There has been a lot of buzz in the media about the AI software ChatGPT and its ability to turn out essays that have at least a superficial plausibility. I decided to give it a spin and gave it variations of the same prompt on a topic about which I knew something but was probably too esoteric for non-physicists, just to see what it would come up with.
Here are the prompts and what I got back.
Prompt: Explain Bell’s theorem
Bell’s theorem is a concept in quantum mechanics that describes the phenomenon of entanglement. It states that when two particles are entangled, the properties of one particle are intimately linked to the properties of the other, regardless of the distance between them.
This means that if one particle is measured and its properties are determined, the properties of the other particle are automatically determined as well, even if the two particles are separated by great distances.
Bell’s theorem suggests that this entanglement is not due to some hidden force or connection between the particles, but rather it is a fundamental property of the universe itself. It challenges our everyday understanding of physics and suggests that there are deeper truths waiting to be discovered.
The implications of Bell’s theorem are far-reaching and have been the subject of much research and study in the field of quantum mechanics. It has also led to the development of new technologies, such as quantum computers and quantum cryptography, which rely on the principles of entanglement to function.
I then decided to play around with the prompt.
Prompt: Explain Bell’s theorem in the style of Ernest Hemingway
Bell found a theorem. It was about particles that are linked in a way that can’t be explained by everyday physics. He called it “entanglement”. It showed that our understanding of the world is incomplete. There are deeper truths waiting to be discovered.
Hemingway is famous for his short sentences and pithiness, so that checked out, though the assertion that it “showed that our understanding of the world is incomplete” is not substantiated.
Prompt: Explain Bell’s theorem in the style of J. K. Rowling
Once upon a time, in the world of physics, there was a brilliant mind named Bell. He delved deep into the mysteries of the universe, seeking to understand the very nature of reality itself. And so it was that he discovered a wondrous theorem, a theorem that revealed the strange and magical nature of entanglement.
Entanglement, you see, was a mysterious force that linked particles together in a way that defied all logic and explanation. It was as if the particles were bound together by some unseen magic, their fates forever intertwined.
Bell’s theorem uncovered the true nature of entanglement, showing that it was not just some fanciful idea, but a very real and powerful force that could not be explained by everyday physics. It hinted at a world of magic and wonder, a world that lay just beyond the veil of our everyday understanding.
The implications of Bell’s theorem were staggering. They suggested that there were secrets waiting to be unlocked, mysteries that could lead to a deeper understanding of the very fabric of the universe. And so, just like the wizards and witches of Hogwarts, Bell ventured forth, armed with his theorem and his wits, seeking to uncover the true secrets of the universe.
Prompt: Explain Bell’s theorem in the style of James Joyce’s Ulysses
In the mystic realm of physics, there was a savant known as Bell, whose quest for knowledge knew no bounds. It was in this infinite expanse that he stumbled upon a theorem that challenged the very foundations of our understanding of the universe.
This theorem, known as entanglement, revealed a profound connection between particles that transcended time and space. Bell’s findings shed light on the mysterious and bewildering nature of this phenomenon, leading to revelations that were beyond comprehension.
His theorem illuminated the vast and intricate tapestry of the universe, exposing the secrets of existence that were hidden from plain sight. It beckoned Bell to journey deeper into the unknown, to unravel the mysteries of the universe and uncover the answers that lay dormant within.
The discovery of Bell’s theorem marked a turning point in the annals of physics, a revolution that changed the course of human understanding forevermore. It paved the way for new discoveries, new horizons to be explored and new truths to be unveiled.
And so it was that Bell, with his theorem as his guide, set forth into the unknown, driven by his insatiable thirst for knowledge and his unrelenting spirit of adventure.
As I read this, it felt like the kind of pseudo-deep woo that Deepak Chopra would come up with, so that gave me the idea for my next prompt;
Prompt: Explain Bell’s theorem in the style of Deepak Chopra
My dear friend, let me begin by saying that the universe is a magnificent and awe-inspiring place, full of mysteries that challenge our understanding of reality. One such mystery is Bell’s theorem, which speaks to the fundamental nature of the universe itself.
At its core, Bell’s theorem suggests that our understanding of the world as consisting of separate, individual parts may be incomplete. It suggests that there may be a deeper, more interconnected reality that underlies the physical world as we know it.
This is because Bell’s theorem deals with the phenomenon of entanglement, which is a strange and wondrous property of quantum mechanics. Entanglement occurs when two particles become connected in such a way that their states are correlated, no matter how far apart they are from each other in space.
This means that when we measure one particle’s state, we can predict the state of the other particle with a high degree of accuracy, even though they are not physically connected in any way that we can see.
Now, some scientists have suggested that this entanglement is simply a result of random chance or coincidence. But Bell’s theorem suggests that there may be something deeper going on. It suggests that the particles may be communicating with each other in some way that we don’t yet understand.
