Woah that’s a big deal if true. I remember Nate Silver had a good chapter on this in his book on statistics. He went through the history of failed attempts in Italy and elsewhere. It’s a famously hard problem that has a long history of false promises and good efforts that simply didn’t work out, where experts just generally concluded it’s an unsolvable problem with current tech and the general random nature of it (in terms of usefully specific accuracy).
Yes, this also made me immediately think of that chapter in Nate Silver's book ("The Signal and the Noise"). Considering what I remember Nate's musings on the topic to be, I highly doubt that this is a particularly significant improvement in earthquake prediction. Sure, with the ability to crunch more numbers than ever before, AI can no doubt up the accuracy a bit. However, same as with the non-AI techniques that it's built upon, only a big-ish improvement in our fundamental understanding of tectonics, and/or a big-ish improvement in the physical sensors that feed data into the models, is really going to move the needle in this space.
> the AI algorithm correctly predicted 70% of earthquakes a week before they happened during a seven-month trial
> The outcome was a weekly forecast in which the AI successfully predicted 14 earthquakes within about 200 miles of where it estimated they would happen
> It missed one earthquake and gave eight false warnings.
So about 36% are false positives (which aren't terrible in this particular field) and about 6% false negatives.
This doesn't sound bad, but I'd love to know the precision of previous techniques / prior art.
It may be better than previous models, but that "70%" number is misleading:
> The outcome was a weekly forecast in which the AI successfully predicted 14 earthquakes within about 200 miles of where it estimated they would happen and at almost exactly the calculated strength. It missed one earthquake and gave eight false warnings.
So 14 true positives and 8 false positives, which means the positive predictive value is not great, less than 65%. And I didn't read the paper for the details of the timing, but the article said "a week out", so I'm assuming it means the predictions meant the earthquake could strike in a minute or in 7 days.
What would governments/policy makers actually be able to do with any of this data? Not to denigrate it as a step forward, but I'm having trouble seeing much practical impact at all.
As someone who lives in an earthquake-prone area, I don't think those stats are high enough that they would lead officials to evacuate an area as a precaution. But it would be a useful indicator to encourage people to make sure they have their earthquake supplies topped up and ready to go.
I disagree, in Japan we just had the "megaquake" warning for a week, maybe it was prudent but it was also very distracting, kind of worrying, and nothing happened in the end. On the other hand we have Typhoon warnings, which are actually much more predictable and quite useful.
My point, unless it's something like 90-95% accurate, I don't think it's useful at all. Mostly a distraction. When you live in a seismically active area, you just need to be prepared all the time.
Fair enough. I'm in Wellington and the risk profile here is a lot different to Japan. From a quick read of a BBC article, I see the "megaquake" you were warned about could have caused hundreds of thousands of deaths in the worst case scenario. Here in Wellington I've read estimates of up to thousands of deaths in the worst case scenario. But regardless, people here tend to be quite blasé about emergency supplies - perhaps if there was a heightened risk, people might actually check their grab-bags and get their supplies in order.
It wasn't a warning with a 70% accuracy though. It was saying it's 3x or 4x more likely than usual, which is still pretty unlikely. I don't think more information is ever a bad thing. Should countries have published covid statistics or hid them because it's very distracting, kind of worrying, and nothing really happened in the end? It was useful to know when cases were going up in your community so you can take precautions.
I took the megaquake warning as me realizing that I need to have at least 3 days of water stockpiled. I had 0. I don't really want to be a burden upon my neighbors in case of an earthquake, so it made sense just to store some water.
> What would governments/policy makers actually be able to do with any of this data? Not to denigrate it as a step forward, but I'm having trouble seeing much practical impact at all.
You can look at how beneficial hurricane forecasting has been in saving lives, which has been increasing in accuracy for longer lead times. It's very useful to know something bad is likely going to happen somewhere so you can move resources and evacuate people.
> So 14 true positives and 8 false positives, which means the positive predictive value is not great, less than 65%.
This actually seems huge to me unless we already are hitting close to 65%. I'm not sure how this wouldn't be a big deal compared to what I understand the status quo to be (unpredictable).
Of course, if the forecasts are just a couple minutes out at best, then that's way less useful. But at the very least an emergency alert could be sent out so people can get to safety which could help.
But if you, when you received a warning, knew that roughly a third of the time nothing would happen, would you still adjust your plans or leave?
And even if you trusted the warning, would you leave immediately? Considering that the warning was for the next week, ie the quake could come next minute or in six days.
Leave. If you live in such an area it would be a way of life. You probably have a spare toothbrush at someone else's house already. Unless you have a bunker or something that is considered safe.
Used to live near Japan, in a 5-story concrete building. We had so many earthquakes, the things were getting thrown off the shelves multiple times a year. There is nowhere to leave for a week. And there is a very slim chance anything would collapse, because the buildings are specifically built to withstand up to around 9 richter.
The sacrifice is loss of income from not working, losing job, or at least using up paid time off. There is also disruption to kids’ schedules and activities, as well as lodging and transportation costs to live somewhere else for a week.
If you could forecast earthquakes with principal component analysis and random forests don't you think someone would have done it by now? This paper is trash and will have zero impact.
> The outcome was a weekly forecast in which the AI successfully predicted 14 earthquakes within about 200 miles of where it estimated they would happen and at almost exactly the calculated strength. It missed one earthquake and gave eight false warnings
There were 600 entries into the competition, I wonder how chance played into this solution winning the contest.
