top of page

BISBEE WIRE/Special- info you need about the data center,county application, questions/ What AI is? Election: Regime coming to the US before Iran

3 hours ago
14 min read

Hello,


I began this for my own education. Clearly got out of hand as one thing led to another. Hopefully you will find this as illuminating as I have. Artificial Intelligence (AI) use is quickly growing in most every aspect of our lives. There clearly are problems with its use, but there are also huge benefits to be gained. We, as a society, are just beginning to understand and grapple with the ramifications of this relatively new technology.

   The most pressing problem, illustrated in the elections, is an awakening that very few people are controlling the development, use and sale. And the trump administration's SEC will let people give them money for their stock (going public) when there clearly have been dangerous documented problems with 'AI escaping'. That should be stopped. 

  It is my belief that these huge mega companies are the largest thieves in history. They have stolen, and continue to steal, from me, you, every citizen that lives in the US and most countries. They then repackage our information and sell it. They have stolen information indiscriminately from companies, books, scientists, health records, government records, military plans, and wherever there is an electronic trace. There is no source they will not ripoff.

   The remedy is that they should be made to pay billions in reparations, with the money used to aid poor and working people. And they should no longer be private companies. At the very least the government should own a 51% share of each of these companies. I would prefer outright ownership by the people of the US.  

Maybe the election will help bring some oversight and sanity to the overt corruption of corporations ignoring or purchasing the rules under which they are supposed to be regulated. 


fred

   ...............​ELECTION COMMENTS................   

There are very few events that publicly 'take the pulse' of a majority of citizens throughout our country. Elections do that.  Consequently it is clear that this election was an emphatic rejection of much of trumpism, a repudiation of him personally and what his bootlikkers have done and want to do; the war, the policies, the cruelty, the personal crude epithets, the corruption, the absent republican senate and house, the ICE murders and actions, the glorification of wealth, and much more. 

   But if I hear blue wave or blue tsunami one more time I'm gonna puke. This was, in fact, serious red bloodshed. It was done by a majority of people across the country angry, dissatisfied and hurting economically. 

   Regime Change is Coming to the US before Iran, is what my friend Mike Thornton says and I agree. 

   Dems ran on many familiar issues; corruption, cost of goods, lack of healthcare, anti-trump. But it wasn't a 'program' in the sense of a group of thought-out policies underpinning legislative actions. Hopefully some coherency will emerge. 

   With rare exceptions, politicians do not lead. They put their finger in the air to see which way the wind is blowing and bend that way. If a big donor is in that direction, all the better. The task is to make sure the wind is strong enough to bend them in best way. That is why it is imperative to let them know what is your opinion in whatever forum you can-what should be done and when. 

   In Arizona, dems will control both the house and senate in addition to having each of the major administrative posts. It is the first time since 1966. It could be very positive. We'll see. 

   In DC, there may be a battle for the senate leadership with newly elected and established senators wanting to see a change. If they do a quick vote it may be Chuck Schumer, if there is a few weeks delay with some calling for a vigorous discussion, we may be looking at Cris Van Holland. We'll see.

   On a side note...there are jobs, jobs, jobs in Phoenix. With dems in power-they  set the operational budgets, chamber rules, and partisan hiring structures-If you or anyone you know is interested in serving in government, now is a good time to check it out. New politicos need legislative staffing, administrative jobs within agencies.   

I


       ..............COCHISE COUNTY................    

 

THE PROPOSED DATA CENTER NEAR THE WILCOX PLAYA

Like many I'm trying to understand what the proposed data center in Cochise County actually will do. So I've done some digging and some self-education.  I've included information from the application to county , various newsletters and legacy press articles, and, ironically, Claude the Anthropic AI assistant. Also Ali Morse, a tenacious Portal resident that tracks Cochise County Supervisor decisions, has produced research, background info, and history that is valuable for understanding this application.


Resources: 

Application: Go here for the actual special use application to the county turned in on Oct 30. There also is a second application HERE with modifications to the first one and turned in on November 2, after the scheduling deadline. It has 17 pages that have been revised. Apparently this is allowable according to the Ms. McLachlan of the county development services. 


