Activation Energy is a good model for a lot of things

Hacker News Top News

Summary

The essay applies the concept of activation energy from chemistry to explain initial barriers in contexts like physics, neuroscience, and personal relationships, emphasizing the importance of low energy for sustaining actions.

No content available
Original Article
View Cached Full Text

Cached at: 08/19/26, 10:30 AM

# Activation Energy is a good model for a lot of things Source: [https://homosabiens.substack.com/p/activation-energy-is-a-good-model](https://homosabiens.substack.com/p/activation-energy-is-a-good-model) I hate every graph I found on Google for “activation energy” so our thumbnail for this essay is just gonna be [![Activation Energy (Ea) Chemistry Definition](https://substackcdn.com/image/fetch/$s_!Ntiy!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F740ae116-a8b6-40da-97ea-be50f953d4af_1500x1233.jpeg)](https://substackcdn.com/image/fetch/$s_!Ntiy!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F740ae116-a8b6-40da-97ea-be50f953d4af_1500x1233.jpeg) Heat some paper to 300ºF \(150ºC\) and mostly what happens is you have some hot paper\. Heat some paper to 500ºF \(260ºC\) and what you have is a*fire\.*You can use that fire to start other fires, and there’s a self\-sustaining thing happening that will go on until all of the locally available fuel is consumed\. If you’re trying to push a table or dresser across a floor, and it takes fifty pounds of force to get it to*start*moving, it probably only takes forty or thirty pounds of force to*keep*it moving\. Static friction is almost always greater than kinetic friction\. Most neurons in your brain are connected to something like a thousand other neurons\. Those other neurons are firing, and firing, and firing, and at some point*enough*of them fire in close\-enough proximity, and that causes*this*neuron to fire\. It’s more complicated than this very crude model, but if 10 of the neurons this neuron is “listening to” say “go,” usually nothing happens\. If 700 of them say “go,” something does\! Why am I writing this essay? Definitely not because an objective assessment of the ~80\-or\-so partially written essays I have in my stack reveals that*this*one is the highest product of effort and impact\. Nope, I’m putting this particular one out because a few musings on the concept of activation energy*got me sitting down at my desk and typing\.*The required initial investment in the essays Making the Other Monkeys Feel Bad, and You Are The Deer In The Subway, and Defense Against the Dark Arts \(which is like 93% finished already, even\) is just … a lot higher\. The initial investment for this one was low, partially because it’s short and simple and partially because my shower thoughts were already going there\. How do you make a certain kind of friendship last? In the summer of I\-think\-it\-was\-2006, Steph and Eric and Rebecca and I ended up hanging out every single day for something like five straight weeks, and it was*in part*because, as we headed home each night, we just naturally found ourselves asking “so, when are we starting tomorrow?” Each hangout begat the next hangout, which begat the hangout after\. In the frame of activation energy, you could either view this as “we made the activation energy for*each additional day*really low, somehow” or as “yeah, once we activated Being The Type Of Friends Who Hang Out, the energy it took to*sustain*that was pretty low\.” There’s momentum coming out of basically anything, and you either sustain it, or it fades, and typically starting over from scratch requires much more effort than keeping it going\. How much energy does it take for your romantic partner to broach a difficult conversation with you? How*hard*is it, for them to arrange the preconditions for that sort of conversation to happen at all \(let alone to arrange the preconditions for it to go well\)? Can you do anything to*lower*that activation energy? Presumably, you want to, because presumably you*don’t*want your partner just … sitting on a bunch of unmentioned resentments that build and build and build and build… Putting the cookies on a high shelf is meant to be an activation energy intervention—it’s meant to make it*harder*to access the cookies, so that you’re less likely to do so\. Putting your exercise bike in your living room, where you see it every day, is meant to be an activation energy intervention, to*reduce*how much oomph it takes to get you to exercise\. But I think a lot more things are accurately modeled by activation energy than just the obvious ones that everybody already talks about\. There’s a way that you could squint, and think about pretty much*everything*in terms of activation energy\. \(You could model sex as happening when both partners exceed the activation energy threshold, and*not*happening when only one does\.\) \(I don’t recommend permanently shifting your worldview in that way, and fusing activation\-energy\-goggles to your face forever, but I*do*sort of recommend spending a whole week keeping activation energy at the front of your mind, and letting your brain sink into the Tetris effect, and seeing it everywhere, from how much activation energy it takes to get you to open your mouth and talk, to how much activation energy it takes to cause the world to[do something useful about the race to lethal superintelligence](https://aistop.watch/)\.\) Where does your activation energy*come*from? “Introverts” get more of it from quiet time, and “extroverts” from other monkeys\. Most people have more of it if they’ve eaten in the last few hours than if they haven’t\. Some people get disproportionately large AE from*negative*stuff—they’re highly motivated to avoid bad outcomes\. Others find that the big boosts come from positive, “toward” energy, visualizing the things that are exciting and happy\. Do you have multiple streams? Is your AE*fungible,*i\.e\. largely repurposeable and redirectable? Does anything ever pull the plug out from under, leaving you with no energy to do anything at all? \(What*do*you do, when it feels like you have no activation energy? What’s your*autopilot*like? Does it take activation energy for you to*change*your autopilot?\) [![](https://substackcdn.com/image/fetch/$s_!nwnY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe61f0d4c-1e2d-4970-bcfa-f87b6ee69a74_749x473.jpeg)](https://substackcdn.com/image/fetch/$s_!nwnY!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe61f0d4c-1e2d-4970-bcfa-f87b6ee69a74_749x473.jpeg) Again, I’m not recommending a total reframe, here\. This isn’t “oh, my whole life will be different, if instead of saying*I don’t wanna,*I say to myself*This doesn’t meet the activation energy threshold\.”* But I do think it’s worth doing that for a*little*bit\. Looking back over the past hour, maybe, or looking ahead to the next week\. What sorts of activities cost you nothing? What sorts of activities cost a lot, but are worth it? What sorts of activities cost a lot, but*aren’t*really worth it \(but also once you’re in them you’re kind of stuck for a while, oh god, why did I even come to this party\)? Are any AEs set at a*weirdly high*threshold, requiring more than you would’ve thought? Do you have the sense that some activity “should” only take so much energy to activate, but in practice it takes a lot more? Where’d that should come from? Etc\. #### Discussion about this post ### Ready for more?

