AI Images For Weather Disasters Now Prevailant, Risks Misinformation And Misrepresenting Victims' Needs

An AI image of Nashville's recent
ice storm. It doesn't look like this
and this image has been used to
"show" what it's like in the ice
storm south. Fake images like
this can foster mistrust in 
disaster news, and potentially
direct aid to the wrong places. 

 Each big disaster has just an iconic image or two, ones that capture the breadth and emotion and tragedy of the event. 

The big Lost Angeles wildfires in January, 2025 has the vintage blue Volkswagen bus gleaming seemingly unscathed in a large, completely leveled, blackened neighborhood. Or the resort towns in western North Carolina wiped out by Hurricane Helene. Or, here in Vermont, downtown Montpelier inundated by the summer floods of 2023. 

Now, we have that horrible ice storm in the south. I've seen so many dramatic photos and videos of collapsed trees, tangled, destroyed electrical systems, damaged houses and highways that resemble miles long skating rinks, 

But the photos that arguably circulating the most on social media do not at all capture the real tragedy, scariness and power of the ice storm. 

The photos I see the most are AI generated. Definitely not the real deal. 

One account on Facebook called Meanwhile in Delco posts a lot of AI generated images of bad weather, and really went to town with the southern Ice Storm. Sometimes images from this account and others get a community note explaining that it's a fake image. 

But most of the time, when I see the AI images the comments are all, "OMG the destruction!" and "Pray for Nashville."

I'm 100 percent in favor of praying for Nashville and anybody else affected by the winter storm. However, I don't think AI helps. 

One person posted this genuinely thoughtful comment under one of Delco's  AI-generated Nashville ice storm images. She wrote: "For those saying this is AI... Does it really matter since this is scary close to what Nashville and the surrounding areas are surviving through the week? 

My answer is a respectful yes. It does matter. 

DANGERS OF AI PICS

Fake images breed distrust.  If it gets hard to tell which images of a disaster are fake and which are real, you start to lose trust in everything. Some people might wonder if the whole ice storm disaster is fake. AI, used incorrectly, can hurt, rather than help recovery efforts after disasters.

I'm picking on Meanwhile in Delco, but that Facebook account holder doesn't' seem intent on causing harm. The text parts of their posts seem accurate when they discuss power outages, recovery efforts or storms elsewhere. So overall, they're probably doing a little more good than harm. Or at least they're not adding much to the chaos. I've seen a LOT worse on Facebook.

But other people on social media can use AI to be misleading. Here's a theoretical: The winter storm was huge and caused misery and danger really from New Mexico to Maine, right?

I can envision efforts to use AI to exaggerate the effects of a particular storm in one area, while saying other areas that were hit hard are fine. For example, somebody could post AI images of New Orleans just obliterated by an ice storm, even though they're fine, while posting AI images of Nashville looking just ducky.  

Imagine this was being done by a crook who wants to use the sympathy of social media viewers to steer donations toward New Orleans (really the crook's pocket) and away from Nashville, which really needs the help. 

I know in the real world under my scenario, Nashville would still be getting help. But a little bit of that potential aid could get diverted from Nashville, or other areas of the South that were blasted by the ice storm. And the crook with the AI account would pocket the money.

My dream is to require all AI images to clearly state they are in fact AI.  I'd like to see that rule in social media terms and conditions. But, their algorithms are so bad that such a rule would probably ban real, unadulterated photos, or legitimate AI art that is not pretending to be news.

So yea, a solution to this isn't easy. 

Don't get me wrong. There's a place for AI on social media and elsewhere. It can be entertaining, even illuminating at times. 

But if there's some weather event or disaster somewhere, I, and I think most people, want to know what really happened. Not just some AI representation that at best is a guess, and at worst, a scam. 

 

Russia's Northeast Coast Is Buried In Unbelievably Deep Snow

Digging out from incredible snows in Kamchatka, a
region on the Pacific Ocean in Northeast Russia 

I saw a lot of images on social media showing ridiculous amounts of snow in Kamchatka, a  region o Russia on the Pacific Coast.

Those images looked so surreal that I thought it was all just AI slop.

Turns out at least a few of those images are the real deal, even if many others were fake.

Kamchatka, a city and region near Russia's Northeast Coast, has had storm after storm since December. The snow depth on the ground is at least as deep as a normal person is tall. 

