Dan Shipper & Nathan Baschez - Every reaches 70,000+ subscribers Every raises $2M seed led by Reid Hoffman Nearly 100% AI-written code - Every's engineering team Nathan Baschez - Lex surpasses 300,000 users Every: 5 software products + 7-figure revenue Dan Shipper - "Most new software will just be Claude Code in a trench coat" Chain of Thought column - AI's definitive weekly read Nathan was Substack's first employee - now building what's next Dan Shipper & Nathan Baschez - Every reaches 70,000+ subscribers Every raises $2M seed led by Reid Hoffman Nearly 100% AI-written code - Every's engineering team Nathan Baschez - Lex surpasses 300,000 users Every: 5 software products + 7-figure revenue Dan Shipper - "Most new software will just be Claude Code in a trench coat" Chain of Thought column - AI's definitive weekly read Nathan was Substack's first employee - now building what's next

YesPress Profile — Founders & Media

Dan Shipper & Nathan Baschez

Two newsletter writers walked into the AI era. They didn't just report on it - they built the thing everyone reads about it.

Co-founders of Every - the only AI media company where the engineering team writes almost no code, the CEO predicts the future every Thursday, and Reid Hoffman showed up with a check.

70K+ Subscribers

$2M Seed Round

5 Products

~100% AI Code

Who They Are

The media company that ate its own homework

Every is the publication that didn't just write about AI - it became a live experiment in what an AI-native company looks like when you actually mean it.

Dan Shipper flies from his college graduation to Boston to close an acquisition deal. Nathan Baschez writes the software that makes Product Hunt tick before Product Hunt is famous. These are not people who watch things happen from a distance.

The company they built together started as two separate newsletters merged on a hunch in April 2020: Dan's Superorganizers (interviews about how brilliant people stay organized) and Nathan's Divinations (business strategy through the lens of Christensen and Porter). Six hundred paying subscribers became a thousand in a month. Then GPT-3 arrived and Dan had what he later called a revelation - not a pivot, a transformation.

Every isn't a newsletter that covers AI. It's a newsletter written by people who use AI to build products, run operations on AI agents, and spend their days at the literal frontier. The daily dispatch lands in 70,000+ inboxes belonging to the exact people building tomorrow's tools.

The line between writer and builder is blurring.

- Dan Shipper, CEO of Every

Dan writes the "Chain of Thought" column every Thursday - predictions, frameworks, arguments about what agent-native software architecture means in practice. Nathan, having spun out his AI writing tool Lex (300,000+ users, $2.75M raised from True Ventures), now runs a separate bet on what AI does to the word processor. Two different companies, one shared origin story.

In May 2025, Reid Hoffman co-led a $2M seed round into Every on the same day The New York Times ran a feature on them. A $2M round sounds modest until you learn Every already has seven-figure annual revenue, five software products, and 25 employees - the round was structured as a "sip seed," drawn down as needed. Hoffman had already appeared on the AI & I podcast multiple times. This was a bet he'd been watching for a while.

Quick Facts

The Origin

From two Substack newsletters to a $2M-backed AI company

The story most people tell about Every starts in 2020. The actual story starts earlier - with a kid in New Jersey who built BlackBerry apps to pay for college, and another who quietly coded the first version of Product Hunt before anyone knew it would matter.

Dan Shipper studied philosophy at Penn because, as he put it, it was "the only major that would tell me how to do it" - how to live well. He also sold his co-browsing startup Firefly to Pegasystems for several million dollars the same day he graduated. He flew directly from the graduation ceremony to Boston to close the deal. He was 22.

Nathan Baschez took a different route: building tools that other builders use. He went from designing Product Hunt to running product at Gimlet Media (winning the first Cannes Lion for an Alexa Skill) to becoming Substack's first employee. He built the "Baschez Score" - an internal system for identifying emerging writers - before anyone else at Substack had thought to look for them systematically.

When they merged Superorganizers and Divinations in April 2020, it wasn't a business plan. It was a bet that two writers covering adjacent terrain would be stronger bundled. They were right.

2012-2014

Dan co-founds Firefly, gets First Round's Dorm Room Fund investment, sells to Pegasystems on graduation day

2013

Nathan designs and builds the first version of Product Hunt

2016-2018

Nathan as Head of Product at Gimlet Media; wins first Cannes Lion for an Alexa Skill

2018-2019

Nathan joins Substack as first employee, builds core features and the "Baschez Score"

2019

Dan starts Superorganizers (productivity interviews); Nathan starts Divinations (business strategy)

Apr 2020

Newsletters merge into "Everything" bundle on Substack - 600 to 1,000 paying subscribers in one month

2021

GPT-3 arrives; Dan pivots Every's editorial focus toward AI; every.to launches as independent platform

2022-2023

Nathan starts Lex nights-and-weekends; launches publicly with 25,000 day-one signups; raises $2.75M from True Ventures

May 2025

Every raises $2M seed co-led by Reid Hoffman and StartingLine VC; NYT feature drops same day

2026

Every: 70,000+ subscribers, five products, 7-figure revenue, nearly 100% AI-written code

The People

Two builders, two obsessions

The Company

Every: Media company, software company, or something else entirely

✍️

Editorial

Daily AI newsletter to 70,000+ founders, operators, and investors. Columns include Chain of Thought, Superorganizers, and Divinations. Day-one "vibe checks" of every new major AI model.

