Social media measures the success of an idea by how many “likes” it gets. Scientific success is measured by how close the idea is to the truth. In the physical sciences, the truth is synonymous with experimental evidence. One would therefore expect physicists to measure success by how well their ideas match data rather than by how popular these ideas are among their peers. Surprisingly, this naive expectation is not manifested throughout the current landscape of theoretical physics.

The mathematical constructions of supersymmetry, string theory, Hawking radiation, anti-de Sitter/conformal field theory (AdS/CFT) and the multiverse  are currently considered irrefutable and self-evident by the mainstream of theoretical physics, even without experimental evidence to support them. In the words of a prominent physicist at a conference that I attended a few months ago: “These ideas must be true even without experimental testimony in their favor, because thousands of physicists believe in them and it is difficult to imagine that such a large community of mathematically-gifted scientists would be wrong.”

Once a mainstream culture grows to this self-sustaining phase, it does not need external verification. The ideas it advocates are reasoned to be inherently correct based on their mathematical beauty, with experiments serving the optional role of narrowing down the wide range of possibilities allowed by the flexible mathematical framework. Past generations of theoretical physicists were less arrogant; among the possibilities they contemplated was one that allowed their theories to be proven wrong by experimental data.

But the current self-sustained culture thrives in its own theoretical sauce, dismissing alternatives because they are getting fewer “likes.” When award or grant-allocation committees are populated by advocates of the popular paradigm, it could take centuries to correct a path that should not have been taken in the first place. Large enough groups can legitimize speculative concepts without adhering to Carl Sagan’s statement “Extraordinary claims require extraordinary evidence.”

To his I would add the basic lesson from Galileo Galilei that experimentation is crucial because extraordinary groupthink leads to extraordinary ignorance. Before Galileo’s observations, it was popular to construct beautiful abstract frameworks on the assumptions that heavy objects fall faster than light objects under gravity and that the sun revolves around the Earth.

Is there anything new and alarming about the self-sustaining culture of some current physicists or was it always around, even after Galileo? My personal impression is that half a century ago, theoretical physicists were far more disposed to the concept of experimental vindication. But right now, if we double down on supersymmetry as being just around the corner when the Large Hadron Collider did not find evidence for it; if we insist that Hawking radiation must exist despite the paradox that is spells between Einstein’s general relativity and the fundamental principle of no information loss in quantum mechanics; if we posit that the multiverse must exist, and anything that can happen will happen in it an infinite number of times—without evidence to support this notion—then we are betraying the trademark of physics as an effort to describe the reality we live in.

After all, when I consider theoretically all possible amounts of money that I could have had in my bank account, I can get very excited. In particular, there is this attractive possibility that I am a billionaire. But going to an automated teller machine (ATM) and finding how much money I actually have, introduces a sobering tone to my spending habits. Getting data and comparing them to our theoretical ideas provides a reality check that we are not hallucinating or engaged in wishful thinking.

Even though I never tried this experience myself, I can imagine that avoiding feedback from experimental data must be similar to being on drugs. Based on the reports from those who had the experience, “you are floating in constructions of your imagination and feeling happy.” A group of dreamers can feel happier together by sharing experiences and supporting each other.

Yet this activity misses the central point. Physics is not supposed to be a recreational activity that makes us feel good about ourselves. History teaches us that groups of humans can feel happy in the company of each other while advocating the wrong ideas.  Science is a learning experience about nature that holds the potential of showing us wrong, irrespective of our popularity status on Twitter. Physics is a dialogue with nature, not a monologue. We are supposed to have skin in the game and make testable predictions.

Nevertheless, there are some physicists who even advocate the hypothesis that we might be living in a computer simulation. Here again, the appropriate response is to be guided by evidence. This proposal should gain traction only if we notice pixels in spacetime as we see on a computer screen or if we detect a bug indicating that the simulation had crashed. So far, reality appears very real to me. Perhaps this is because, unlike some advocates of this idea I have to check the balance in my bank account in order to avoid the consequences of acting irresponsibly.

Social pressure does not only fashion popular speculations in the realm of theory, but also limits the empirical exploration of far less speculative notions with instruments that we readily have. For example, the search for technological signatures of civilizations on exoplanets—in the form of industrial pollution, artificial lights or heat, photovoltaic cells, structural artifacts or artificial satellites—could have been conducted with far more rigor if not for the reluctance of the mainstream to pursue this task. The related prejudice and peer pressure echo the refusal of some philosophers to look through Galileo’s telescope.

How can physics recover its traditional humility with respect to evidence in the age of social media? The path forward is simple: priority should be given to gathering experimental data and ruling out theoretical ideas. There is nothing more humbling than being guided by data. Rather than spending full careers going down mathematical alleys that might be declared irrelevant by future generations of physicists, it would be prudent for young scientists to focus on those areas of research where the value of ideas can be tested and cashed during their lifetime.

Without experimenting at a nearby ATM, we run the risk of discovering one day that we are bankrupt. And the speculative notion that in some parts of the multiverse there might be a wealthier version of ourselves will not save our skin in this one reality to which we are held accountable.