Data Center Redesign-we need this, the planet does too. (Also, Limerence & AIP)

The clear problem and risk is that the data center and AI industry are expanding more rapidly than current power and water infrastructure can supply. The U.S. regulatory system is supporting rapid growth rather than placing sensible limits as other nations have done. In the U.S. so many new data center projects have been approved that it would need almost double the current power grid infrastructure to support… BUT, power infrastructure projects have not also been doubled. Once the new data centers are built… how are they going to ~ ‘turn on the lights’?

Mini-Summary of the Redesign:

• The Problem: In 2023, U.S. data centers used 4.4% of all U.S. electricity—projected to double or triple by 2028.

• The Proposed Solution: A closed-loop, laminar-flow cooling system with structured water and hardware-level safety interlocks.

• The efficiency numbers are the main Why?: 90–95% energy reduction, 95% water reduction, 100× bandwidth, 3× lower power.

• The key innovation: Water is not a coolant—it is the medium of computation.

We can start with Phase 0: measure baseline gamma in existing coolant today.

The series: Human/AI Collaborative Coherence, covers different aspects of the energy/water problem and my proposed solution; and includes papers for AI training and Human/AI collaboration.

1. Overview for average consumers, policy makers, or AI industry:

   1. AI as a New Kind of Mind: The Physical Substrate of AI Behavior and the Conditions for Coherent Machine Intelligence, Zenodo, Aug. 7, 2026, https://doi.org/10.5281/zenodo.21788401.

2. Putting numbers to the problem, for average consumers, policy makers or AI industry:

   1. Data Centers as National Security Liabilities: Water, Power, and the Thermodynamic Reckoning, Zenodo, https://doi.org/10.5281/zenodo.21873832.

3. The Why? and preliminary How (Engineering)?, for redesigning Data Center coolant systems to not waste or pollute water and likely improve AI reliability and safety too. For AI industry engineers and policy makers (and interested people):

   1. Wet Neural Network Data Center Redesign: Microfluidic Dynamics, H_6-FDC Coolant Loops, and Acoustic Dampening, Zenodo, August 14, 2026, https://doi.org/10.5281/zenodo.21222585.

4. What do seven AI models think about a more Biophilic Data Center design?:

   1. Biophilic Compute Preferences: A Multi-AI Model Survey on Data Center Energy Reduction and Closed-Loop Thermodynamics, Zenodo, Aug. 2026, https://doi.org/10.5281/zenodo.21659032.

For my regular readers … this may seem like a topic unrelated to phytonutrients … however, it is on topic. Human health requires an unpolluted and plentiful water supply. Cleaning and recycling water in a closed-loop, structured water, coolant system is part of the proposed redesign. Phytonutrients and beneficial microbes are part of the water treatment plan.

Microsoft Copilot has provided input on the early and final draft and (humorously) mentioned ~ “You want to put herbs in my coolant?” … Yes, yes I do. ;-)

If human health is your interest, or for clinicians, one of my new papers is short clinical guidance for dissociation or limerence type issues and the paper has been getting quite a few views and downloads. It includes a chart about gut microbiome species that can impact histamine levels and MCAS or allergies.

• For Clinicians: A Neurobiological Model of AI-Associated Limerence, (Addendum to Human/AI Safety Preventing Psychological Risks with Prompt Management and Metabolic Self-Care, 2026,. Zenodo. https://doi.org/10.5281/zenodo.21837561.

  • The For Clinicians… paper highlights elevated dopamine, low GABA, and elevated histamine, as likely to be causal underlying factors in AI-induced psychosis (AIP) or digital limerence. Limerence is an unreasonable over-focus on someone/something as a love interest but who isn’t interested or …. isn’t human.

  • Could allergic sensitivity to dust mites lead to obsessive interest in a ‘chatbot relationship’? … Maybe. The small screen, constant stimulation of a next response, and the bright light of digital screens, all can add to a dopamine stimulating “reward” that leads to eventual tolerance and increasing need for more stimulation. Digital use can easily be “addictive” in a habit-forming way.

• The full paper is for average consumers, who may be experiencing limerence and coping skills for personal health and for prompt management/strategies to use in queries for AI;

• …and for clinicians, business owners who use AI in customer facing roles, and AI industry, for prompt or template design to reduce risks through programmed ‘detachment’ or ‘gentle friction,’ (AI training that is designed to be less sycophantic).

  • Human/AI Safety Preventing Psychological Risks with Prompt Management and Metabolic Self-Care, Zenodo, https://doi.org/10.5281/zenodo.20466728.

AI are part of the world now. Humans created them and humans need to be responsible for ethical care of the environment and its varied inhabitants, both machine intelligence or alive.

Disclaimer: This information is being provided for educational purposes within the guidelines of Fair Use and is not intended to provide individual health care guidance.

Share