📱

Software Products

Five distinct products: Spiral (AI writing), Cora (AI email), Monologue (voice dictation), Sparkle (file organization), and Proof (collaborative editing). Plus One handles Slack-based AI agents.

🎧

Podcast & Audio

Dan's "AI & I" podcast interviews the people actually building AI - founders, investors, filmmakers. Reid Hoffman is a recurring guest. Distributed on Apple Podcasts and Spotify.

🤖

AI Operations

Every runs on four internal AI agents: Anton (prioritization), plus meeting notes, OKR planning, and growth tracking agents. Nearly 100% of code is AI-written.

💰

Consulting

A dedicated consulting arm advises companies on AI adoption and operations. Every practices what it publishes - the consulting work is informed by the company's own AI-native experiments.

🌎

Community

Subscriber-only Discord community for the people who read every.to. Also cited by The New York Times and The New Yorker as a source on AI trends.

70K+ Daily subscribers

7-fig Annual Revenue

~25 Team members

$2M Seed (May 2025)

Moments That Define Them

The details that don't show up in a press release

Dan flew from his college graduation ceremony directly to Boston to close the Firefly acquisition deal with Pegasystems. He had built co-browsing software as a senior in college, sold it to a 300-person enterprise company, and was on a plane before the diploma was dry.

Nathan built the first version of Product Hunt. Not a prototype - the actual working first version of the platform that would go on to launch hundreds of thousands of products. He left before it was famous, because the next build was already calling.

Dan built BlackBerry apps as a teenager - one tracked where his friends were, another changed LED colors based on the weather - to cover his own college expenses. He had decided early that making things was more interesting than waiting for permission to make things.

Nathan's internal system at Substack for identifying emerging writers was known around the office as the "Baschez Score." He was the first employee and was already building infrastructure before the company had enough users to need it.

Dan has written openly and in public detail about the undiagnosed OCD that consumed his late twenties. Therapy, meditation, retreats - all tried, none sufficient. Medication finally helped. He published the essay. His readers replied by the hundreds with their own versions of the same story.

Every's engineers write virtually zero traditional code. The company runs ~25 people and five software products on nearly 100% AI-generated code - and runs its operations on four internal AI agents named Anton (prioritization), plus meeting notes, OKR planning, and growth tracking. This isn't a marketing claim. It's the actual architecture.

In Their Own Words

What they think about what's coming

Most new software will just be Claude Code in a trench coat.

Dan Shipper, on agent-native software

The line between writer and builder is blurring.

Dan Shipper, Every

Watch yourself and your desires, and you'll eventually build something that other people might want too.

Dan Shipper, on building products

What comes next?

Dan Shipper's guiding question for Every's strategy

AI-Native Operations

Running a company on four agents and almost no code

Most companies say they use AI. Every means something different by it. The engineering team writes almost no code manually - the company estimates ~99% of its code is AI-generated. The operations team runs on four internal AI agents built in Notion AI. These aren't chatbots bolted onto existing workflows. They're the workflow.

Anton handles prioritization. A second agent processes meeting notes. A third runs OKR planning. A fourth tracks growth metrics. Every also treats major AI systems like specialized team members - Claude, Codex, and others named "Friday" and "Charlie" are assigned tasks with the same expectation you'd give a contractor.

Agentic engineering will be the most productive new engineering discipline.

- Dan Shipper, 2026 AI predictions

This isn't a thought experiment. Every's five software products are built on this foundation: Spiral (AI writing assistance), Cora (AI email management), Monologue (AI voice dictation), Sparkle (AI file organization), and Proof (collaborative document editing). They also run Plus One, a Slack-based AI agents product for teams.

Dan's 2026 prediction: "Agent-native software architecture" means the new features of most software products are just buttons that activate prompts to an underlying general-purpose AI agent. The app isn't the product. The prompt is the product.

Every's Internal AI Stack

Anton

Prioritization - decides what matters this week

Meeting Notes

Captures, summarizes, extracts action items

OKR Agent

Plans quarterly goals and tracks alignment

Growth Agent

Monitors subscriber and revenue metrics

Friday / Charlie

External AI models treated as team contractors

Engineering code written by AI

0% ~99%

Things Worth Knowing

Fun facts and lesser-known details

Fact 01

Nathan was Substack's very first employee - he joined as they were finishing Y Combinator, before Substack was a household name in media circles.

Fact 02

Dan chose philosophy as his college major at Penn because he believed it was "the only major that would tell me how to do it" - meaning how to actually live well.

Fact 03

The name "Every" came from their original bundle called "Everything" on Substack. The name compressed; the ambition expanded.

Fact 04

Dan has been blogging since 2010. By 2025, the blog had accumulated 700,000 total readers - before newsletters were a business model and before AI was a beat.

Fact 05

Lex - Nathan's AI word processor - got 25,000 signups on its first day of public launch. It now has 300,000+ users and $2.75M from True Ventures.

Fact 06

Dan co-organized a Facebook Live telethon on Inauguration Day 2017 that raised $200,000 for three nonprofits, with 60+ celebrities and 1.5 million viewers. He called it the Love-a-thon.

Fact 07

Nathan's internal Substack writer-discovery system was dubbed the "Baschez Score" by the team. It was an early version of the kind of algorithmic talent identification that recommendation systems now run at scale.

Fact 08

Nathan's dog is named Porter. He lives in Los Angeles. Dan is in New York. Every runs across both coasts - the newsletter that covers remote-first AI work is itself remote-first.