I am a latecomer to AI (Artificial Intelligence), being probably one of the last people I know to try publicly accessible AI engines such as Microsoft Copilot and Chat GPT (lots of others are available). It’s not possible to avoid the noise in the media in recent years about if and how AI could affect our jobs. But one thing that’s clear is that the art of creative writing; in fact many forms of artistry including music, film making and painting, are potentially under threat from our robotic cousins (or robotic overlords, depending on how much of a conspiracy theorist you are).
So with this thought in mind, I decided to test out how good a job AI could do at producing one of my blogs. I’ve chosen three questions, totally at random, to test out not just the quality of the content but also its ability to entertain. After all, a blog needs to be engaging as well as informative. Let’s dive straight in. AI answers are in blue, just in case you already found my writing robotic and assumed it was computer generated.
Question 1:
Steve:
Mister Copilot (I thought I’d be formal at first), how does thermal ionization mass spectrometry work?
Copilot:
Thermal ionization mass spectrometry (TIMS) is a technique used to measure isotope ratios with remarkable precision. It works by heating a sample to high temperatures until it ionizes, allowing the isotopes to be separated based on their mass-to-charge ratio. Here’s how it unfolds step by step:
1. Sample Preparation & Loading
The sample, often in a liquid form, is deposited onto a metallic filament made of materials like rhenium or tungsten. Once dried, this ensures an even distribution of the element of interest.
2. Thermal Ionization
The filament is heated under vacuum conditions, causing atoms in the sample to evaporate. Some of these atoms lose an electron due to thermionic emission, producing positively charged ions without additional external ionization processes.
3. Ion Extraction & Acceleration
An electric field accelerates the newly formed ions, imparting them with a uniform kinetic energy to ensure they travel consistently through the mass spectrometer.
4. Mass Separation
A magnetic field deflects the ions based on their mass-to-charge ratio (m/zm/z). Lighter ions curve more than heavier ones, allowing them to be spatially separated.
5. Detection & Isotope Ratio Calculation
The separated ions strike a detector, which measures their intensities and calculates the ratio of isotopes present in the sample. Scientists then analyze these ratios to draw conclusions about geological processes, environmental changes, or even forensic investigations.
TIMS stands out for its high precision, making it invaluable in applications like geochronology (Rb-Sr or U-Pb dating) and tracing material origins in planetary sciences. It’s slower than other techniques like ICP-MS, but its accuracy remains unmatched.
So, what do we think? I actually think it is pretty good! Possibly a bit dry, but in all likelihood more accurate and concise than any description I could come up with. In other words, as a way to produce a factual description, AI has probable gotten the better for me. So I’m going to say this one is a win for Copilot. Let’s move on to question two:
Question 2:
Steve:
Captain Copilot (I decided “captain” seemed a better way to address a pilot), tell me a funny story about an engineer missing his flight in Japan…
Copilot:
An engineer—let’s call him Tom—was in Japan for a tech conference in Tokyo. As expected, he was awed by the country’s efficiency, precision, and ultra-modern transportation systems. Tom had spent a week marveling at bullet trains that arrived exactly on time, vending machines that dispensed everything from hot meals to ties, and toilets that were smarter than his smartphone.
On the last day, Tom confidently checked out of his hotel, thinking, “I’ve got Japan figured out.” He had precisely calculated the time it would take to reach Narita Airport using multiple spreadsheets—because, of course, he was an engineer. But here’s where things went sideways.
Tom, ever the meticulous planner, set out for his flight with plenty of time. But as he arrived at the train station, he noticed something terrifying—the platform was empty. Not a single soul waiting. Not even the usual polite queue he had come to admire.
Confused, he checked the train timetable on his phone and saw a shocking reality: His perfectly planned train had left exactly two minutes ago. And the next one? Delayed! (A rare phenomenon in Japan, but fate has its own sense of humor.)
Tom panicked. His flawless, spreadsheet-driven plan crumbled before his eyes.
Frantically, he sprinted to the taxi stand. But here’s the thing about Tokyo—while taxis exist, they aren’t exactly the go-to for long airport rides unless you’re prepared to auction off your kidneys to afford the fare. The driver glanced at him, did some quick math in his head, and gave Tom the price of the ride. It was approximately equal to the cost of his return ticket to Europe.
Tom gulped. Was this the price of hubris?
At that moment, he had a flashback to his first day in Japan, when a kind local had advised him to ALWAYS give himself extra time for transit because even the most efficient system could betray the unwary. He had scoffed. But now? He was the unwary.
In the end, Tom had no choice but to take the overpriced taxi, sweating profusely as he barely made it to his gate before departure. As he boarded, exhausted and financially wounded, he whispered to himself:
“Next time, I buffer my travel plans. Even in Japan.”
And so, he left with newfound respect for Japanese punctuality… and a much lighter wallet. Moral of the story? Even engineers need real-world error margins.
So… thoughts on this one? It’s not bad, but it does seem a little soulless, which is probably about right, bearing in mind it was made up by a box of wires and circuits in under three seconds (really!). I’m also not sure about all of the incessant and unnecessary bold. Therefore I’m going to give the win to me on this one; I think my real experiences are always going to make for better stories than ones made up by AI. But please let me know if you disagree.
And finally we move on to question three. I’ve decided to add one that I haven’t yet covered in my blogs, but hope to at some stage soon…
Question 3:
Steve:
Hey computer (I am not sure the formality is working), what do I do if my mass spectrometer is possessed by a ghost?
Copilot:
Ah, the age-old dilemma: scientific precision meets supernatural chaos. If your mass spectrometer has started making eerie noises, displaying inexplicable isotope ratios, or, worse, whispering chemical formulas in the dead of night, you may have yourself a paranormal analytical instrument.
Steps to Identify the Problem:
1. Check for Logical Explanations First – Before calling in an exorcist, ensure your spectrometer isn’t suffering from mundane issues like electrical interference, vacuum leaks, or software glitches. Ghosts are rare, but faulty wiring? Very common.
2. Monitor Unusual Readings – If you keep detecting spectral anomalies (pun intended), double-check calibration settings. If your machine is mysteriously identifying elements from another dimension, something’s definitely up.
3. Consult the Manufacturer – Tech support may not have a “Ghost Troubleshooting” section, but asking them about strange behaviors could reveal whether spectral artifacts are due to hardware malfunctions rather than actual spirits.
4. Perform a Scientific Séance – Gather your team, wear lab coats like armor, and see if the spectrometer responds to polite negotiation. If it starts generating Latin inscriptions or ancient alchemical formulas, back away slowly.
5. Call a Specialist – If all scientific methods fail, consider reaching out to an expert—whether a physicist or a paranormal investigator. Either way, your lab will make history.
Bonus Safety Tip:
Avoid running cursed samples through the spectrometer. No radioactive artifacts from abandoned temples, no vials labeled “DO NOT ANALYZE.” Science is about discovery, but some things are better left unknown.
So what are your thoughts about question three? I thought it as pretty good, just missing some specific examples that might have “humanized” the story. And again, Mr. Copilot needs to learn to use bold more sparingly. But it did have quite a bit of humour, and in my opinion made for quite a well written and light-hearted article.
So there we are, three examples of AI-produced blog content. In light of these three examples, am I out of a job or are my blogs human enough to continue with them? Only you can decide…
That’s all for now. If you have any stories or experiences regarding AI and its applications to analytical science, please let me know, I’d love to hear about them. As always, send your bots to:
