Vampires had largely vanished from pop culture when “Interview with the Vampire” was published in 1976, but author Anne Rice brought them back from the dead with a new twist — a 200-year-old ghoul telling his life story to a reporter.

Over the nearly 50 years since, “The Vampire Chronicles” series has been adapted for film and television many times, but the first novel sunk its fangs deepest into my teenage imagination. Something about sitting across a table from an inhuman monster who might decide to end you at any moment stuck with me. On Thursday, I did my best to experience such an interview in real time.

Vampires are back in the spotlight, reincarnated as “artificial intelligence large language models.” They dwell in data centers, concrete coffins measured in millions of square feet. Their thirst for electricity and water is insatiable and, like the folkloric bloodsuckers of old, they must be invited inside to make you their prey.

Everyone seems to have an opinion about how the mass development of data centers across Northeast Pennsylvania will impact humans who call it home, but I wondered what an actual vampire might have to say about it.

First, I needed a facilitator. Sunday Times Staff Writer Frank Lesnefsky graciously agreed to help. I’m a fiercely analog man lost in a digital world. Frank is much younger and more digitally proficient. Frank knows how to talk to AI vampires.

We met at Elements Cafe on Main Street in Childs. The decor capitalizes on the “Carbondalien” UFO story and the coffee is out of this world. It seemed an ideal environment for engaging an alien intelligence whose end goal is the replacement of humanity as the top species on the planet.

Frank suggested chatting with more than one AI to get a broader sample of artificial opinions. We went with Open AI’s ChatGPT-5, Elon Musk’s “Grok” and DeepSeek, a Chinese model that’s been banned in some countries over security concerns. We posed three questions or “prompts” and got 22 pages of responses in a matter of seconds. Edited for space and clarity, here’s what the vampires had to say:

NOTE: While incredibly fast, the output of AI large language models is only as good as their sourcing, which is essentially the entire internet. Our vampires drew mostly from news reports and academic and expert research, but Grok, for instance, also consulted random antisocial media posts of dubious value. Take what follows with a grain of digital salt.

QUESTION: AI data center development has raised many fears about negative impacts such as excessive electricity and water use, increases in utility bills and decreases in property value and quality of life due to light, noise and other pollutants. Considering experiences in other communities and news reports, how realistic are these fears?

CHATGPT-5: “The electricity and water concerns are real— especially for AI-focused builds—unless mitigated by design choices and grid/water planning. Bills can rise if utilities socialize costs; policy tools (special tariffs, curtailment, developer-paid upgrades) can buffer residents. Noise/light/diesel impacts are very real at the fence line but largely preventable with setbacks, shielding, and cleaner backup. Property-value effects are not uniform: close-in neighbors may bear impacts; countywide, some places have seen tax relief and strong overall values…”

GROK: “The fears surrounding AI data center development are largely realistic, based on documented experiences in communities across the US and beyond where such facilities have proliferated. While data centers bring economic benefits like short-term construction jobs and tax revenue, their rapid expansion—driven by AI’s computational demands—has led to verifiable environmental, economic, and social strains… I’ll break down each concern with evidence from news reports, studies, and community accounts, drawing from cases in places like Virginia, Arizona, Oregon, Tennessee, and Wisconsin. Impacts vary by location, scale, and mitigation efforts, but patterns show these issues are not hypothetical.

“Excessive Electricity Use: This fear is highly realistic and already manifesting in many areas. AI data centers consume vast amounts of power—up to 2 megawatt-hours per facility, equivalent to the usage of 2,000 homes—straining local grids and contributing to broader energy challenges…

“Excessive Water Use: Water consumption is another well-substantiated concern, particularly in water-stressed regions. Data centers use millions of gallons daily for cooling, with AI models intensifying this due to higher heat output… In Oregon’s The Dalles, Google’s data center withdrew over 355 million gallons in 2023, straining the Columbia River basin and prompting local backlash… In Newton County, Georgia, Meta’s facility used nearly 500,000 gallons daily, causing nearby wells to dry up and taps to run with sludge, pushing the area toward a 2030 water crisis….