Similar Articles

Learning concepts with energy functions

OpenAI Blog

OpenAI presents a technique using energy functions to enable agents to learn and extract abstract concepts (visual, spatial, temporal, social) from tasks, then transfer these concepts to solve related tasks in different domains without retraining. The approach uses energy-based models with neural networks to perform both generation and recognition of concepts.

Energy per Successful Goal: Goal-Level Energy Accounting for Agentic AI Systems

arXiv cs.AI

This paper proposes A-LEMS, a framework that redefines AI energy accounting from per-inference to Energy per Successful Goal (EpG), and introduces the Orchestration Overhead Index (OOI) to measure energy costs of multi-step orchestration in agentic systems. Empirical results show agentic workflows consume 4.33× higher mean energy per goal than linear baselines, but OOI can invert for tool-augmented tasks, demonstrating goal-level accounting is necessary.

The Energetic Costs of Cellular Computation (2012)

Hacker News Top

This paper theoretically computes the energetic costs of a simple cellular network computing ligand concentration, showing that breaking detailed balance and consuming energy is necessary for learning, with implications for resource-limited environments.

Implicit generation and generalization methods for energy-based models

OpenAI Blog

OpenAI presents implicit generation and generalization methods for energy-based models (EBMs) that use Langevin dynamics for iterative refinement to generate samples without explicit generator networks. The approach offers advantages including adaptive computation time, flexibility in learning disconnected data modes, and built-in compositionality through product of experts.

Constructed Reality, Contested Priors: Decoupling and the Architecture of Cognitive Relapse Under the Free Energy Principle

arXiv cs.LG

This paper investigates the concept of 'ontological inversion' under the free energy principle, using a convolutional variational autoencoder to explore whether a synthetic environment can permanently replace a system's default generative model. The study finds a decoupling between representational accuracy and default behavior, introducing the phenomenon of 'cognitive relapse' where the system partially reverts to its original model.