-be) fundamental particles are actually composite particles, they don’t display that compositeness under any of the experimental conditions we’ve ever probed.
- And if there is no dark matter or dark energy, but rather a field explanation instead, then that explanation requires at least two new free parameters: a “clumpy” one that behaves like dark matter, and a “smooth” one that behaves like dark energy.
You don’t gain anything by making these reformulations, and in many cases, you’ve just added more complexity to account for a puzzle that gives you no additional predictive power over the prevailing theory. There’s no reason you can’t explore these avenues, but unless you can either explain something the leading current theory cannot or you can reduce the number of free parameters required by the theory overall, you haven’t done anything more than play in the sandbox.
3.) It’s fundamentally unscientific to begin with an ideologically-motivated conclusion.
This is one of the most dangerous pitfalls that scientists — particularly young and/or inexperienced scientists — can fall into. If you have a puzzle or problem that vexes or fascinates you, you might have a thought along the lines of, “Wouldn’t it be fascinating if (insert your favorite speculative idea) explained what we were seeing?” There’s absolutely nothing wrong with having that thought, and there isn’t even anything wrong with exploring the theoretical consequences of what your idea would imply for the rest of the Universe: particularly if it’s connected to anything we have the capacity to observe, test, or measure.
But there’s a line that, once you cross it, pushes you over the line from legitimate scientist into crackpot territory: when you become convinced that your idea must be correct. As soon as you take that leap, you’ve decided that “I know what the conclusion must be,” and that leads down a dangerous road that too many people have fallen into. What always happens is that the idea gets fiddled with, fine-tuned, or introduces extra mechanisms until it reaches the conclusion that you were attempting to draw from the outset. This type of model-building-by-working-backward might give you the result you want, but there’s an enormous problem with it: the result you obtained was antithetical to the practice of science.
Many extremely good, world-famous scientists have fallen right into this trap.
- Fred Hoyle became convinced that the Universe must be in a steady state and could not have a hot, dense origin, despite the overwhelming evidence supporting the Big Bang.
- Arthur Eddington was convinced that the stars in the Universe could never achieve properties beyond certain limits, despite the observational evidence that those limits were frequently exceeded.
- Even Einstein himself, later in life, became convinced that quantum “randomness” must have a deterministic explanation and that gravity and classical electromagnetism would lead to a unified force; these avenues led to no consequential results during the 1940s and 1950s: the final 20 years of Einstein’s life.
In many ways, these influential scientists held back progress in their field substantially until their deaths: many scientists followed them down these dead-ends, despite the lack of merits of their arguments on scientific grounds. The lesson you should take away is that your physical intuition — regardless of who you are or what you’ve accomplished — is no substitute for the legitimate information that we gain by asking the Universe questions about itself. It’s why Johannes Kepler, who threw away his “beautiful” theory of nested spheres and perfect solids for the “ugly” theory of elliptical orbits that fit the data better than any other, remains such a spectacular (and rare) role model for how to conduct science properly.
4.) Perhaps the most important part of a scientist’s job is to rigorously attack their own hypothesis, whereas “new idea proponents” often fail to even attempt that very job.
Did you have an idea, and did you fall in love with it? Do you want it to be true, and are you convinced that when you’ve finished working it all out, it’ll be so beautiful that it must be correct? Many of us go down that road, and this is a tremendous problem. Not for science, mind you, but for ourselves. In science, it falls to us, the scientists doing the developmental work on a theory, to be the harshest critics of our own ideas. Because we will be the first ones to explore these ideas in-depth, before we present our findings to the world, we must raise key objections and answer them ourselves, before they are evaluated by others. If you fail in the attempt to knock down your own ideas — to find its weak points, to expose where its range of validity ends, to identify where it compares unfavorably with the theory it seeks to replace — others will do that work for you: very likely exposing not only your incompetence, but your failed methodology.
That is not cruelty.
That is not close-mindedness.
And that is certainly not adherence to dogma.
That is a necessary part of science: to subject any new hypothesis to rigorous scrutiny and evaluation. Though it may be unfortunate, most “new ideas” will fall apart under the weight of the evidence that has already been collected, just as most of the ideas that are originally proposed to explain a novel phenomenon turn out to fail spectacularly at describing the full suite of evidence that the Universe provides.
Credit: M. Cordiner et al., Astrophysical Journal Letters submitted, 2025
It’s easy to understand why, if you have an idea that you love, you’d want others to love it, too. But it’s very difficult to convince other scientists — particularly the scientists who embrace the notion of having appropriately high levels of skepticism for new ideas — that your idea is worth loving if you haven’t subjected it to the necessary scrutiny. If you want to propose a theory where the speed of light is different for different wavelengths of light, it had better not disagree with any of the multiwavelength observations we’ve already collected about light from distant objects, for example.
If you have an idea that falls outside of the mainstream, there are some questions you’ll definitely want to ask.
- What is the problem you’re considering that motivated this idea?
- How does this idea compare to the prevailing theory when applied to this specific phenomenon?
- How does this idea compare to the prevailing theory when applied to the other major successes of the prevailing theory?
- And what are some critical tests that you can legitimately perform (with current or near-future technology) to further discern your idea versus the prevailing theory?
As Richard Feynman once put it so eloquently, “The first principle is that you must not fool yourself — and you are the easiest person to fool.”
It is not an act of cruelty, dogmatism, or close-mindedness to demand scientific rigor. Instead, it is a mark of integrity and a commitment to finding the scientific truth surrounding whatever issue or phenomenon you’re investigating. There are many great, brilliant ideas that have been relegated to the historical dustbin of failed theories for the best reason of all: because they did not successfully line up with our observed reality. No matter how fanciful or compelling an idea may be, if it disagrees with experiment, measurement, and observation, it’s wrong, and must be revised or discarded in its current incarnation.
There are plenty of compelling, interesting, and viable ideas that are out there, and there will always be plenty of room for speculation about the unknown. But whenever we consider a novel, alternative idea, we have to do it through the lens of scientific rigor. We cannot simply pick-and-choose the phenomena we wish to pay attention to while ignoring the aspects of reality that are inconvenient for our pet ideas. Here in 2026, in the era of sycophantic LLMs, this is a greater danger than ever.
In the end, however, the Universe will always be the ultimate arbiter of what is real and what theories best describe our reality. That’s why it’s up to us — the intelligent beings that conduct the enterprise of science — to rigorously uncover those truths scrupulously: by interrogating the Universe in ways that force it to give up its secrets, and then to let the data we’ve collected guide us towards our conclusions. Unless we do it responsibly, we run the risk of fooling ourselves into believing what we want to be true, rather than what the Universe is actually telling us.
This article was first published in February of 2023. It was updated in May of 2026.
This article Yes, scientists can be hostile to new ideas. So should you is featured on Big Think.