Data Centers in Brighton: Electricity, Part 2

This is a continuing discussion (part 4) on data centers possibly coming to Brighton, and part 2 of talking about electricity/power. You can find previous articles here:

If you read the last article on electricity, you will remember that when we talk about our homes, we talk about watts and kilowatts. Now we need to scale up to talk about megawatts, gigawatts, and terawatts.

1 terawatt = 1 trillion watts (the US produced 4,430 terawatts hours of power in 2025. In 2024, electricity accounted for 1.8% of GDP.

1 gigawatt = 1 billion watts

1 megawatt = 1 million watts (United Power’s biggest output is around 680 MW)

1 kilowatt = 1 thousand watts

1 watt = very small (could power an LED indicator light)

When we talk about data centers and electricity, we obviously need to talk about AI. Pretty uniformly across everybody we (Jordan and I) spoke to, we got an education on how we’re all using data centers currently for our phones, that we stream movies and tv, that we do internet searches, etc… and they’re right, however, that is well-established technology data centers use. That is not where the growth is. The vast majority of growth of data centers right now is for AI. Although we can’t say for sure, that is likely what would come to Brighton in part or fully, due to the massive growth in AI data centers.

I want to introduce a documentary to you that I recommend you watch. My brother texted me about it last night, as he just saw it on a plane. I had to do an online rental. It’s called, “The AI Doc: Or How I Became an Apocaloptimist.” I highly, highly recommend you watch it, then take a long walk and practice some meditation.

This documentary is interesting in what is said, what is not said, and alarming to me, the body language of the people interviewed. The basics that I got were:

They (the billionaires developing the technology) were also adamant that we (the U.S.) need to be “first,” before China or Russia, and that some sort of international agreements need to be drawn up, similar to how we deal with nuclear warheads. Think about this type of tech currently strictly being in the hands of billionaires and having no laws around it!  Anybody hear of the federal government working on this?  Me neither.

Breathe! We can’t “lose it” now! And watch the movie.

Locally, both 27J Schools and the City of Brighton are supporting the use of AI.  Brighton, for example, has 8 Flock spy cameras around the city (in addition to others owned by businesses). With these cameras, local police can track your car (color, license plates). They currently share use of the cameras with 1500+ other entities. Meanwhile the ACLU reported that even though our local police can’t view it, Flock can do a lot more tracking than the police can (see links at bottom of this page).  In addition, the city is now using “Brightly” an AI bot powered by Citibot on its website. This school year, 27J Schools implemented Intellisee, AI used on school cameras to spot activity Intellisee considers concerning or dangerous in some 27J schools.

Brighton’s AI on the city’s webpage

Is this what we really want for our community?

Ok, back to electricity! If you try to look up how much energy it takes to do an AI search, you will get all sorts of answers.  AI companies are very secretive about it, and people who study are estimating that some companies use 10x the amount of energy as others.  Per a July, 2026 United Power presentation document, old school Google searches used 0.3 kW of energy, while AI searches took 2.7 kW. I calculated that out on our home electricity use on a daily average across 4 years.  We could power our whole house for the same amount of power as 6-8 AI searches per day. How sustainable and realistic is this? Are we living for technology or humanity? What kind of environmental impact will this have? Well…..

A May, 2025 article by MIT Technology Review stated:

“In analyzing both public and proprietary data about data centers as a whole, as well as the specific needs of AI, the researchers came to a clear conclusion. Data centers in the US used somewhere around 200 terawatt-hours of electricity in 2024, roughly what it takes to power Thailand for a year. AI-specific servers in these data centers are estimated to have used between 53 and 76 terawatt-hours of electricity. On the high end, this is enough to power more than 7.2 million US homes for a year.

If we imagine the bulk of that was used for inference, it means enough electricity was used on AI in the US last year for every person on Earth to have exchanged more than 4,000 messages with chatbots. In reality, of course, average individual users aren’t responsible for all this power demand. Much of it is likely going toward startups and tech giants testing their models, power users exploring every new feature, and energy-heavy tasks like generating videos or avatars.

By 2028, the researchers estimate, the power going to AI-specific purposes will rise to between 165 and 326 terawatt-hours per year. That’s more than all electricity currently used by US data centers for all purposes; it’s enough to power 22% of US households each year. That could generate the same emissions as driving over 300 billion miles—over 1,600 round trips to the sun from Earth.