And this communication, my friend, may be happening on a level that is beyond our current ability to measure or comprehend. It may be happening on a level that is deeply connected to the very fabric of the universe itself.
So let us embrace the mystery of Bell’s theorem, and let us continue to explore the wonders of the universe with an open and curious mind. For who knows what other mysteries we may uncover as we journey deeper into the heart of the cosmos.
Apart from saying that Bell’s theorem deals with entanglement, which the Hemingway essay succinctly captured, the rest of the essays were just fluff, though I must admit that the fluff captured the magical adventure aspects of Rowling and the woo of Chopra.
There have been concerns that this software would enable students to produce essays without doing any work.This article says that if you know the material that you are requesting essays about, you will be able to identify these essays because they are vacuous.
But university professors are catching ChatGPT assignments in the wild for a different reason: because the AI-produced essays are garbage.
“The first indicator that I was dealing with AI was that, despite the syntactic coherence of the essay, it made no sense,” wrote assistant professor of philosophy at Furman University, Darren Hicks, in a Facebook post after confronting his first ChatGPT-generated essay on ‘Hume and the paradox of horror’.
For another professor who asked to remain unnamed, it was also that the essay was jarringly, clangingly wrong, that first raised their suspicions ChatGPT may have been involved. The essay, which addressed the work of critical theorist Judith Butler, “was just nonsense,” they said. It appeared to have mashed together various sources that “talked about Butler and sexuality and gender and whatever… It was a series of sentences that made their own kind of sense individually, but together made very little sense.”
Students are often wrong too, but it’s the idiosyncratic ways in which ChatGPT is botching assignments that professors are starting to recognize. They say the essays embody a constellation of traits that trigger an uncanny valley effect in the reader. Educators are starting to share these early encounters, and their clues on how to spot the ghostly imprints of ChatGPT.
“Normally, when a student plagiarizes, it’s a cost-benefit analysis that comes up because they’re desperate,” Hicks told Motherboard. They don’t know the material and they don’t have time to complete the assessment, so they ask how likely it is that they’ll get caught. “Because it’s a last minute scramble, in most cases the essays are terrible,” he said.
Not so with ChatGPT. In the case of Hick’s ChatGPT essay, “it was wrong, but it was confident and it was clearly written,” he said. “If I didn’t know the material better, it would have looked good. And that, that’s a weird combination of flags which I’d never seen before.”
There are also the tell-tale stylistic cues. “It tends to produce essays that are filled with bland, common wisdom platitudes,” said Bret Devereaux, visiting history lecturer at the University of North Carolina at Chapel Hill, who recently encountered his own first ChatGPT-created assignment.
When I first started assigning essays that required making an argument, students would argue in favor of something they agreed with but give short shrift to opposing arguments, counter-evidence, or to other views. Some tended to write the essays at the last minute, which never produces good writing. I realized that I had been focusing on getting a good final product rather than teaching students what it takes to write well. I had been acting like an editor rather than a teacher of writing.
So I changed my teaching strategy to require students to think about the topic at length and make many revisions. Students would choose a topic early in the semester. The first essay required them to present three different viewpoints on that topic in such a neutral way that I should not be able to identify which one they preferred. The second essay required them to pick one of the viewpoints and spilt the essay into two parts, one part arguing in favor of it, the other part against it, again in such a way that I should not be able to tell which side they preferred. The final essay was their synthesis, where they presented the argument in favor of their own position while fairly treating objections to it as well as all the other points of view. Each of the three essays allowed for revisions based on feedback from their fellow students and from me. I ended up not just getting more thoughtful and balanced essays but giving students practice in what it takes to produce a good piece of writing.
I am not sure whether ChatGPT and its alternatives would be able to game my system. Since I am retired, I may never find out.
rich rutishauser says
Mano, where were you when I was learning to write! I have always felt that K-12 education and even some college teaches writing poorly…a somewhat scattershot method that requires young minds to mostly guess at how to write something, get a poor grade on it with vague suggestions like “more detail” or “be clear”, adjust what they write for the next assignment and maybe get a better grade, maybe not. There is actually very little teaching going on with this method. When I went back to college as an adult I was very lucky to have an instructor who recognized the same faults you found and taught an almost identical method…the whole class discovered we actually could write pretty well!
Rob Grigjanis says
Cranky old bastard here. Any explanation of Bell’s theorem which doesn’t include (violation of) Bell inequalities isn’t worth shit.
Steve Morrison says
Since the Vice article mentioned Bret Devereaux, here is his blog post about ChatGPT and what happened when he prompted it to write an essay on his own topic of expertise (Roman military history). Spoiler: most of what it wrote was BS.
John Morales says
Rob, too harsh. I doubt your average punter could be so accurate.
I mean, you’re complaining that it’s incomplete, not that it’s wrong.
Rob Grigjanis says
John @4: Indeed. “So vague as to be useless” is not necessarily wrong, except perhaps in an ethical sense.
