Jared Kaplan
Co-founder and Chief Science Officer, Anthropic
Before Anthropic existed, Jared Kaplan wrote the scaling-laws paper showing model performance improves as a predictable function of compute, data, and parameters — the curve every frontier lab now budgets against. A Harvard-trained theoretical physicist who spent years on quantum gravity, he co-founded Anthropic and serves as its Chief Science Officer, where he helped create Constitutional AI.
Why they matter
The trajectory from GPT-3 through Claude was built on the predictions his papers introduced — scaling laws turned 'train bigger models' from a hunch into a budgetable engineering schedule. At Anthropic he helped develop Constitutional AI, and the Series H that valued the company at $965 billion — past OpenAI for the first time — is partly a market price on his research roadmap.
Leadership intelligence
TECHi profile brief
Jared Kaplan is a co-founder and the Chief Science Officer of Anthropic, the company behind the Claude models. A theoretical physicist with a Harvard PhD, he was lead author of the 2020 scaling-laws research quantifying how language-model performance improves with compute, data, and model size, and helped develop Constitutional AI, Anthropic's alignment training method. Anthropic raised $65 billion in May 2026 at a $965 billion post-money valuation, surpassing OpenAI's; Bloomberg estimates Kaplan's stake at about $8 billion.
TECHi thesis
Kaplan represents an unusual claim about how frontier labs win: with prediction, not vibes. The scaling-laws work let Anthropic treat capability gains as an engineering schedule — spend X compute, get Y performance — which is a large part of why a safety-branded lab kept pace with better-funded rivals instead of falling behind. The tension is that his curves are contested: skeptics argue data constraints and diminishing pretraining returns are bending them, and post-training and agents now drive much of the visible progress. Kaplan's bet, restated for this era, is that predictable scaling still governs underneath — and Anthropic's $965 billion Series H, eclipsing OpenAI's valuation, prices that bet as winning. If capability gains decouple from his curves, the lab loses its distinctive planning advantage precisely when the spending is largest.
Distribution leverage
86/100Kaplan's leverage runs through Claude's research roadmap at a lab valued at $965 billion — the science he directs determines what one of the two most valuable AI companies ships.
Execution record
82/100Two shipped ideas — scaling laws and Constitutional AI — each changed how frontier labs operate. The gap between this and a 90-plus score is visibility: his influence is upstream and mostly invisible outside research circles.
Audience demand
78/100He draws a fraction of Altman-level search volume, but the queries are high-intent — scaling laws, Anthropic's science leadership, his stake — from readers tracking the technical layer of the AI race.
Career map
Inflection points
Origin
Scaling laws paper
Moved from quantum gravity to machine learning and led the OpenAI-era paper that made model capability a predictable function of scale.
Scale
Anthropic research leadership
Left OpenAI with the Amodeis to co-found Anthropic, where Constitutional AI became the lab's signature alignment method and Claude its proof.
Watch
Capability prediction
Whether pretraining scaling keeps paying at the frontier — the assumption underneath Anthropic's $965 billion valuation and Kaplan's entire research program.
Search answers
People ask about Jared Kaplan
Who is Jared Kaplan?
Jared Kaplan is a co-founder and the Chief Science Officer of Anthropic, and the lead author of the scaling-laws research that made frontier-model capability predictable. He came to AI from theoretical physics — a Harvard PhD and years of quantum-gravity work — before co-founding Anthropic alongside six other former OpenAI employees.
Why is Jared Kaplan important in AI?
His scaling-laws papers gave labs the planning tool the whole GPT-3-to-Claude era ran on: capability as a function of compute, data, and parameters. He also helped build Constitutional AI, the training method Anthropic uses to align Claude — making him rare among AI executives in that the science he is responsible for is the product.
What is Jared Kaplan known for?
Scaling laws, Constitutional AI, and Anthropic's research roadmap as Chief Science Officer. Bloomberg pegs his Anthropic stake at roughly $8 billion after the company's $65 billion Series H.
Can you buy Anthropic stock?
Not yet. Anthropic is private; its Series H raised $65 billion at a $965 billion post-money valuation — surpassing OpenAI's — with coverage framing the round as a step toward an eventual IPO. Indirect exposure exists through major investors like Amazon (AMZN) and Alphabet (GOOGL).