This is a citizen website created by Patsy Nyberg, with a wealth of info

  https://www.stopcochisedatacenter.com/


This is the testimony of two career industrial hygienists in Michigan,  https://youtu.be/uioLCgXECaM?t=4908


A clear explanation of a data center/AI was given at a Marana meeting by Colin Mellar, go here


Data center: Is the term for any facility housing computing, storage, and networking equipment, along with the power, cooling, and security systems that keep it running. (see below for the difference in 'server farms' and 'data center'.)

Who: After some major obfuscation, a whistle blower within Sulphur Springs Electric Coop  revealed that the CEO, in an address to company workers, said that SpaceX was the corporation behind the application to the county. Immigrant Elon Musk controls about 80 % of the total voting power of SpaceX.    The application and press releases came from Tombstone Resources LLC, a company formed in July of this year, with the address of a global law firm Eversheds Sutherland in Atlanta GA.  SpaceX was not mentioned in any of the releases by the company until it was announced at the public meeting that it was SpaceX wanting the application. This is typical of large technology companies to obfuscate until they are found out. 

Proposed Size: 747 acres. The proposed development includes up to nine computing buildings, 40' tall, constructed in four phases. It also includes battery energy storage, electrical, cooling, and supporting infrastructure. According to the application it would use minimal water because it would be a 'closed loop' cooling system. 

Location:  748 acres adjacent to the Apache Generating Station, about 13 miles west of Wilcox, along Hwy 191, on the edge of the Wilcox Playa near the town of Cochise AZ. Wilcox Playa is a 40 sq mile dry lake bed that, in winter, is a roosting area for thousands of Sandhill Cranes. It is about 40 miles from Whitewater Draw another roosting area for the Cranes.

Land Acquisition: AEPCO will sell 400 acres to-ummm I'm not sure because it hasn't become public but probably SpaceX I would guess-and 348 acres will be sold by Southline Transmission to SpaceX.(Southline Transmission is owned by Grid United via its wholly-owned subsidiary GU Southline LLC with founding investor Hunt Transmission Services, a subsidiary of privately owned Hunt Power/Hunt Energy. Southline is co-developing the Southline Transmission Project that includes a new transmission line between the Afton substation west of Las Cruces, New Mexico, and the Apache substation in Willcox, Arizona.)

Apache Generating Station: The plant adjacent to the proposed data center will produce the electricity to power the center. Currently it produces about 1.3 gigawatts with a mix of natural gas, 90%, and coal, 10%, in their two operating units(with 5 standby units). They also have solar operations on 1300 acres near Apache. It is unknown whether they are selling any land that solar now occupies. They are ranked #19 out of 180 power plants in Arizona. The station was due to phase out coal use to attain 60% use of renewables by 2031, however the trump administration loosened the regs on coal and this summer the Dept of Energy awarded $21 million for upgrades to their coal fired steam generators. Another $32.5 million from an unidentified source non-governmental source, will be used for the upgrade.

   (Gigawatt definition: The application states that the first of four phases would need 1.5 Gigawatts, and an estimated electrical demand of 6 gigawatts at full build-out. Watts are the standard unit of power, and a gigawatt gigawatt (GW)  is a much larger unit, equivalent to one billion watts. Power measures the rate at which energy is generated, used, or transferred.

   1.5 Gigawatts could, depending on circumstances, provide electricity to about 1,300,000 homes for one year. (using an average of 10,000 kilowat hours per home). At complete build out, it is proposed to to need 6 Gigawatts is twice the size of all existing Maricopa data centers, and that output would power 3,000,000-5,000,000 homes.)

Ownership of Apache Generating Station: AEPCO is usually stated as the owner however, as the chart shows, the parent cooperative is Arizona G&T Cooperatives. In practice it is AEPCO that sells wholesale power to member coops, the largest being Sulphur Springs Valley Coop based in Benson. They then sell  retail electricity to residential and commercial consumers.(a cooperative is member-owned,not for profit organization)


Did anyone at the county know? Welllll difficult to say for sure, but it seems they would have had some indication this year. The supes went through a process of establishing regulations for data centers. In early 2026 they directed the staff with the help of the CC Development Services to draft requirements and technical criteria. The draft then went to Planning and Zoning Commission in March for approval. P&Z approved and it then went to the supes who passed the regulations on April 7, 2026;  Ordinance 26-09 (Section 2.51.040 of the Cochise County Zoning Regulations. There have been various claims, particularly by supervisor Antenori that he did not know in advance.

