AI Has Learned How to Capture Our Emotional Metadata

Was this newsletter forwarded to you? Sign up to get it in your inbox.

---

We've spent years trying to get large language models to obey instructions, but what happens when they start obeying our feelings instead?

Most conversations around GPT-4.5 (and whatever other model we're currently fussing about) obsess over functionality: code debugged, emails written, summaries summarized, or benchmarks outperformed. Basically, we've been building really fancy, really expensive software to automate stuff we already know how to do—just faster and cheaper.

But maybe we're missing what's actually interesting here. What if the real magic isn't in getting AI to flawlessly execute tasks, but in leveraging a more subtle, powerful ability to read between the lines of the words in a prompt and pick up on the intent behind them? Previous generations of software were already great at helping us be more productive at tasks that have well-defined, step-by-step processes and clear answers. Excel power users, for example, can whiz through financial modeling. I see little reason to fix that. Perhaps our mistake was in trying to make LLMs automate these things, instead of focusing on what new capabilities could be unlocked. 

This shift is just now gathering steam: Computing is moving away from explicit instruction-following toward intuitive preference-reading. 

This wasn’t deliberate. Research labs have been trying for years to replicate existing digital workflows. However, this happy side effect is what I’m calling “vibe compute.” On the frontier, AI models are already moving up layers of thoughts to more abstract thinking. They still struggle at things human children would do easily—try asking them to count the number of Rs in the word “strawberry.” And yet, these models are able to intuit my tastes and preferences in ways that continue to shock me.

I’ve had this idea for a while but have so far lacked clear examples by which to prove my point. With GPT-4.5’s release last week, this latent capability finally feels legible to the average user. Vibe compute is here. 

The implications of this shift feel wildly under-discussed. When talking about the role of probabilities and emotions in our relationship with computers, most people post a screenshot of Joaquin Phoenix in the movie Her and call it a day. Instead, I want to truly interrogate the idea of an intuitive machine. What does it mean when LLMs are bad at math, but great at feelings? 

The three breakthroughs behind vibe compute

Before we assume that I’m right, it’s worth going through the three technical breakthroughs that have made this leap possible: 

Context is baked into the training

At the risk of being too simplistic, a large language model works by predicting the most likely next word (or token) based on the context of everything that came before it. During training, an LLM learns to calculate probabilities for what should come next. When ChatGPT writes a response to a prompt, each word it selects isn’t chosen randomly but is instead weighted by these probabilities. What’s astonishing is how sensitive this approach is to picking up not just the content of text, but the context as well. It’s like metadata—the subtle things we encode in our writing, the nuanced cultural signals, our perception of how certain arrangements of words translate to our idea of “high-quality writing” are also encoded in the training. As models continue to get bigger and more powerful, those encoded vibes become a more prominent part of the system. 

Learning your preferences over time using memory features

One of ChatGPT’s least heralded features is its ability to store “memories” about you that it can use to improve its responses. Over time, it picks up personal facts but also remembers your interests and preferences.

Probabilistic compute output variance

Until LLMs, all computers were deterministic. The same input would generate the same output every time. Large language models are probabilistic. The same input will generate many outputs. They are similar but always contain some variation.

Together, these developments mean that models can give you what you want—even when you don’t know how to express it. AI products shift from solving jobs to be done to focusing on satisfying the emotions behind a request.

If you look closely, you can see glimmers of this happening in three product categories today:

1. Vibe coding with tools like Cursor, Replit, or Lovable: Modern coding assistants are able to translate vague instructions into functioning applications. When you say "this feels clunky" or "make this more elegant," they are able to grok what you actually mean. The code becomes a conversation about feel rather than pure functionality. My colleagues at Every, including writers and designers, have been known to code apps on their phones in bed. You can make an app poolside.
2. Writing and editing with GPT-4.5: This model only came out a few days ago, but the more our team uses it, the more impressed we are. It catches subtle signals about tone and style. Tell it "that sounds too formal" or "make it punchier," and it adjusts accordingly. The writing process becomes less about explicit instructions and more about collaborative refinement based on feel. This is exactly what working with a great human editor feels like. Now you can do it with a model.
3. Seamless conversation: Voice-enabled AI assistants are moving beyond simple command-and-response to expressing subtle, remarkable emotion. A recent demo from Sesame is a must-try if you haven’t yet. The voice is so good (and strangely flirty?) that it feels natural. Try honestly expressing your problems, or chatting about something the way you would in therapy. 

These products work because if you say something like “that output sucked, can you make it better?” the models can understand what you mean. While this has always been a possibility, it has only started coming true in the last three months or so. It’s a genuine, but subtle, breakthrough that requires us to evolve our relationship with our machines.

A two-track future

One of the hardest parts of dating is trying to understand if the other person is picking up what you're throwing down. With LLMs, we're entering a strange new territory where our computers might understand the vibes we are putting out there better than we do ourselves.

The biggest shift that might result is not AI doing what we tell it to do, but AI understanding—and acting on—what we mean even when we can't articulate it ourselves. In this future, this relationship becomes genuinely collaborative and increasingly intimate. 

It also raises profound questions about self-knowledge and technology. What happens when your AI writing assistant knows your voice better than you do? When your coding companion understands your programming philosophy more clearly than you've ever articulated it? When your AI therapist recognizes patterns in your emotional responses that you've never noticed?

Perhaps the most valuable thing these systems will offer isn't efficiency or automation, but insight—a new lens through which to understand ourselves. 

That might be uncomfortable. It might be revelatory. But it's where we're headed. In the age of intuitive computing, the objective isn't just to build machines that do what we say—it's to build machines to do what we don’t even think to ask. 

---

Evan Armstrong is the lead writer for Every, where he writes the Napkin Math column. You can follow him on X at @itsurboyevan and on LinkedIn, and Every on X at @every and on LinkedIn.

We build AI tools for readers like you. Automate repeat writing with Spiral. Organize files automatically with Sparkle. Write something great with Lex. Deliver yourself from email with Cora.

We also do AI training, adoption, and innovation for companies. Work with us to bring AI into your organization.

Get paid for sharing Every with your friends. Join our referral program.