According to the Moscow Times:

"In December, Petropavlovsk-Kamchastsky saw triple the monthly average of snowfall, while January 1-16 saw around 150% more than the monthly average for the period. The average height of snow in the city reached 170 centimeters (5.5 feet) with drifts exceeding 2.5 meters (8 feet) in certain neighborhoods.

'The last time we saw something like this was over 50 years ago, in the early 1970,' said Vera Polyakov, head of Kamchatka's Hydrometeorology Center. 'These conditions are exceptionally rare as far as modern observations go.'"

Reuters reports that close to 33 feet of snow has fallen since early December in the region around Kamchatka.

Schools have been closed and public transport is shut down. 

In the past couple of days, temperatures have risen to just above freezing. That is causing huge amounts of snow to slide off roofs, which is dangerous for anyone standing nearby. A 63-year-old man died when  he was struck by a massive snowfall off of a nearby roof. 

Enormous, falling icicles are another danger. 

More snow and wind is forecast for Kamchatka today through next Monday with temperatures hovering near freezing. 

The Kamchatka region is on Russia's northeastern coast, across the Bering Sea.  The area has a population of roughly 310,00 people.

Storms similar to nor'easters in New England and southeastern Canada regularly sweep through the region. Kamchatka often sees big snowstorms. But they've never seen anything like this in many decades, at least.

If you think you've heard of Kamchatka before, it might be from this summer, when a powerful offshore earthquake rattled the region and generated a damaging tsunami that slammed into the coast. 

Video:

A lot of videos I've seen of Kamchatka's snow appear to be at least partly AI. I've never seen so much AI slop online connected to one incident. There were several "photos" of snow drifts reaching to the 10th or higher floors of apartment blocks. Didn't happen. Welcome to the grim new world in which everything is fake, and nothing can be trusted, I guess.

But I found one video that I THINK is genuine. Part of it contains images of people gleefully jumping from upper story windows in apartment blocks and floofing into deep snow, Yes, I know "floofing" isn't a real word, but you get the picture. 

  

Click on this link to view, or if you see the image below click on that. 

 

Later in the day, the New York Post put up this even better video of the incredible snow in Kamchatka. Again, click on this link to view or if you see the image below, click on that. 

Will AI Fight Climate Change Or Make It Much Worse? Hot Debate On The Topic Now

AI is obviously the Next Big Thing.

Illustration from Appen 

Like it or not, it's invading all aspects of our lives. And of course, AI will have an enormous effect on other Big Things. Like climate change.  

Whether or not AI is good or bad, there was certainly a lot of interest in it at the recently concluded COP30 sessions.

That's the big annual UN shindig where countries from around the world try to come up with agreements on how to fix climate change. 

Usually, including this year, they didn't make much progress. And, the United States sat this one out all because our Glorious Leader says climate change is a hoax. 

But back to AI and climate. 

The Associated Press said they counted at least 24 sessions related to AI during the conference. 

They included using AI to help neighboring cities share energy, and a ceremony for the first AI for Climate Action Award for a project on water scarcity and climate variability in Laos.

All the tech bros at the converse are telling anyone that will listen that AI will solve climate change. Or at least make it less bad.  Or maybe help us deal with it better. They give examples. For instance, it can increase the efficiency of electric grids. 

AI can help farmers more accurately predict whether to save their crops. It can design buildings, roads and other infrastructure to withstand ever-increasing weather extremes. 

AI can also directly fight climate change. It's being used to detect methane leaks. Methane is an even more potent greenhouse gas than carbon dioxide. 

As NPR reported back in January, 2024 that methane leaks were seldom found and monitored. Now, scientists and companies use AI daily to find methane emissions by by interpreting enormous amounts of satellite images.   

That data is now used by the United Nations to verify the accuracy of corporation and government reports on methane emissions. 

Forest managers use AI to detect forest fires before they get out of hand. The technology can also ensure conditions are right for controlled burns to prevent future huge forest fires 

These are just some examples of how AI could save the world from a climate catastrophe, goes the theory. Or something like that. 

OR IS AI KILLING US 

It takes a TON of energy to run the data centers that keep AI afloat. So maybe AI will just make climate change even worse.

Per the Associated Press: 

"Climate groups, however, are sounding the alarm about AI's growing environmental impact, with its surging needs for electricity and water for powering searches and data centers They say an AI boom without guardrails will only push the world farther off track from goals set by 2015 Paris Agreement to slow global warming."

We'll probably blow past that 1.5 degree Celsius goal with or without sincere climate action. But AI might make us go far beyond that goal.

