Debugging Done Diligently

by Brian Dunning
September 9, 2025

This episode was sponsored by The Illuminati Quality Control Officer

Sponsor an episode!

In a 1957 book, author Orlando Aloysius Battista wrote "An error doesn't become a mistake until you refuse to correct it." Well, it might take me a few months — or in some cases, a few years — but eventually I do try to get around to every error in Skeptoid and prevent them from becoming mistakes. And that's what I have for you today: another episode of corrections, as we do here on Skeptoid — and as we have always done — since the very beginning. Thus, although we're not always error-free, we do try to remain mistake-free.

We're going to get started with a correction pertaining to cryptocurrency, artificial intelligence, and generating the power that's needed to make them possible.

Is Bitcoin "Legal Tender"?

In episode #994 on the challenges of making artificial intelligence environmentally friendly (as its infrastructure is such an incredible electricity hog), I happened to mention that Japan had been the first country to declare Bitcoin as legal tender. Listener Ron wrote in with a seemingly minor, but actually quite important, correction:

Technically this isn't quite right. In 2016 Bitcoin was classified (or recognised) in Japanese law as property or an asset (rather than simply a ‘token'). In practice, this meant crypto-asset exchanges would be regulated by Japan's Financial Services Agency. Regulation should boost confidence and user protection.

This is not quite the same as becoming legal tender. The term legal tender refers to money that must be accepted if offered in payment of a debt under a country's laws. This means for instance, you can pay your taxes using this money, or it must be accepted in a court of law for the repayment of debts.

In Japan people can accept payment in Bitcoin if they want to... but they do not have to accept Bitcoin. Nor can anyone pay their taxes in Bitcoin.

All the best, Brian. Thanks for everything you do to promote scientific thinking.

I will confess that's a distinction that I was not aware of. But there it is: Japan was the first to recognize Bitcoin as an asset that can be regulated, but that's not the same as legal tender. Thank you, Ron! —and the transcript has been duly updated. Ron also provided a valid reference, which I always ask for with corrections.

The First Paranormal Prize

In episode #989, guest host Jeff Wagg talked about paranormal prizes such as the great James Randi's famous Million Dollar Challenge. In the show, Jeff said that such challenges have a history going all the way back to the first one, which was in the 1920s when magician Harry Houdini offered a prize of $10,000 to anyone who could demonstrate a paranormal ability. Listener Richard wrote in:

I have to tell you that it is not exact: the first to my knowledge is that of Gustave Le Bon, who founded a prize for 500₣ in 1905 (raised to 1,000₣ by Prince Roland Bonaparte, member of the Academy of Sciences, then to 2,000₣ by Dr Xavier Dariex, director of the Annales des sciences psychics) to any medium who moves an object, without touching it, in broad daylight and under the control of two conjurers.

We spoke about it ourselves, with Henri Broch, in our book 50 Years of Zététique two years ago.

Bonjour Brian, thanks for your amazing work.

My thanks to Richard, not just for the correction but also for introducing me to Le Bon, of whom I had not heard before. Check out the Wikipedia page for Gustave Le Bon. He wasn't a magician, but rather a polymath with a surprisingly wide range of real expertise. He was so accomplished that I'll say no more, but rather leave it to you to discover on your own.

Hazard vs Risk

Episode #985 was about supervolcanoes and super earthquakes, and in it, I discussed the important difference between hazard and risk. The example I gave was that falling into the Sun would be a very hazardous place to be; but your risk of that happening is very low. However, I gave a wrongly incomplete definition of risk as simply being the likelihood of a hazard happening. The correction came in quickly from John Sifling, Skeptoid's board chair — technically and (ahem) currently my boss:

Actually, Risk is the product of Likelihood (probability of a hazard occurring) and the Severity (potential impact or consequences of the hazard), i.e. R = L × S. There are more sophisticated formulations, but this is the basic one that is most often used.

This having been said, I thought your overall discussion of Hazards and Risk in the episode was excellent and a very good topic as people tend to both over and underestimate actual Risk based on various cognitive biases (e.g., the availability heuristic).

John has 35 years of emergency management expertise, including both formal training and practical experience in risk management in such roles as an Incident Commander with the Coast Guard during events like the Exxon Valdez oil spill — so we'll throw him a bone and give him the benefit of the doubt on this one.