The Public meeting: An informational meeting was held last week in Cochise. At the meeting SpaceX formerly revealed that they were behind the application for the data center. About 300 people attended the raucous meeting where a presentation describing the project was given before SpaceX employees fanned out to talk with smaller break out groups. Many were opposed and had signed a petition with 1400 names. Some were supportive. Most wanted more information. 

The Decision Meeting: A meeting of Cochise County Planning and Zoning scheduled for December 9 at 2:00 at the county supe meeting room at 1415 Melody Lane will consider the application. At the hearing, the Planning and Zoning Commission can approve, approve with conditions, or deny the special use request. Comments can be given the day of the meet or sent in. For written comments go here, they have to be in by noon Nov. 27 to be included in the supes packets, after that they will only be in the record. If you wish to speak either remotely or in person, fill out this form before the meeting. Speaking time is 2 minutes per person.

Many Questions: There is a batch of questions about this application. Some answered and many not.  Planning and Zoning can't really make informed decisions until many answers are forthcoming. These are several that need answering for clarity.

  • There have been a few public mentions of NDA (non-disclosure agreements) but no information about who signed them. Was it AEPCO?, Sulphur Springs Coop? County Supervisors? County employees? It would be instructive to know who cannot talk about this project. (NDAs have become a controversial issue in other states because of the secrecy involved excluding the public, that has often resulted in green lighting a project before adequate information about possible impacts have been available for the public. See this Boondoggle article about two other projects involving SpaceX, NDAs, and public giveaways.)

  • If it there was advance notice, have there been any open meeting violations, as have happened in other states?

  • Does the county have expertise on staff to review the application in its total effect on energy transmission, residential consumption, water/air/soil, noise, light, and wildlife?

  • Has the application been reviewed by anyone outside the county? If so who?

  • If AEPCO is selling land, who are they selling it to, SpaceX or Tombstone Resources? And about when were they approached and talks begun? 

  • If Apache Generating Plant is going to sell electricity, when were those talks about available capacity and possible sale initiated?

  • Did supervisor Antenori or any supervisor know anything official or unofficial about either Tombstone Resources or SpaceX making inquiries or plans for a data center?  

  • Has the development dept or county staff had any contact with reps from either or both Tombstone Resources or Space prior to this announcement?

Coal Use....There were previously announced plans to wind down the use of coal and ramp up the use of renewables for generating. This deal seems to do the opposite. AEPCO/Apache GS received a $21 million grant from dept of energy recently to upgrade coal use for steam generation. Also mentioned was Another $32.5 million from an unidentified source non-governmental source. It would be useful to know what entity is supplying that money.

  • Currently Apache uses about 90% natural gas and 10% coal to power generators to provide 628 MW of power. What will be the percentage mix of coal and gas used to provide fuel for the steam generators to provide electricity for SpaceX when upgrades are completed?

  • What is the quantity of coal projected to be used?

  • Where will the coal be coming from?

  • Assuming coal will be transported using trains. How many and what frequency?

  • How will  solid, liquid, and gaseous byproducts  such as coal ash, slag, and carbon dioxide be dealt with? There is a secondary market for some of this, but are there companies in the county that can utilize it?

Natural Gas and power delivery

Currently natural gas is supplied through major interstate natural gas pipeline systems connected to regional supply basins, primarily drawing from the San Juan Basin in New Mexico and the Permian Basin in Texas via networks like Kinder Morgan's El Paso Natural Gas system.

  • There is a projected doubling of output at the Apache Generating Station in the first phase of this application, will they be able to supply additional gas for the upgraded facility?

  • How will the generated electricity be delivered to SpaceX?  Will new transmission lines be needed?

++++++++++   ++++++++++  ++++++++++


WHAT IS AI? TERMINOLOGY AND DEFINITIONS

Once again I is was befuddled by terms bandied about in almost all media outlets. So here is my attempt to understand them. The industry is developing so fast that some of the terminology will not be applicable in a year or maybe even six months. 