I run 200 blogs, each of them predicting a moderate earthquake at a particular place but at different times. When one of them finally “hits”, I need monetization ideas.
Practical AI: 2025 7.1 Cascadia Subduction Zone Earthquake predicted with Large Seismic Model
Here’s the flim-flam I had ChatGPT generate for blog #78:
So, it turns out AI might be better at predicting earthquakes than we thought. A group of us have been working on a project using a massive AI model to analyze seismic data from the Cascadia Subduction Zone. And here’s the wild part: the model is saying there’s going to be a 7.1 magnitude quake in 2025. We didn’t believe it at first either, but after triple-checking everything and running more data, it keeps spitting out the same prediction. Only time can tell, so by all means check back here next year to see if we got pie on our faces. And maybe consider insurance. You know we wre.
ASIDE BOX:
Read “10 Best Earthquake Policies You Can’t Afford to Ignore (take it from a data scientist)”
small text: the lawyers say we must disclose that we are not insurance agents, but we may from time to time receive commissions from insurance agents.)
The model itself is built on a mix of historical seismic data and live feeds from monitoring stations scattered across the Pacific Northwest. It's not just looking at standard quake indicators, though—it's picking up on micro-signals, plate movements, and some weird patterns in ocean temperature shifts that we hadn’t considered before. It’s not perfect (nothing ever is), but this is the first time any of us have seen a prediction this specific and confident.
The prediction, released in a public report last week, has sent shockwaves—pun intended—through both the scientific community and public safety organizations. The AI's model incorporates everything from tectonic stress accumulation, plate motion, and fault line behavior to oceanic temperature variations, allowing it to anticipate tremors with remarkable precision. According to the developers, the LSM's ability to map out micro-movements in the Earth's crust is at least five times more sensitive than existing seismic monitoring systems, marking this prediction as a potential game-changer for earthquake preparedness across the Pacific coast.
No, the blog with the correct prediction would have 100% success. Counting all the blogs would be like counting the other 599 entrants into the contest.
Given the map of the predictions versus the actual locations of the earthquakes, I'm not prepared to celebrate the accomplishment here. It looks as if in the essential details, the prediction is just flat out wrong--it's consistently predicting earthquakes in the wrong basin.
Actually, digging in a little more, I'm even more suspicious. The article says
> The outcome was a weekly forecast in which the AI successfully predicted 14 earthquakes within about 200 miles of where it estimated they would happen and at almost exactly the calculated strength. It missed one earthquake and gave eight false warnings.
A "weekly forecast" isn't terribly descriptive, but it sounds to me like a prediction "Will an earthquake occur this week? If so, where and at what strength?" Given that it's 14 earthquakes over a 7 month period (i.e., about 30 weeks), that means you're looking mostly if not entirely at small, probably unnoticeable earthquakes. It also means that there's basically a coin flip of whether or not an earthquake will occur--and if you score it on the accuracy of predicting such, it comes out to 30% wrong (so the p-value, if I'm doing it right is 0.02, which I guess is significant, although if another commenter is right and this is the best of 600 competitive entrants, it should be expected that one would look this good).
Given that both the timing and the location accuracy look less than impressive, the next question is how good a job it did at predicting the magnitude. There's no details on the accuracy here, but given the location accuracy is hailed as impressive despite being clearly visually less than so, it wouldn't surprise me that the magnitude predictions are similarly garbage.
In short, this feels like merely continued evolution in the history of earthquake prediction techniques rather than a revolution, which is to say something that is loudly hailed as being a good start yet turns out to go absolutely nowhere.
They dumped a bunch of data on pca and random forests then only talked about when it happened to be right. Don't get your hopes up that this paper will go anywhere.
I wish the article would explain the criteria for winning the competition. If it's simply percentage of actual earthquakes that were predicted I could easily write an algorithm that would score 100% by always returning the value `true`.
For me, its quite clear from the article, that always returning `true` is not what they did:
"[the ai] predicted 14 earthquakes within about 200 miles of where it estimated they would happen and at almost exactly the calculated strength. It missed one earthquake and gave eight false warnings."
So its:
14 Positives,
8 False Positives and
1 False Negative
Precision was 64% and recall was 93%. In this context recall is a lot more meaningful (how many real earthquakes were predicted) than precision (how many predictions were earthquakes) as long as there aren’t too many false alarms.
> the AI successfully predicted 14 earthquakes within about 200 miles of where it estimated they would happen and at almost exactly the calculated strength. It missed one earthquake and gave eight false warnings.
It gave slightly more than 1 in 3 false alarms unfortunately.
This might go into hands of insurance companies who will decide to not insure areas where Earth Quake is likely to hit and could also do some more forecasting to avoid further regions.
But that's kind of ok? I mean, insurance is ultimately about amortizing the cost of unexpected adverse events. It's not a way of fobbing off the cost of expected damage onto someone else, even though many people intuitively think of insurance this way.
If it's able to predict them, then there must be some sort of pattern and it's using a self-created algorithm to predict them. Has it spit out the algorithm so that analysis and enhancements can be done?
Enhancements? ML usually beats feature engineering humans. See: chess, computer vision. So the enhancements might be more data, better data and better models.
However knowing why is probably if scientific interest and of interest to people deciding where to live.
This is amazing. A scientifically literate society would budget 100s of millions on huge sensor networks to collect what remaining data is needed to increase predictive power above 90%.
"The researchers said that their method had succeeded by following a relatively simple machine learning approach. The AI was given a set of statistical features based on the team’s knowledge of earthquake physics, then told to train itself on a five-year database of seismic recordings."