Planet Killers

Episode #982 was about defending against the planet killers, all the various countermeasures we have against all the various ways the universe might try to kill us (spoiler alert: that's a battle we'll almost certainly lose if and when it happens). I mentioned mass extinctions that happened 372 million years ago and 445 million years ago. Listener Jeff — or in this case, reader Jeff — pointed out the following:

Doesn't Ga represent a billion years? Feel like maybe you meant Ma?

He's correct. Although I correctly said millions in the podcast, anyone reading the transcript would have seen that I incorrectly wrote "372 Ga and 445 Ga" — Ga standing for gigaannum, a billion years before today. The transcript has been corrected to Ma, megaannum, a million years before today.

We had a second correction on this same episode. A lot of the methods discussed had to do with changing the orbit of planet-killing asteroids to make them miss the Earth, and a lot of these had to do with adding or removing material from these objects, and I said something like this change to the object's mass would change its orbit. Listener Juan wrote:

You made an error in your discussion of asteroid planet killers. Changing the mass of anything orbiting anything else under the influence of gravity has absolutely zero effect on the orbit. If the mass goes down the gravitational force goes down in an exactly proportional way, meaning that there's absolutely no change in the acceleration on the asteroid. All those methods that you described changing the mass of the asteroid, the important thing is that they are ablating or otherwise blasting material off the asteroid which applies an additional force on the asteroid changing its orbit.

I'll admit it to it being somewhat of a detail, but as a former astrophysics professor, I care about these details.

Thanks for the pod! Keep up your efforts at fighting woo!

Now I've known for years that changing the mass of an orbiting body has no impact on its orbit — I learned that when I first started studying exoplanets and learned that although we can sometimes detect the details of its orbit, that tells us nothing about its mass. So when I was researching this episode, a couple of articles that I used for the source on this indicated that changing the objects' mass was what would change their orbits. I read them carefully and shook my head and decided these writers must know something I don't, or there must be some other effect at play or something I didn't understand — I mean, come on, celestial mechanics is complicated stuff.

After getting Juan's email, I went back to the sources and re-read them with the newly refocused context, and sure enough, I must have been misreading them or something. It is indeed the force exerted by the ablation method itself that causes the change in orbit. The sources are correct, Juan is correct, and now the Skeptoid transcript is correct as well.

How Big Is Germany?

I threw a nice goof in episode #980. Plato's description of the size of the central plain of Atlantis converts to just about 150,000 km² which I said was bigger than Germany. Listener David was quick to jump on that:

It seems you might have mixed up square kilometers with square miles regarding the size of Germany.

And it turns out that's exactly what I did. I had a list of countries sorted by area, and didn't notice that it was in square miles; I just assumed it was in square kilometers. That would be bigger than Germany, which is about 138,000 square miles. But in square kilometers it's 357,000 — so in the transcript I replace it with Illinois, a better comparison to Plato's description of Atlantis, at 150,000 km².

Work long enough and you're going to make bonehead errors, just as I did here.

Jailing the Italian Seismologists

Next we have a mistake that I caught on my own. Episode #873 was about earthquake prediction — a thing that's still impossible to do with any useful combination of accuracy and lead time. You may recall a news story that after a major quake in Italy in 2009 that killed over 300 people, Italian prosecutors charged and convicted seven seismologists with manslaughter. The way this was popularly characterized on social media and low-quality news sources was that they were convicted for failing to predict the earthquake and warn everyone — a thing that's impossible to do; but if you don't have a decent level of science literacy (as the Italian prosecutors likely did not), you might not know that. This is the way I characterized the incident in my episode as well.

And as you may be able to guess, that's not what happened at all. The situation was more nuanced and had nothing to do with failing to predict the earthquake. The seismologists who were convicted had been charged with downplaying the likelihood of a deadly quake, which they did under instructions at a media event — they stated that a series of smaller quakes meant energy had been released, minimizing the chances of a larger quake — though that's not how earthquakes work. Also their convictions have since all been overturned, as the media event was done under orders.

It would be wrong to prosecute scientists for not knowing things we can't yet know, and it's also wrong to tell people they don't have to worry because there will be no earthquake.

Well that's all we have time for today, which is a bummer because I still have a few left. But luckily there's no end in sight for Skeptoid, thus there is no upper limit on the number of errors we'll continue to make and to correct. That number of errors can actually be computed: given N as the number of episodes, k as the frequency at which errors occur, and E as the total number of errors after N episodes, we have:

So now you know that.

By Brian Dunning
Follow @BrianDunning