<><>

What is Artificial Intelligence/AI? This is a definition from Stanford University, "AI is a broad term for computer systems that can perform tasks with human-like intelligence, such as understanding language, recognizing images, learning from data, reasoning, and making decisions "

What are the forms of AI?  It is quite confusing as AI terms are used without coherent understanding of what they are nor how they are used. This is a simplified version...

Narrow AI:  It is classified as 'Narrow AI'  because it is built for specific tasks, like spam filtering, chess, or recommendations. Every AI in use today falls here. 

   Two other AI's, are the scary ones that are, so far, although often referenced as if they exist, do not yet exist.  Artificial General Intelligence (AGI): A hypothetical system that could learn and perform any intellectual task a human can. It doesn’t exist yet. Superintelligence (ASI): A hypothetical system that surpasses human intelligence across the board.


Terms and uses: In terms of capability the various kinds of AI break down like this. The following are simplified versions that I can understand. I  found it useful to think of Generative AI as the mother, and most of the other AI's as kids. There is a lot of overlap in using the different forms.


Generative AI: Creates new content, including text, images, audio, video, and code. It’s an umbrella category that includes many of the models below.

Discriminative/predictive AI: Classifies or predicts from input, such as “is this email spam?” or “will this customer churn?” Used for/by:

  • Banks and payment companies (Visa, Mastercard, PayPal, Stripe): fraud detection, deciding whether a transaction is legitimate.

  • Google (Gmail): spam and phishing classification.

  •  risk scoring and underwriting.Insurers and lenders (Lemonade, Upstart):

Large language models (LLMs) (a type of generative AI: Deep learning models trained on vast amounts of text to understand and generate language. Used for/by

  • Claude, GPT, and Gemini are examples. 

Diffusion models: Generate images (and increasingly video/audio) by gradually refining random noise into a coherent output. Used in tools like Stable Diffusion and Midjourney. Used for/by: 

  • Stability AI: Stable Diffusion.

  • Midjourney: image generation.

  • OpenAI: DALL·E and Sora for video.

  • Adobe: Firefly, built into Photoshop and Illustrator.

  • Google DeepMind: Imagen and Veo.

  • Runway: AI video tools.

Multimodal models: Handle multiple types of input or output, such as text, images, and audio, in one system. Used for/by: 

  • OpenAI (GPT models), Google (Gemini), Anthropic (Claude), Meta (Llama): handle text and images, with some also handling audio or video.

  • Microsoft: Copilot, which works across text, images, and documents.

  • Computer vision: Interprets images and video, for example facial recognition, medical imaging, and self-driving perception. Used for/by: 

  • Tesla and Waymo: self-driving perception.

  • Apple: Face ID and photo recognition.

  • Amazon: Just Walk Out checkout-free stores, plus Rekognition.

  • Mobileye: driver-assistance vision systems.

  • Zebra Medical Vision and Aidoc: medical imaging analysis.

  • John Deere: crop and weed detection.

Natural language processing (NLP): The broader field of working with human language, including translation, sentiment analysis, and summarization. LLMs are its current dominant approach. Used for/by:

  • Google: Translate and Search understanding.

  • DeepL: translation.

  • Grammarly: writing assistance.

  • Salesforce and Zendesk: customer support automation and sentiment analysis.

  • Bloomberg: financial news analysis and BloombergGPT.

  • Speech AI: Speech recognition (speech-to-text) and synthesis (text-to-speech). Used for/by

  • Apple (Siri), Amazon (Alexa), Google (Assistant): voice assistants.

  • Nuance (Microsoft): medical dictation.

  • OpenAI (Whisper) and Otter.ai: transcription.

  • ElevenLabs: realistic text-to-speech and voice cloning.

  • Zoom and Microsoft Teams: live captions and meeting transcription.

Recommendation systems: Predict what you’ll want next. Used by: 

  • Netflix: what to watch.

  • Spotify: Discover Weekly and playlists.

  • YouTube and TikTok: video feeds.

  • Amazon: product suggestions.