AI relies on data centers that use tremendous amounts of electricity, Data centers accounted for 1.5 percent of the world's electricity consumption in 2024. Since 2017, electricity consumption in data centers has grown by 12 percent, more than four times faster than the total rate of electricity consumption. 

Since a pretty big share of electricity is still generated via fossil fuel generation, data centers contribute to climate change. 

According to cornell.edu:

"Cornell researchers have used advanced data analytics - and, naturally, some AI, too - to create a state-by-state look at that environmental impact. The team found that, by 2030, the current rate of AI growth would annually put 24 to 44 million metric tons of carbon dioxide into the atmosphere, the emissions equivalent of adding 5 to 10 million cars to U.S. roadways. 

It would also drain 731 to 1,125 million cubic meters of water per year - equal to the annual household water usage of 6 to 10 million Americans. The cumulative effect would put the AI industry's net-zero emissions targets out of reach."

Sounds really bad. But then again, maybe not that bad?

Says cornell.edu again:

"On the upside, the study also outlines an actionable roadmap that would use smart siting, faster grid decarbonization and operational efficiency to cut these impacts by approximately 73% (carbon dioxide) and 86% (water) compared with worst-case scenarios."

So, the bottom line is AI is maybe one of the biggest new thing of all big new things we've ever had. AI could let us swim or make us sink. 

Only the future will tell us. And the future might come sooner than we think. 

Hurricane Melissa Is the First AI Generated Fake Video Disaster

A screenshot of a fake AI-generated view of Hurricane
Melissa from above. Hurricanes don't even look like
this in real life. Hurricane Melissa inspired a 
lot of fake AI videos, confusing and in some
cases endangering the public. 

 A lot of the dramatic videos you saw of Hurricane Melissa in the Caribbean were probably AI fakes. 

True, there are many, many, videos that show the true tragedy of what happened on Jamaica in that terrible storm. Those videos shock us, upset us, make us want to help. And they get ton of views. 

There's also a lot of fake AI Hurricane Melissa videos going around. Some of them really look like the real thing. 

But if you look closely, they were generated by whatever AI service the maker used. The motivation? Make a pile of money out of somebody else's tragedy. 

 Though AI videos have exploited disasters and other big events  for awhile now, Melissa
a is arguably the first calamity to get the full AI treatment on social media. 

Per PBS News:

"Experts noted that Melissa is the first big natural disaster since OpenAI launched the latest version of its video generation tool Sora last month. 

'Now, with the rise of easily accessible and powerful tools like Sora, it has become even easier for bad actors to create and distribute highly convincing synthetic videos,' said Sofia Rubinson, a senior editor at NewsGuard, which analyzes online misinformation. 

I supposed the AI videos of Melissa are less harmful than political deep fakes, ones that make leaders really appear to be something they're not, to sway the public based on big lies, to create a political environment detrimental to everyone except a cadre of billionaires and their brown nosing fans.  

Disaster AI videos are just clickbait. Sure, they're scammy. But they're easy and cheap to produce, and if whoever makes any of these gets lucky, they can get hundreds of thousands or even millions of views. A really lucky strike with a fake disaster AI video can rake in a few thousand dollars.  

These fake AI videos can spread heartache, and interfere with storm recovery.  

As CNET reports;

"In times of crisis, like a dangerous and imminent natural disaster these fake videos can create confusion panic and distraction at a time when accuracy can be life-saving."

One of the AI videos going round depicted the main airport in the Jamaican capitol of Kingston completely destroyed. If taken at face value, that video could have affected or delayed relief supplies.

In fact, the airport in Kingston was only lightly damaged. It opened for both relief and commercial flights last Thursday, October 30, just a couple days after the hurricane. 

"I am in so many WhatsApp groups and I see all of these videos coming. Many them are fake....And so we urge you to please listen to the official channels" said Dana Morris Dixon, Jamaica's Education Minister. 

AI videos can also falsely minimize the severity of a disaster. As Forbes noted, one widely circulated Melissa AI video shows Jamaicans partying, boating and swimming, which made the hurricane appear much less dangerous than it really was.  

It's trickier when you're in a disaster zone and trying to find solid information to keep  you safe. Often people in the middle of a disaster look to social media for guidance. An AI video made by someone who is unfamiliar with the area affected and how to respond to disasters can easily lead storm victims into danger, rather than helping them.  

The AI videos are confusing the public so much that people are writing off legitimate videos as AI. One video shows the dramatic views inside the eye of Hurricane Melissa taken from aboard a hurricane hunter plane. The video is legit, but it's so incredible, people tried to flag it as AI.

