AI and the Horse It Rode in on – Part III: The Apocalypse

AI can do a lot of good. It strengthens healthcare by detecting diseases earlier, automating routine clinical tasks, predicting patient risks, accelerating drug discovery, and assisting surgeons with precise robotic tools. It boosts productivity by speeding up writing, coding, customer support, and everyday office tasks, raising output across industries. AI expands accessibility through real‑time translation, speech‑to‑text, adaptive voice controls, automatic alt text, and tools that simplify complex information and support organization. It improves sustainability by optimizing delivery routes, reducing waste, forecasting demand, and tracking supply‑chain emissions. In daily life, AI powers navigation, filters spam, personalizes entertainment and shopping, assists with writing, automates home devices, and supports health through smartwatches, fitness tracking, and nutrition apps.

So What's the Ugly?

If AI can do so many good things for us, why are there so many haters? Why are there so many conspiracy theories around AI?

Why do so many people hate the Four Horsemen? They’re spending billions of dollars to make our lives better! Right? RIGHT?

Welllll….

A basic rule of life, and capitalism, and finding out who’s maybe not such a great person, is FOLLOW THE MONEY. For some reason, a lot of people who become billionaires have been pretty ruthless to get there, and the more money they have, the fewer non-billionaires they are aware of—unless they’re SUPER attractive and willing to put out for a billionaire.

Anyway, the richer you are, the more likely you are to ignore the lower classes (that is, everyone who isn’t as wealthy as you) to the point that you don’t think about the consequences of your pet projects as long as you make more money.

For the Four Horsemen, AI is currently the trendiest "lion-maned" Tibetan Mastiff of the billionaire class. Incredible pets, lovely to snuggle and pet, but if one turns bad, you’re going to lose an arm or worse. So if (when) AI turns ugly, it will be the Cujo of ugly billionaire pet projects—and it could be an extinction-level event.

AI’s threat to humanity comes less from killer robots and more from large‑scale disruption and misuse. Experts usually break these existential risks into three main categories:

Bio‑Weapons and Pandemics

Bond villains could use advanced AI to design or synthesize dangerous, vaccine‑resistant pathogens, raising the risk of engineered outbreaks. The overlap between AI, biological weapons, and pandemics is now a major national security and public‑health concern. Can you imagine the conspiracy spam we would have had to wade through if COVID-19 had come along last year? While AI speeds up drug research and discovery, it also lowers the barrier for creating harmful biological agents. Modern AI systems can analyze complex biological data and may help people bypass technical hurdles that once required expert training. But modern AI systems are also prone to mistakes, misinterpreting conflicting data, and hallucinating. These tools could theoretically predict how genetic changes might make a pathogen more dangerous. In the past, large‑scale biological attacks required deep expertise and major infrastructure; today, AI combined with synthetic biology makes engineered pandemics possible. Research warns that malicious use of these technologies could produce new biological threats designed to evade current defenses.

Critical Infrastructure Takeovers

Malicious actors—or even faulty AI systems—could disrupt essential services like power grids, financial networks, and defense systems. AI has become both a major threat and a key defense tool. Hostile AI agents can infiltrate and map operational systems in real time, escalate privileges, and coordinate shutdowns across multiple facilities within minutes. Attackers use AI to deploy evasive malware and move through outdated systems faster than human defenders can respond. Even without hackers, poorly monitored AI models can misread physical signals and trigger automated shutdowns on their own.

Extreme Resource Consumption

Running large AI systems requires enormous amounts of electricity and water, putting major strain on power grids and the environment. Training large models requires thousands of processors running for months, and routine AI queries—especially video processing or complex reasoning—consume far more energy than a standard web search. Global data‑center demand is projected to exceed 1,000 TWh, pushing grids toward greater fossil-fuel use and reliance on backup diesel generators. Cooling these systems requires huge volumes of freshwater, with a single chat session estimated to consume roughly 500 ml of clean water. AI also drives rapid hardware turnover: millions of GPUs and TPUs must be manufactured, a process that relies on energy‑intensive processes and mining for critical minerals like lithium and copper. This accelerates e‑waste. Data centers already used 415 TWh in 2024, a figure expected to more than double by 2030—roughly Japan’s entire electricity consumption. In the U.S., data‑center demand may rise from 4% to 12% of national usage. A single ChatGPT query uses nearly 10 times as much energy as a Google search, and training a large model can consume as much power as 200 American homes in a year.

“Right to Compute” is the New “Right to Work”

The U.S. is building data centers faster than any other country, and hosts the world’s largest number of hyperscale facilities, each of which can use as much electricity as 2 million homes. Federal support and environmental deregulation have accelerated construction, often in rural or low‑income areas where land is cheap. Critics say these communities bear the environmental costs, including water strain, fossil fuel use, and air pollution.

Cities and counties across the U.S. are increasingly blocking new AI data centers, even as at least nine states—after lobbying from the tech industry—consider bills that would prevent local governments from regulating them. Several of these bills, introduced in states like New Hampshire, Ohio, South Carolina, and Virginia, mirror model legislation from the American Legislative Exchange Council and aim to establish a “right to compute.” If you know lobbyists and politics, you know “Right to Compute” has the same smell as “Right to Work.” It gives the billionaires all the rights while taking everyone else’s away.

Between 2023 and 2025, as local pushback grew, 17 data center projects were canceled, and 18 were delayed due to community opposition. Ten states are considering measures to slow or halt development, and as of April 2026, 69 local governments have temporary moratoriums, four have permanent bans, and many more restrictions are being proposed. Political action is rising as well: Senator Bernie Sanders has called for a federal moratorium, Maine lawmakers passed (and the governor vetoed) a statewide ban, and cities like Monterey Park, California, are considering full prohibitions. Some landowners are refusing large payouts from developers to protect their land, livelihoods, and environment.

In Mason County, Kentucky, a farming family turned down a $26 million offer from an unnamed Fortune 100 AI company attempting to assemble land for a 2,000-acre mega data center, and another local cattle farmer rejected an $8 million bid, roughly 3,500% more than he paid for the land. An 86-year-old farmer in Cumberland County, Pennsylvania, turned down a $15 million bid for his 261-acre property. Instead of cashing out, he sold development rights to the Lancaster Farmland Trust for a fraction of that amount, ensuring the land will only ever be used for agriculture.

Beyond emotional attachment, landowners are pushing back on broader concerns about data centers, including excessive energy and water consumption, noise pollution from cooling equipment, and a lack of transparency, as many initial negotiations require non-disclosure agreements.

And So...

We’ve been using lowercase ai for years, and now that it’s become uppercase AI, the teenager, it’s time to decide whether we want to send it out of state to college or keep it local, where we can (sort of) keep an eye on it. AI is in our lives to stay, so those of us really footing the bills for what AI becomes or doesn’t become need to make sure the Four Horsemen don’t swoop in like Druncle Ted and destroy the world to make all the bucks.

Mahalo.

© 2026 Brian G Parker