The researchers were clear that adoption of AI and the accelerated server technologies that power it has been the primary force causing electricity demand from data centers to skyrocket after remaining stagnant for over a decade. Between 2024 and 2028, the share of US electricity going to data centers may triple, from its current 4.4% to 12%.”

I had to read that a couple times. By 2028!  That’s a little over a year away!

So, what about data centers coming to Brighton?

We don’t know any details. None, zero. We can, however, look at data centers going in around us in other communities.

Data centers are described in 4 sizes:

Micro/small: Occupy under 5,000 square feet and draw under 100 kW to 1 MW of power, handling local or low-latency processing. 

Medium: Ranging from 5,000 to 20,000 square feet, these typically require 1 MW to 20 MW of power and host thousands of servers. 

Large: 20 MW to 100 MW 

Hyperscale: Cover 100,000 to 10 million square feet, with 100+ MW or even gigawatts of continuous power 

The hyperscale data center, QTS, being built in Aurora is reported to be built to use 160-177 MW, taking up 60-80 acres. We are finding a lot of reports, however, that data centers will start at one size and expand, sometimes without permission, and using generators to supplement their power (and denying that’s what they’re doing while poisoning the local air). Of interest, Aurora’s QTS data center has 40 diesel generators and they are requesting 98 more, which would give them 345 MW of power backup for a 160-177 MW facility . Via calculations I found online, if they were all used at once, this would burn over 24,000+ gallons of diesel per hour. Aurora is thankfully not on United Power’s service delivery area. For comparison, United Power’s entire peak power delivery is about 680 MW for the entire grid.

In Elyria-Swansea, on the north side of Denver, a CoreSite data center is being built, and Denver has now put a 1 year data center moratorium in place due to the backlash. This facility is reported to use 75 MWof power with 14 diesel backup generators. A July 21, 2026 Denverite article (worth reading, as there is a video of the sound from the data center) stated:

“It’s now clear that the community can expect irregular floods of generator noise from the facility. On Monday afternoon, a low-frequency hum — similar to an overhead jet or a massive air-conditioning unit — emanated from the site, masking the noise of construction at a long-planned senior living facility across the street.

It could also be heard in a nearby park and from a sidewalk outside Tepeyac Community Health Center, a low-income health clinic near the data center.

‘It is impossible not to notice the testing runs,’ said Dr. Pamela Valenza, the clinic’s chief health officer, who first noticed the noise last week. ‘As I got out of my car, the sound coming from the generators was deafening.’”

In Windsor in Weld County, they are planning an even bigger data center than Aurora, Global AI.  An August 4, 2026 article from Colorado Public Radio indicated that the county has asked the data center to stop building three times due to permit issues, but the work has continued.  The article goes on to describe that CPR found discrepancies in energy and water permits, and supply.  What happens when we have allowed billionaires and trillionaires with tons of investor money to so monopolize the country that we lack the money at the local level to stop them?  Sadly, this is not a rhetorical question.

As with most newer technology, developers pose a lot of hype to get investors and money to spend. You see that everywhere, the inevitability of AI. The problem is that the amount of energy it takes to run AI (remember 2.7 kw for an AI search and 6-8 AI searches to power my house for a day), helps you see the problem. AI was free at first, but now they’re starting to charge a little bit. In the long run, they’re going to have to recuperate their electric bills AND make a profit. Some estimates,for heavier home users are $200/month with custom company use being up to $500,000/year in a subscription style cost.  For now, AI companies are keeping those costs lower.

On July 2, 2026, Forbes said that many for many jobs, people are cheaper than AI, and the spending by AI companies to keep up the illusion is starting to fail.  OpenAI is offering to sell 5% of its company to the government.  Some are questioning if this is the start of a government bailout.

With all of that information, you are starting to see the problem.  These companies have GOBS of investor money that can beat out the local government’s ability to cope with them.  They need a lot of energy, and are polluting the air all over the country.  I have concerns about the logic going on around our ability to get power to our local grid in a fashion that makes sense to us and our local grid.  Sadly, this is going to go into a third article about power/electricity.  There’s so much to cover.  More to come.