Epstein and The End of “Pure” Science
HOBOKEN, FEBRUARY 26, 2026. Two things have been happening: One: Podcasters are asking me how I feel about The End of Science, which was published 30 years ago. Do I still cling to my ridiculous claim that science is over? Amateur astronomers in Princeton have invited me to give a talk on the book in May, and I’m trying to gather my thoughts.
Two: I’ve been asked for my take on the Epstein scandal. I never met or corresponded with this world-class sexual predator, but I know many people who did, and my name pops up in files posted by the Department of Justice. Details below.
Here’s the hybrid question I’m pondering: Is the Epstein mess making me reconsider my end-of-science argument? Short answer: Yes, I now think my gloomy book wasn’t gloomy enough.
Let me explain.
The End of Science argues that science is becoming a victim of its own success. Modern scientists will never top science’s greatest hits, notably the big bang, relativity, quantum theory, evolution, DNA-based genetics, the second law of thermodynamics.
Nor will science solve the big mysteries of existence: how our universe came to be, how life began, how matter makes a conscious mind. As they flail against science’s limits—physical, economic, political, cognitive—scientists will become increasingly desperate. They will indulge in the unconfirmable speculation that I call “ironic science,” and they will hype their work more aggressively.
All this, I humbly submit, has come to pass. The big mysteries look more intractable than ever. Physicists have failed to find a “theory of everything” that cracks the riddle of the cosmos. Theorists are still stumbling around in the cul de sacs of string and multiverse theories, my prime examples of ironic science.
We still don’t know how life began or whether it was inevitable or a fluke. And “ironic science” is too mild a term for all the “explanations” of consciousness floating around out there. In my recent books My Quantum Experiment and Mind-Body Problems, I argue that science, far from converging on answers to big questions, is drifting further into incoherence.
If anything, science is in worse shape than I envisioned. End of Science focuses on what I call “pure” science, which seeks understanding for its own sake, for the sheer thrill of illumination. I resist offering predictions on applied science, other than an ironic riff on the future of superintelligent machines (yeah, that was a thing in the 1990s too).
But applied as well as pure science seems to be running out of gas. A 2023 study of scientific papers and patents reports that “progress is slowing in several major fields.” Papers and patents “are increasingly less likely to break with the past in ways that push science and technology in new directions.”
This study corroborates others showing declines in science's productivity. As genuine progress has stalled, hype has surged. A 2015 study of the biomedical literature found that between 1974 and 2014 terms like “novel,” “innovative” and “unprecedented” multiplied 15-fold.
Then there is the so-called replication crisis, spawned by the revelation that most peer-reviewed papers are “false,” as statistician John Ionannidis puts it in his blockbuster 2005 paper. Ioannidis asserts in a 2024 update that “most research done to date has used nonreproducible, nontransparent, and suboptimal research practices.”
He warns that science, because of its increasing dependence on computation, “is becoming more massive and more complex,” and there is “an inverse relationship between transparency and complexity.”
Speaking of computation, doesn’t the recent surge of artificial intelligence demolish my end-of-science thesis? Won’t this applied-science advance catalyze further leaps forward, both applied and “pure”? That’s what boosters like Sam Altman assure us.
I admit I didn’t see ChatGPT and other new AIs coming, nor did I anticipate how rapidly the public would embrace these clever technologies. But will AI lead to genuine scientific breakthroughs? Cures for cancer and schizophrenia? Solutions to the cosmic mysteries? I doubt it.
I’ve been asking my students about their use of AI. Almost all of them use AI almost all the time for schoolwork and almost everything else. Advice on food, exercise, entertainment, relationships. My students fear they’re becoming too reliant on AI, it’s becoming a crutch. So they confess in AI-cowritten papers.
AI is enabling scientists to churn out papers at an accelerating rate, leading to the “AI slop” crisis, a monstrous, mutant version of the replication crisis. Far from saving science, I see AI as catastrophic for the whole enterprise of serious intellectual inquiry. We’re outsourcing our thinking to machines.
When I wrote End of Science 30 years ago, I assumed that what might be called “civilization”—which upholds justice and truth as ideals--would endure. Increasingly, I see that assumption as sentimental and naïvely optimistic. The president of my country disdains justice and truth, he cares only about power. He has slashed funding for both “pure” and applied science.
Altman and other tech moguls who have allied themselves with Trump don’t give a shit about justice and truth, either. These rich, powerful guys care only about becoming richer and more powerful.
And that is the norm, historically, for patrons of science, according to James McClellan, an historian of science. Through the ages, my pal Jim informs me, emperors, kings, politicians, CEOs and other funders of research have seen it primarily as a source of profits and power. “Pure science”? Pursued for its own sake? For the sake of “truth”? Ha.
I can no longer say “pure science” without scare quotes—which brings me back to Epstein. The scandal provides especially sordid evidence of how depressingly short modern science falls from the ideal of passionate truth-seeking.
Exchanges between Epstein and scientists remind me of argumentative theory, a conjecture about the evolutionary origins of science’s wellspring, “reason.” The theory says “reason” (again, scare quotes seem appropriate) has always been more about boosting status and hence sexual opportunities than discovering truth.
Like his old buddy Trump, Epstein harbored eugenic fantasies, in which he and other superior men spread their seed widely. Epstein’s narcissistic delusions dovetail with the sci-fi aspirations of tech billionaires such as Musk and Bezos, who dream of colonizing space and becoming immortal.
Epstein hung out with scientists I profiled in End of Science, including Murray Gell-Mann, Marvin Minsky, Richard Dawkins and Noam (oh say it ain’t so!) Chomsky. Epstein, I’ve learned, funded platforms to which I contributed, such as the website Edge.org and the syndicated series “Closer to Truth.” He purchased End of Science from Amazon in April 2019, just months before he hung himself in prison.
I was never invited to Epstein’s island or one of his parties. If I had been, I probably would have gone and then written something snarky about him. That’s how I dealt with my ethical qualms about taking money from the Templeton Foundation, Deepak Chopra and a CIA contractor.
I’m not so pure either.
Further Reading:
For more, check out the preface of the 2015 edition of The End of Science and the full list of columns on this site, which you can find under “About Cross-Check.”
This column borrows from my 2019 Scientific American column “Jeffrey Epstein and the Decadence of Science.”