Today, we're going to discuss AI and Energy Simulation. AI suggests you will find these topics life-changing, so please read and explore the links! Machine learning isn’t just a buzzword; it has already impacted all of us, whether we realize it or not. Many of us are familiar with digital twins and other innovative tools. A significant question arises: Will AI replace our jobs?

Over the past few months, I've explored AI tools extensively. The team at Energy-models.com can now accomplish tasks that previously required a freelancer. I can complete many tasks faster and more accurately on my own. Here are a few things we've achieved:

• Developed a new type of video for faster, more concise “how-to” videos
  Here's an example from the latest eQUEST tutorial
• Cloned my voice, allowing any editor to update my videos
• Optimized and enhanced old audio (a remastered OpenStudio demo)
• Generated captions with punctuation and transcripts, reducing costs significantly

AI has surprised us in many ways. For instance, from a few transcripts, it managed to create eQUEST and OpenStudio code snippets for ASHRAE 90.1 items. AI has undoubtedly made certain tasks easier and faster. It has also replaced several jobs, removing pay from our transcriptionist, content organizer, graphic designer, and video editor. Even our web developer went on vacation, and I was able to do his job in ways I never imagined possible. AI even replaced me as an instructor in a random video (we'll see if anyone notices and if it engages users more effectively).

AI rewrote this entire post. No, I'm joking. It proofread, and I had to undo some of its work. I believe I'm still a more engaging writer.

Everyone is crazy about ChatGPT. 

• eQUEST Modeling Assistant
• OpenStudiOracle - An Oracle to inquire about OpenStudio
• Optimize Architect - A game of 20 questions that suggests Energy Efficiency Measures based on your answers.

The Big Picture Simulation

Back in 2019, which feels like the Stone Age for artificial intelligence, a team used AI to simulate the universe. The simulation began with a certain amount of matter and energy, then projected what would happen over time. Questions like, "Do stars form?" "How many and how big are they?" and "Does the simulation collapse into a black hole?" were explored. Given the complexity of the universe, it's understandable that initial simulations required 300 computation hours.

The team developed faster methods, reducing calculation time to two minutes, though this speed increase came at the expense of accuracy. The fast models had a relative error of 9.3% compared to the initial model. Scientists developed an AI method, called the Deep Density Displacement Model (D3M), focusing on one parameter (the amount of ordinary matter in the universe) and leveraging thousands of models for machine learning. This effort paid off, producing an AI simulation that computed within 30 milliseconds and improved the relative error to 2.8%—making it three times more accurate and 3,000 times faster than the other rapid simulation method.

Mysteriously, the D3M model worked for parameters it hadn't been trained on. When the team altered the starting conditions, such as the amount of dark matter, the model still simulated accurately.

For those familiar with machine learning, this outcome is puzzling. None of the scientists understood how the AI model worked. I've wondered if any of them considered, "Maybe it's because dark matter doesn't exist, and the AI figured that out." Fun fact: I named my business "Cosmo Logic" because I believe simulations will solve the cosmologic constant problem without needing the "boogeyman" of dark matter and dark energy.

It remains unknown whether the D3M model discovered an algorithm that functions without dark matter or if it transforms Planck's constant into something entirely unexpected. Machine learning doesn't "show its work," after all.

Why Is This Relevant to Building Energy Simulation?

Physicists have led the way in simulations and AI, yet robots haven't replaced their jobs. In fact, there are more simulation scientists now, conducting far more advanced simulations. The roles lost have been in data collection, data entry, and post-processing.

Simulations serve as a tool to explore "What happens if?" The value of a simulation scientist lies in answering "Why did it happen?" and "What can we do about it?"

As building simulation specialists, this is crucial for us. We're in a field that will grow with AI. Those who merely input data into models may find their roles evolving, but those who excel will be:

• Experts in understanding complex methods
• Inquisitive individuals who ask the right questions and apply AI effectively
• Professionals who know how to translate data into actionable insights

So, if your current role involves basic model input, consider this a call to deepen your knowledge and secure your position in an exciting future.

Thanks for reading,

Bob (Or was this entire thing written by AI?)