Human-Written Or Machine-Generated: Finding Intelligence In Language Models

What is the essential element which separates a text written by a human being from a text which has been generated by an algorithm, when said algorithm uses a massive database of human-written texts as its input? This would seem to be the fundamental struggle which society currently deals with, as the prospect of a future looms in which students can have essays auto-generated from large language models (LLMs) and authors can churn out books by the dozen without doing more than asking said algorithm to write it for them, using nothing more than a query containing the desired contents as the human inputs.

Due to the immense amount of human-generated text in such an LLM, in its output there’s a definite overlap between machine-generated text and the average prose by a human author. Statistical methods of detecting the former are also increasingly hamstrung by the human developers and other human workers behind these text-generating algorithms, creating just enough human-like randomness in the algorithm’s predictive vocabulary to convince the casual reader that it was written by a fellow human.

Perhaps the best way to detect machine-generated text may just be found in that one quality that these algorithms are often advertised with, yet which they in reality are completely devoid of: intelligence.

Continue reading “Human-Written Or Machine-Generated: Finding Intelligence In Language Models”

DB Cooper Case Could Close Soon Thanks To Particle Evidence

It’s one of the strangest unsolved cases, and even though the FBI closed their investigation back in 2016, this may be the year it cracks wide open. On November 24, 1971, Dan Cooper, who would become known as DB Cooper due to a mistake by the media, skyjacked a Boeing 727 — Northwest Orient Airlines Flight 305 — headed from Portland to Seattle.

During the flight, mild-mannered Cooper coolly notified a flight attendant sitting behind him via neatly-handwritten note that he had a bomb in his briefcase. His demands were a sum of $200,000 (about $1.5 M today) and four parachutes once they got to Seattle. Upon landing, Cooper released the passengers and demanded that the plane be refueled and pointed toward Mexico City with him and most of the original crew aboard. But around 30 minutes into the flight, Cooper opened the plane’s aft staircase and vanished, parachuting into the night sky.

In the investigation that followed, the FBI recovered Cooper’s clip-on tie, tie clip, and two of the four parachutes. While it’s unclear why Cooper would have left the tie behind, it has become the biggest source of evidence for identifying him. New evidence shows that a previously unidentified particle on the tie has been identified as “titanium smeared with stainless steel”.

Continue reading “DB Cooper Case Could Close Soon Thanks To Particle Evidence”

Predicting The A-Bomb: The Cartmill Affair

The cover of the infamous issue of Astounding, March 1944

There’s an upcoming movie, Argylle, about an author whose spy novels are a little too accurate, and she becomes a target of a real-life spy game. We haven’t seen the movie, but it made us think of a similar espionage caper from 1944 involving science fiction author Cleve Cartmill. The whole thing played out in the pages of Astounding magazine (now Analog) and involved several other science fiction luminaries ranging from John W. Campbell to Isaac Asimov. It is a great story about how science is — well, science — and no amount of secrecy or legislation can hide it.

In 1943, Cartmill queried Campbell about the possibility of a story that would be known as “Deadline.” It wasn’t his first story, nor would it be his last. But it nearly put him in a Federal prison. Why?  The story dealt with an atomic bomb.

Nothing New

By itself, that’s probably not a big deal. H.G. Wells wrote “The World Set Free” in 1914, where he predicted nuclear weapons. But in 1914, it wasn’t clear how that would work exactly. Wells mentioned “uranium and thorium” and wrote a reasonable account of the destructive power: Continue reading “Predicting The A-Bomb: The Cartmill Affair”

Ask Hackaday: Why Are Self-Checkouts Failing?

Most people who read Hackaday have positive feelings about automation. (Notice we said most.) How many times have you been behind someone in a grocery store line waiting for them to find a coupon, or a cashier who can’t make change without reading the screen and thought: “There has to be a better way.” The last few years have seen that better way, but now, companies are deciding the grass isn’t greener after all. The BBC reports that self-checkouts have been a “spectacular failure.” That led us to wonder why that should be true.