But yet another video purportedly taken above the eye of Melissa portrays it as sort of a wormhole, with clouds swirling around the eye, and pouring into the "hole."  If you understand meteorology, you would catch the problem right away. 

A hurricane's eye is the storm's "chimney," and the clouds around it are rising, not sinking. 

Though it's still pretty possible to distinguish a real video from AI nonsense, it's getting harder.  

Per PBS News:

"In the past, people could often identify fakes through telltale signs like unnatural motion, distorted text, or missing fingers. But as these systems improve, many of those flaws are disappearing, making it increasingly difficult for the average viewer to distinguish AI-generated content from authentic footage.'"

It's kind of annoying that nowadays, you have to look at everything with a super critical eye to distinguish fact from fantasy. 

But it's still possible. With AI video, as always, check as best you can to determine whether the video is from a reliable source. Check the account's history on social media. If the past is pretty clickbait-y, you are probably looking at more AI clickbait in the video you're watching. 

You can sometimes see a Sora watermark, which shows it is AI.  Apps can get rid of the watermark, but often smudge is left behind where you'd expect a watermark, so that can act as a clue. 

Although scammers are getting better at producing videos, look for strangely shaped objects or blurred commercial signs in the background.  Sometimes one object in the video has odd lighting, is strangely blurred or doesn't seem to have quite the right shape. 

If you still can't tell whether it's real or fake, welcome to the club. I have a good eye for this type of thing and sometimes I'm still not sure whether what I'm seeing is real or AI. In those cases, let's play it safe and don't share the video on social media. 

Credible sources usually end up verifying legit videos. The one you're fascinated with will either be proven true and debunked. Once that happens, then you can share it if you feel safe about it.

Meanwhile our social media feeds are getting more and more clogged with annoying fake AI videos posing a news. 

It's just another example of how we just can't have nice things. 

 

 

Will AI Make Climate Change Worse?

A coal-fired electricity generating plant in Pennsylvania'
AI is greatly increasing electrical demand, meaning
decommissioning fossil fuel run plants like this
are being postponed, harming the 
battle against climate change. 

 It turns out it takes TONS of electricity to generate AI. 

So much so that electricity demand in the United States, which had been on the stable side in recent years, is set to soar.  And that might screw up the battle against climate change. 

The amount of electricity needed for data centers will more than double between now and 2035. The increase is the equivalent to the amount of electricity the entire state of California used in 2022.

According to Environment America Research Policy Center: 

"Data centers currently make up 3 percent of America's energy demand but that is expected to rise to 8 percent by 2030.

The increase in data center electricity is primarily due to the significant data processing needs of Artificial Intelligence. As AI development accelerates, the need for data processing grows, and companies have to build more and larger data centers to meet those needs - and those data centers use a lot of electricity."

Moving toward EVs and other conversions that supposedly move away from fossil fuel is of course a good thing if you don't want to worsen climate change. 

But if demand for electricity skyrockets, the worry is some of that needed electrical generation won't be clean energy.  

Until now, electricity use hasn't grown much because of efforts to improve energy efficiency. As Environment America points out, the cleanest energy is the energy that never gets used in the first place.

They go on:

"When electricity demand is lower or at least stable and predictable, it is much easier to meet that demand with clean energy like wind and solar. A rapid and unexpected surge in electricity demand is the last thing we need, particularly when energy demand is already expected to grow due to electrification of buildings and transportation."

Environment America notes that utilities have extended the life span of coal plants and proposed new methane gas plants to meet the expected growth in demand from all these data centers. 

Each AI query a person does uses up a lot of juice. As NPR reported, one query to ChatGPT uses about as much electricity as what could power one light bulb for 20 minutes.  One ChatGPT query needs about ten times the amount of electricity as a Google search query.

So imagine what happens when millions of people are poking around Chat GPT or other sources of AI help. 

And as NPR tells us, a majority of that energy needed for AI still comes from the burning of fossil fuels like coal and gas.

The Washington Post tells us that soaring power consumption is delaying coal plant closures in Kansas, Nebraska, Wisconsin and South Carolina. 

Not only does this spiraling demand threaten to aggravate climate change, it also is making regulators worry that residential ratepayers will be stuck paying for the upgrades to the electrical grid

 Then there's this question. Like almost every Next Big Thing I've ever seen, there tends to be an over-enthusiasm bubble which might hit here. 

Will tech companies overbuild data centers only to see demand for AI fall short, maybe because of government regulation,  just too much capacity, or maybe the Next, Next Big Thing will come along after AI to wow the masses. 