John Morales says
Rob, beware trying to meet me on my ground. Never ends well.
If it’s not necessarily wrong, it follows that it sometimes it is not wrong.
Now, I know you’re alluding to Pauli’s “not even wrong” quip, but still.
Again: better than the average punter for sure.
Rob Grigjanis says
John @6: I wasn’t alluding to Pauli at all. Not sure why you’d think I was (because I used the word ‘wrong’?).
If it’s useless, so what?
John Morales says
Rob, yes, that was it. https://en.wikipedia.org/wiki/Not_even_wrong#History_of_the_expression
False positive, in this case.
If it’s useless, then it is useless; but if it’s also not wrong, it’s not wrong.
Like Bell’s theorem to the average punter.
Holms says
Oh shit, red alert -- John is about to unleash his full arsenal of pedantry! Snide disdain!! Repeated instances of ‘heh’!1!
…
/eyeroll. Get over yourself.
Marcus Ranum says
I wonder what it would say in the style of John Ringo, or Richard Feynman.
Marcus Ranum says
“Mark Twain explains physics” I’d buy 2 copies.
Rob Grigjanis says
Marcus @10: Feynman already went there, in a very Feynmanesque way.
https://aapt.scitation.org/doi/10.1119/1.4948268
Brilliant man, and a jackass.
marner says
@2 Rob (or anyone qualified to answer)
I asked it to explain it like a physicist. Is this better?
Explain Bell’s Theorem like a physicist.
Bell’s Theorem is a fundamental result in quantum mechanics that challenges the classical view of reality. To understand it, let’s start with the concept of entanglement.
Entanglement is a phenomenon where two or more particles can become correlated in such a way that their states are intrinsically linked, even when they are far apart. For example, imagine two particles, A and B, that are entangled. If we measure particle A, we instantaneously affect the state of particle B, regardless of the distance between them.
Now, let’s imagine two experimental setups, each consisting of a source that produces pairs of entangled particles and two detectors that measure the properties of these particles. We can set up these experiments so that the detectors measure two different properties of the particles, such as spin or polarization.
Classically, we would expect that the properties of the particles are determined before the measurements are made, and that the measurements simply reveal these pre-existing properties. However, in quantum mechanics, the act of measurement changes the state of the particle, so the properties are not determined until they are observed.
Bell’s Theorem shows that if we assume that the properties of the particles are pre-determined and that the act of measurement only reveals these properties, then there are certain statistical correlations that we would expect to observe in the experimental data. These correlations are known as Bell inequalities.
However, experiments have shown that these Bell inequalities are violated in entangled particle systems. This means that the classical view of reality, where particles have pre-existing properties that are revealed by measurements, cannot explain the observed correlations between entangled particles.
Instead, quantum mechanics suggests that particles do not have pre-existing properties, but rather exist in a superposition of states until they are observed. This has profound implications for our understanding of the nature of reality, and challenges some of our most deeply held assumptions about the world around us.
Rob Grigjanis says
marner @13: That’s much more meaty than the other explanations, but it’s wrong.
Bell made two assumptions, and showed that one of these must be wrong for agreement with experiment, because both would imply that Bell inequalities are satisfied, while experiment says they are violated.
The assumptions are:
(1) The axis chosen to measure the spin of the particle at one location cannot affect the measurement of the particle at the other location.
(2) The spins of the particles before measurement are determined by some hidden variable λ (this is the ‘pre-existing properties’ part).
If one of these is false, it’s still possible that the other is true. So, either a state can’t have pre-existing values, or it can have pre-existing values, but then the detector settings at one location can instantaneously affect the outcome of the other measurement.
Rob Grigjanis says
BTW, there’s nothing particularly complicated about Bell inequalities. They’re just expressions of classical logic.
For example, let A be the number of red-haired people. Let B be the number of blue-eyed people. Let C be the number of people born in Europe. Obviously, these properties will overlap. A Bell-type inequality is just this;
Let N1 be the number of people with red hair and blue eyes. Let N2 be the number of people with red hair who were born in Europe plus the number of people with blue eyes not born in Europe.
Then N1 ≤ N2
It’s easy to see this if you draw a Venn diagram of the three properties.
marner says
Thanks, Rob. I’m going to read some more about it.
As to the ChatGPT, it appears that a better question may at least moderately improves its answer. It’s interesting that someone with little to no knowledge of a topic, but is good at formulating questions and possesses superior editing/proofreading skills, might be able to quickly mimic a C student.
jrkrideau says
Steve Morrison @ 3 mentioned Brett Devereaux. Here is a Classicist who is just about as impressed. I rather liked the fictional references. Lies, damned lies, and ChatGPT
I get the feeling that a lot of students will soon be returning to traditional essay mills /s
I did see what appears to be a student on a stats programming forum yesterday write that ChatGPT had not helped with the problem they had.
John Morales says
jrkrideau:
I get the feeling you don’t get that this is still a primitive, nascent technology.