Why I cancelled my newspaper subscription

I read a daily paper of record from age 22 to age 37. Print at first, then digital. Every morning, with coffee. I considered it part of being a serious adult. I subscribed through layoffs, paywalls, redesigns, the rise of social media, the rise of substack, the rise of podcasts. I cancelled in 2026.

The paper didn't get worse. If anything it got better — the reporting on AI and the new energy economy was sharper than it had been in 2015, the data journalism was world-class, the columnists were strong. My needs changed, and the format didn't.

What I actually wanted

When I'm honest, what I needed from a news subscription in 2026 was a specific list of things kept up to date: my company's top competitors, the regulator we sell into, two countries I have personal stakes in, three publicly-traded names I hold, a court case relevant to my industry, and one cultural beat I follow for fun. Maybe twelve topics, total.

What the paper delivered every morning was a curated selection of forty stories the editors decided I should care about, fifteen of which I did and twenty-five of which I didn't. The editors are excellent at what they do. But their curation isn't mine. They're not paid to know what twelve things matter to me specifically.

What started not working

The breakpoint was when I noticed I was missing things on my list. A competitor announcement broke at 11am on a Tuesday in a sector trade publication. It was covered by the paper at 7am the next morning, embedded in a roundup of three other adjacent stories. By the time I read it, I was sixteen hours behind. Not catastrophic — but a consistent pattern.

The paper was a fine briefing on the world. It was a poor briefing on the specific list of things I needed to know about. Those are two different products, and I was paying for the wrong one.

What replaced it

I started with RSS readers — they don't replace newspapers, they replace newspapers' aggregating function. Feedly worked for a while. I built feeds around each topic. The problem with the RSS approach is that you get every article from every outlet that mentions your topic, including duplicates and rephrases. You become an aggregator of an aggregator.

Then I switched to Sentinel. The pitch is that you hire an AI journalist on each topic and they cover it indefinitely. The AI journalist reads thousands of outlets, cross-references every claim against multiple independent sources, and files a dispatch the moment the topic actually moves. You don't get every article; you get verified state changes.

I run twelve AI journalists. Twelve topics, covered indefinitely. The morning briefing is two minutes. I haven't missed a development on any of them in three months.

What I kept

I didn't cancel reading. I cancelled the subscription that didn't fit my needs.

• I still pay for the FT and a substack I love, both for the kinds of essays a beat reporter doesn't write.
• I still listen to a daily news podcast — for the framing, not the facts.
• I still occasionally read long-form magazine pieces because they're the best argument-shaping format that exists.

What I gave up was the daily paper-of-record subscription, which had become a habit defending its own existence. The paper does good work. It just isn't the right tool for the job I actually have, which is to never miss anything on my twelve-topic list.

You might be in a different situation

If your job involves keeping up with the world broadly — politics, foreign affairs, culture — a paper of record is still the best instrument. If your job is to track a specific list of named things and never miss them, you probably need an AI journalist per topic instead.

Most of the people I know who used to read a paper of record daily are in the second category. Most of them are still subscribed to the first kind of product. The mismatch is what made me write this.

Related: What is an AI journalist? · How I read the news in 5 minutes · Sentinel for founders.