If that happens, will we have built climate-damaging fossil fuel generating plants for no good reason?  And how much unnecessary emissions would that add to the world? 

I guess as the world gets more complicated, so does the battle against climate change. 

Google's AI Weather Forecasting Model Might Improve Forecasts

Old fart that I am, I tend toward skepticism with AI, or to spell it out, Artificial Intelligence.

Google's GraphCast AI weather forecasting model
performs better than current weather models,
according to studies. 

I know that's where we're going and there's certain benefits to the technology, but don't get me started on the downsides. 

However, if I must focus on the positive,  Google has developed some AI technology consisting of a weather forecasting model that experts say could vastly improve predictions. 

For years now,  it seems like every time there's a storm, you hear about the American model and the European model disagreeing on the track and intensity of expected bad weather. Both are computer models of course, with the European model arguably the better of the two.

According to studies, the AI weather forecasting model from Google is a big improvement over these old school models. I guess all roads lead to Google these days. 

According to the Washington Post: 

"Google DeepMind's AI model, named "GraphCast," was trained on nearly 40 years of historical data and can make a 10-day forecast at six-hour intervals for locations spread around the globe in less than a minute on a computer the size of a small box. 

It takes a traditional model an hour or more on a supercomputer the size of a school bus to accomplish the same feat. GraphCast as more accurate than the European model on more than 90 percent of the weather variables evaluated."

GraphCast is actually one of several AI weather forecasting models under development. Government agencies world wide are looking into AI models because they are fast, efficient and can save money. 

As the Washington Post explains, current models like the American and European make forecasts based on really complex mathematical equations that would make anybody's head spin. These models are expensive because they require tons and TONS of computing power.

The AI models work differently. They suck up pretty much all the historical weather data available, then generate forecasts by looking at current conditions and figuring out what will happen next based on that historical data.

In a sense, AI weather forecasting models are old school.  In addition to relying on computer models, many human meteorologists often compare historical weather events to what's going on now to help them predict what the weather will be like tomorrow. 

One question about AI and weather forecasting is its ability to forecast extreme events.  There are not many super wild weather events from the past from which AI can "learn" about and then project into the future.  

This is especially true in our age of climate change, considering that we're seeing more and more record storms, record temperatures and weird storm behavior. 

However, GraphCast performed well in these event. Says the Washington Post: 

"GraphCast reduced cyclone forecast track errors by around 10 to 15 miles at lead time of two to four days, improved forecasts of water vapor associated with atmospheric rivers by 10 to 25 percent, and provided more precise forecasts of extreme heat and cold five to 10 days ahead of time."

However, questions remain as to whether GraphCast and other AI models can accurately forecast small scale weather trouble, like severe thunderstorms, tornadoes and localized flash floods. We also don't know how well AI does with larger storms that produce heavy precipitation in one area, but light rain or snowfall just miles away. 

That's particularly important here in Vermont. The most dangerous weather in the Green Mountain State include flash floods roaring out of the mountains;  downslope winds during winter storms that can be destructive in one town and no big deal the next; and nor'easters that can dump a couple feet of snow in say, Ludlow, but just a few inches up the road in Rutland.  

The Washington Post also cautions that AI isn't going to replace traditional computer models, or human meteorologists for that matter. At least not right away. Traditional models offer the initial conditions that AI uses to project the future. Also, the old computer models are still necessary to "teach" AI how to interpret historical weather. 

Also, it'll take real-life meteorologists to transmit what AI is telling us what will happen into the Plain English we need to know whether to take an umbrella with us tomorrow.  

 Some of the headlines are already overblown on the idea of AI "replacing" meteorologists. The headline about GraphCast Tuesday in Gizmodo was "Google's New AI Weatherman Will Leave Forecasters In The Dust."

Read the actual Gizmodo article and you discover, not quite. As that article states, "The study authors say their work is meant to work alongside the standard systems meteorologists rely on."

So AI isn't going to chase away your favorite meteorologists like Al Roker, Jim Cantore, Ginger Zee or local Vermont favorites like Gary Sadowski, Tyler Jankowski or Mark Breen.   

If it all works out, AI will be another tool these meteorologists use to give us their most accurate forecasts.  I'd hate to see our forecasts just coming from some sterile machine.  I need the human touch when I'm being told how much snow I'm going to have to shovel from my driveway. 

I just hope the bean counters at major media outlets don't think they can fully replace humans. Because they can't.