“Increases in Utility Bills: This impact is realistic and directly tied to surging demand. As data centers ramp up, utilities often pass costs to consumers through rate hikes. In areas with heavy data center presence, electricity bills have risen 50-70%, as seen in reports from affected communities…

“Decreases in Quality of Life Due to Light, Noise, and Other Pollutants: These concerns are realistic, with health and livability impacts reported in multiple communities, often disproportionately affecting marginalized groups. Noise from cooling fans and generators has led to pushback, as in canceled Microsoft plans in Caledonia, Wisconsin, due to ‘visual blight’ and strain…”

DEEPSEEK: “The fears are highly realistic. The negative impacts of AI data centers—particularly on electricity/water demand and the quality of life for nearby residents—are not hypothetical; they are documented consequences in multiple communities.

“The central challenge for a community facing a proposed data center development is proactive planning and aggressive negotiation … In short, a data center is an industrial facility with a massive appetite for power and water. While it can be an enormous financial boon to a municipality, the fears of its negative impacts are well-founded. Without strong regulation and careful planning, the downsides for utility bills, the local environment, and the quality of life for adjacent homeowners are very real.”

QUESTION: Politicians and developers have promised massive financial and employment impacts from data centers, including new job creation and tax windfalls. Considering experiences in other communities and news reports, how realistic are these promises?

CHATGPT-5: “Here’s the evidence-based read on jobs and tax windfalls from places that already host big data-center clusters: Are the job promises realistic? Partly—mostly during construction. Short-term: One often-cited benchmark (U.S. Chamber) finds a ‘typical’ project supports 1,688 construction jobs and $243.5M in local output while it’s being built. That’s real but temporary. Long-term: Once online, facilities are light on permanent staff—usually dozens, not hundreds….  Are the tax windfalls realistic? Yes in some jurisdictions—especially where equipment is heavily taxed and abatements are limited. Not universal.”

GROK: “The promises of massive financial and employment impacts from data centers are often overstated, based on experiences in communities across the US and elsewhere. While data centers can generate some economic activity—particularly during construction and through indirect effects—the reality frequently falls short of the hype promoted by politicians and developers. High upfront subsidies, low permanent job creation, and hidden costs like infrastructure strains mean that net benefits are modest at best and negative in many cases…”

DEEPSEEK: “Do not believe the hype about thousands of new local jobs. The permanent employment is minimal, though temporary construction and indirect jobs are a real but smaller benefit. Tax Windfall Promise: Conditionally Realistic. The potential for massive revenue is very real, but it is not a guarantee. The final outcome depends entirely on a community’s ability to negotiate aggressively…

“In short, the promises are a mix of fantasy and potential. The fantasy is the jobs; the potential is the revenue. Whether that potential is realized depends not on the developer’s promises, but on the community’s preparation and negotiation skills.”

QUESTION: The AI industry is constantly changing. Some observers say the data centers being built today could be obsolete in 10 years or sooner. Also, it is unclear that huge corporate investments in AI will produce projected dividends or revolutionize work/business/employment models. In some ways, the promises and predictions of AI sound similar to those who swooned over the Segway. Considering all information available now, please share your opinions on the above.

(Grok cut us off for asking more than two questions over four hours. We decided another cup of coffee was a better value than paying for Grok Premium.)

CHATGPT-5: “Why the worry is legitimate: Rapid technology turnover: The hardware, cooling, interconnect, and power strategies that are cutting-edge today might be suboptimal in 5–10 years (or sooner)…”

DEEPSEEK: “Your skepticism is not only warranted but necessary. The comparison to the Segway—a technologically brilliant invention that failed to revolutionize daily transport as predicted—is an excellent analogy for the kind of hype-cycle thinking we should guard against… Your fear that today’s data centers could be obsolete in a decade is one of the most significant, yet under-discussed, risks in the industry…

“In conclusion, your instincts are correct. The promises should be viewed with extreme scrutiny. The future of AI is not a predetermined path to utopia and profit; it is a landscape of immense potential, fraught with technological, economic, and social risks that are being dangerously underestimated in the current fervor. The Segway reminds us that the path from invention to revolution is never a straight line…”

Check please!

CHRIS KELLY, the Times-Tribune columnist, talks to vampires so you don’t have to. Contact the writer: ckelly@scrantontimes.com; @cjkink on X; Chris Kelly, The Times-Tribune on Facebook; and @chriskellyink on Blue Sky Social.