In the Weights: Check Your Influence Score in the AI World

What Is In the Weights?

In the social media era, we've grown accustomed to measuring personal influence through follower counts and likes. Now, a new tool called In the Weights brings this kind of "vanity search" into the AI domain — attempting to quantify how much of a presence and influence you have in the world of artificial intelligence.

Simply put, In the Weights is an AI-centric personal influence scoring tool. You can search your own name to see how much "weight" you carry in the AI field. The concept itself is loaded with AI-community humor — "weights" refers both to the core parameters of deep learning models and serves as a metaphor for your "clout" in the field.

"Weights" in deep learning: In neural networks, weights are the core parameters a model learns through training. GPT-4, for example, is estimated to have over a trillion parameters. These weight matrices encode compressed representations of human language, knowledge, and reasoning in a distributed fashion. When we say someone's ideas are "in the weights," we mean that their writing, code, and opinions have diffused throughout the network structure, shaping how the model understands and responds to related topics — not stored in any specific location, but leaving a kind of statistical "imprint."

Why Does an AI Influence Score Matter?

A New Attempt to Quantify AI Influence

Traditional academic influence metrics — such as the H-index and citation counts — are well-established but tend to cover only the academic world. The H-index, proposed by physicist Jorge Hirsch in 2005, is defined as: a researcher has an h-index of h if h of their papers have each been cited at least h times. This metric accounts for both quantity and quality of output, but its limitations are clear: it relies heavily on the peer-reviewed academic publishing system and is nearly useless for AI practitioners whose primary contributions take the form of open-source code, technical blog posts, or engineering work. Linus Torvalds created the Linux kernel; Andrej Karpathy drove the engineering adoption of computer vision — yet these kinds of contributions are often severely undervalued by traditional H-index metrics.

The AI field's reputation system has undergone a profound transformation over the past decade. The academic prestige once centered on top-conference papers at NeurIPS and ICML is now being supplemented by a far more diverse ecosystem: GitHub star counts measure the popularity of open-source projects, model download numbers on Hugging Face reflect the real-world impact of engineering contributions, technical discussions on Twitter/X shape immediate community consensus, and arXiv preprints have broken down the time barriers of traditional peer review. This engineering culture — where "building it" matters more than "writing about it" — is precisely why the AI field needs a new kind of influence evaluation framework.

The emergence of In the Weights reflects the industry's demand for multi-dimensional influence assessment. It's not just an entertainment tool; it represents a broader trend: the AI field is developing its own unique reputation system.

From Googling Yourself to AI-Searching Yourself

"Vanity search" is nothing new. Ever since the search engine era, people have enjoyed Googling their own names to see how much of a footprint they've left on the internet. In the Weights transplants this behavior into an AI context, but gives it a more structured presentation — not a simple list of search results, but a calculated "score."

This design hints at an interesting technical question: How much information about you is actually in an AI model's training data? The pretraining corpora of mainstream large language models typically draw from sources like Common Crawl (a web crawl dataset covering billions of pages), GitHub repositories, academic paper databases, and book corpora. Whether a person's name and ideas make it into the training data depends on how frequently they appear online and the authority of those sources. If your name, opinions, and code show up repeatedly in training data, then in some meaningful sense, you really are "in the weights."

It's worth noting that models don't simply "store" information — they learn associative patterns through statistical regularities. This means that even if someone's information appears in the training set, the model's "knowledge" of that person may still contain hallucinations, biases, or gaps.

The Deeper Thinking Behind In the Weights

Reimagining Digital Identity in the AI Era

This tool raises a deeper question: in the age of large language models, our digital identities are being redefined. In the past, your online presence was determined by how many web pages a search engine could find about you. Now, it may be determined by how much an AI model "remembers" you — whether your ideas have been internalized by the model, whether your code has influenced the behavior of AI systems.

There's a profound paradox here: your existence within a model is a compressed, distorted, and potentially warped statistical projection — not a faithful representation of your true self. The model's "understanding" of you is essentially a statistical average of everything you've ever said publicly. It may capture your most representative views, or it may form a skewed impression due to biases in the training data. This makes an "AI influence score" both a mirror and a funhouse mirror at the same time.

Motivation and Anxiety, Side by Side

Of course, any scoring system is a double-edged sword. It can motivate practitioners to engage more actively in the AI community, share knowledge, and contribute to open-source projects — but it can also intensify anxiety and competitive pressure within the industry. After all, not every excellent AI practitioner operates in the public eye; many key contributors quietly drive technological progress inside companies. In the Weights attempts to build a unified quantitative framework within this complex ecosystem, and the challenge lies in fairly integrating these highly heterogeneous dimensions of contribution. How should this system compare an engineer who trained a breakthrough model on a closed-source team with an AI science communicator who has a million followers?

Go Check Your AI Weight Score

Regardless of whether In the Weights ultimately becomes a widely adopted tool, it at least raises a question worth pondering: In an era where AI is reshaping everything, how do we measure an individual's participation and contribution to this transformation?

If you're an AI practitioner, researcher, or enthusiast, it's worth giving it a try to see your own "weight score." Even if the result is underwhelming, don't take it too seriously — real influence has never been something a single number can fully capture. But as an interesting industry phenomenon, In the Weights does offer a unique window into the AI community ecosystem.

Key Takeaways

• In the Weights is an AI-centric personal influence scoring search tool that quantifies a user's "weight" in the AI field
• The tool's name is a clever double entendre: it refers both to the trainable parameters in deep learning models and to a person's standing in the AI world
• It reflects the industry's need for multi-dimensional influence assessment, going beyond traditional metrics like the H-index that only cover academia
• It sparks deeper thinking about digital identity in the AI era — your presence in a model is a statistical projection of your real self, subject to compression and distortion
• Scoring systems are double-edged swords: they may encourage community participation, but can also heighten industry anxiety, and struggle to fairly account for closed-source contributors