As a concept, everyone loves it. Stores can hire fewer cashiers. Customers, generally, like having every line open and having a speedy exit from the store. The problem is, it hasn’t really panned out that way. Self-checkout stations frequently need maintenance, often because it can’t figure out that you put something in the bag. Even when they work flawlessly, a customer might have an issue or not understand what to do. Maybe you’ve scanned something twice and need one of them backed off. Then, there are the age-restricted products that require verification. So now you have to hire a crew of not-cashiers to work at the automated not-register. Sure, you can have one person cover many registers, but when one machine is out of change, another won’t print a receipt, and two people are waiting for you to verify their beer purchase, you are back to waiting. Next thing you know, there’s a line.

Continue reading “Ask Hackaday: Why Are Self-Checkouts Failing?”

Solar Chimneys: Viable Energy Solution Or A Lot Of Hot Air?

We think of the power we generate as coming from all these different kinds of sources. Oil, gas, coal, nuclear, wind… so varied! And yet they all fundamentally come down to moving a gas through a turbine to actually spin up a generator and make some juice. Even some solar plants worked this way, using the sun’s energy to heat water into steam to spin some blades and keep the lights on.

A solar updraft tower works along these basic principles, too, but in a rather unique configuration. It’s not since the dawn of the Industrial Age that humanity went around building lots of big chimneys, and if this technology makes good sense, we could be due again. Let’s find out how it works and if it’s worth all the bluster, or if it’s just a bunch of hot air.

Continue reading “Solar Chimneys: Viable Energy Solution Or A Lot Of Hot Air?”

NASA Adjusts Course On Journey To The Moon

It’s already been more than fifty years since a human last stepped foot on another celestial body, and now that NASA has officially pushed back key elements of their Artemis program, we’re going to be waiting a bit longer before it happens again. What’s a few years compared to half a century?

The January 9th press conference was billed as a way for NASA Administrator Bill Nelson and other high-ranking officials within the space agency to give the public an update on Artemis. But those who’ve been following the program had already guessed it would end up being the official concession that NASA simply wasn’t ready to send astronauts out for a lunar flyby this year as initially planned. Pushing back this second phase of the Artemis program naturally means delaying the subsequent missions as well, though during the conference it was noted that the Artemis III mission was  already dealing with its own technical challenges.

More than just an acknowledgement of the Artemis delays, the press conference did include details on the specific issues that were holding up the program. In addition several team members were able to share information about the systems and components they’re responsible for, including insight into the hardware that’s already complete and what still needs more development time. Finally, the public was given an update on what NASA’s plans look like after landing on the Moon during the Artemis III mission, including their plans for constructing and utilizing the Lunar Gateway station.

With the understanding that even these latest plans are subject to potential changes or delays over the coming years, let’s take a look at the revised Artemis timeline.

Continue reading “NASA Adjusts Course On Journey To The Moon”

Vulcan Nails First Flight, But Peregrine Falls Short

For those with an interest in the history of spaceflight, January 8th promised to be a pretty exciting day. Those who tuned into the early morning live stream were looking forward to seeing the first flight of the Vulcan Centaur, a completely new heavy-lift booster developed by United Launch Alliance. But as noteworthy as the inaugural mission of a rocket might be under normal circumstances, this one was particularly special as it was carrying Peregrine — set to be the first American spacecraft to set down on the lunar surface since the end of the Apollo program in 1972.

Experience has taught us that spaceflight is hard, and first attempts at it doubly so. The likelihood of both vehicles performing as expected and accomplishing all of their mission goals was fairly remote to begin with, but you’ve got to start somewhere. Even in the event of a complete failure, valuable data is collected and real-world experience is gained.

Now, more than 24 hours later, we’re starting to get that data back and finding out what did and didn’t work. There’s been some disappointment for sure, but when everything is said and done, the needle definitely moved in the right direction.

Continue reading “Vulcan Nails First Flight, But Peregrine Falls Short”