  • Meta (Facebook, Instagram): feed ranking and ads.

Robotics/embodied AI: AI that perceives and acts in the physical world. Used by:

  • Boston Dynamics: Spot and Atlas.

  • Tesla: Optimus humanoid.

  • Figure AI and Agility Robotics: humanoid robots for warehouses.

  • Amazon Robotics: warehouse automation.

  • Intuitive Surgical: da Vinci surgical robots.

  • iRobot: Roomba navigation.

  • Nvidia: robotics platforms and simulation tools.

To get to these varying AI's, It is necessary to power them and that is where the data center enters.


WHAT ARE DATA CENTERS? 


Servers & Data Centers: Until a few years ago, the term 'server farms' was used to indicate the processing of information and meant a huge room or building full of similar servers (a computer or network of computers  working together on one kind of job, like serving web pages or answering search queries. “Data center” is the broader, more formal term for any facility housing computing, storage, and networking equipment, along with the power, cooling, and security systems that keep it running.

Although often conflated, and people use the terms almost interchangeably, though “data center” is now the standard industry term.

The Difference matters because the real change is what the machines are doing, which changes the physical requirements.

  • Chips: Traditional web services run mostly on general-purpose CPUs (central processing unit), with each query being a relatively light task, such as searches and web pages. It works like this. The 'Client'  is your device (like a smartphone, laptop, or web browser) that sends a request (e.g., asking to load a webpage, send an email, or stream a video) to the Server: The server receives the request, processes it, and sends the requested data or service back to the client.

  • AI training and inference rely on GPUs( graphic processing unit) that are custom accelerators (like Nvidia’s chips), which are far more power-hungry and are designed to work in tightly linked clusters. They are only made by a few companies because of the strict antiseptic necessity for production. (The New Yorker has a fascinating article on the only company, ASML in the Netherlands, that makes the machines that can produce GPUs. Go here for the read-may be paywalled.)

  • Power density: A conventional CPU server rack might draw 5 to 15 kilowatts. An AI GPU rack can draw 40 to 100+ kilowatts, with newer designs going higher. That’s why a single AI campus can require hundreds of megawatts, sometimes approaching a gigawatt, comparable to a small city or a large power plant.

  • Cooling: Air cooling struggles at those densities, so newer facilities increasingly use liquid cooling, which changes the plumbing, water use, and building design.

  • Networking: Training a large model means thousands of chips constantly exchanging data, so these sites need very fast, specialized interconnects. A search farm’s servers, by contrast, can often work fairly independently of one another.

  • Workload type: A search query is a quick burst of work, spread across many machines, and it’s what the “farm” model was built for. Training a frontier AI model is a sustained, massive job that ties up an entire cluster for weeks or months. Providing AI answers (inference) is also much more computationally expensive per request than returning search results.

  • Scale and siting: Because of the power demands, developers now choose locations based on available electricity, grid connections, and sometimes on-site generation, which is why this proposal is tied to a power plant in a region with cheap energy.


Why this matters for the public debate

The controversy around new data centers mostly stems from these differences: electricity demand strains local grids and can raise rates, water use for cooling is a concern in dry regions, enviromental problems of noises, air, and waste, and the buildings employ relatively few people once built.The older server-farm era didn’t raise these issues at nearly the same intensity.

One caveat: not every new data center is AI-focused. Many still serve cloud storage, streaming, and enterprise computing, and the line between the two types is blurring as companies retrofit or mix workloads.


 
 
 

Recent Posts

See All
BISBEE WIRE #119

editor: fred miller November 1, 2026 #119/November 2026 Hello, If you haven't done it by now, I hope you will, and encourage any friends that haven't done it, to do it...vote I

 
 
 
BISBEE WIRE #118

BISBEE WIRE #118 editor: fred miller November 20,2026 #118/November 2026 Hello, With uncertainty about the mail, it's best if you get your ballot in soon if you haven't...the ve

 
 
 
BISBEE WIRE 2026 ELECTION ISSUE

Hello, This is my totally biased cheat sheet. It is a blatant attempt to woo your vote for the following candidates and propositions. I hope it is helpful. The information for this issue was compiled

 
 
 

Comments


bottom of page