What does tech news actually cover? Not the headlines you remember — everything. This study analyzes all 443 articles Calder Brief published between September 30 and October 14, 2026, the site’s first two weeks of full operation. Every number below was computed by script from the site’s own publishing corpus: titles, excerpts, sections, dates, and body text. No surveys, no estimates, no vibes.
The short version: it’s a firehose, and it points in a few very clear directions. The longer version is below, with charts.
Key findings
Key findings
- 443 articles in 15 days. Across the four full publishing days (October 10–13), Calder Brief averaged 91 stories a day — roughly four an hour — with a single-day peak of 97.
- AI is nearly a third of everything. 132 of 443 stories (29.8%) match a fixed AI keyword rule applied to headlines and excerpts — about one in three stories published.
- Money and launches dominate. Funding, investment, and M&A appear in 23.7% of stories; product launches and announcements in 23.5%. Together, nearly half of all tech coverage.
- Microsoft and Meta lead the company leaderboard with 12 headline mentions each, ahead of Apple and Nvidia (9 each) and SpaceX (8).
- The median story runs 1,071 words — these are full reported pieces, not blurbs — across 23 sections, led by Science (35 stories), Technology (32), and Business (26).
Journalists: all figures are quotable with attribution to “Calder Brief Data Desk.” The full dataset and methodology are at the bottom of this page.
The cadence: 91 stories a day
Calder Brief launched quietly at the end of September — one story on September 30 — then ramped through the first week of October before switching to its current cadence: four researched articles an hour, around the clock, starting October 10. The result is visible in the daily counts:
Articles published per day, September 30 to October 14, 2026. *October 14 is a partial day (count as of the morning analysis run). Source: Calder Brief publishing corpus.
The four full days at the current cadence — October 10 through 13 — produced 365 articles, an average of 91.25 a day. That pace held steady across all four days (77, 96, 95, 97), which suggests it’s systematic rather than a burst.
What gets covered
Coverage spans 23 sections. Science leads with 35 stories, followed by Technology (32) and Business (26) — but the long tail is the real story: even the smallest sections (Smart Home, Repair, Work Tech, Climate Tech, Future of Food) each got 15 stories in two weeks, about one a day.
Stories per section, top 12 of 23 sections. The full 23-section table is in the dataset section below. Source: Calder Brief publishing corpus.
No single section accounts for more than 7.9% of output. Whatever “tech news” is, on this site it isn’t one beat — it’s two dozen of them running in parallel.
The beats: money and launches
Sections say where a story lives. Beats say what it’s about — and here the distribution is much more concentrated. Classifying stories by keyword rules (a story can match more than one beat):
Share of the 443 stories matching each beat’s keyword rule. Beats are not mutually exclusive — one story can count in several — so shares do not sum to 100%. Source: Calder Brief publishing corpus; rules in methodology.
Funding, investment, and M&A touch 105 stories (23.7%) — rounds raised, valuations, IPOs, acquisitions. Product launches and announcements touch 104 (23.5%). Combined, that’s 209 of 443 stories: nearly half of all tech coverage is either money moving or products shipping. Policy and regulation is a distant third at 70 stories (15.8%), followed by science results (31), security incidents (21), and earnings (15).
AI’s share of the news
Apply a fixed keyword rule — AI, artificial intelligence, LLM, GPT, Claude, Anthropic, OpenAI, chatbot, machine learning, neural, deep learning, generative AI, agentic, foundation model, transformer, diffusion model — to every headline and excerpt, and 132 of 443 stories match: 29.8%.
That is roughly one in three stories. It cuts across sections: not just the Technology desk, but Science (AI-designed materials, protein models), Business (funding rounds), Semiconductors (AI chips), and Tech Policy (AI regulation). Whatever the beat, there’s a good chance AI is in it.
For calibration, the next most common headline theme words are far behind: “quantum” appears in 20 headlines, “battery” and “chip” in 11 each, “robot” in 8. AI isn’t just the biggest topic — it’s an order of magnitude bigger than the next one.
Headline occurrences of selected theme words (word-boundary match, titles only). Source: Calder Brief publishing corpus.
The company leaderboard
Which companies does tech news actually name? Counting word-boundary matches against a fixed list of major companies in headlines only:
Headline mentions, top 10 companies. Word-boundary match on titles only; fixed company list (see methodology). Source: Calder Brief publishing corpus.
Microsoft and Meta tie at 12 headline mentions each, followed by Apple and Nvidia at 9, SpaceX at 8, and Google at 7. The AI story shows up here too: four of the top six (Microsoft, Meta, Nvidia, Google) are companies whose 2026 news cycles revolve around AI infrastructure and models — plus Anthropic (3) and OpenAI (2) further down the list. Sony, Samsung, and Amazon round out the top ten at 5 each; NASA, the only non-company in the top ten, also lands at 5, a mark of how much space coverage this fortnight carried.
How long is a story
Body text runs from 343 to 1,268 words, with a median of 1,071 words and a mean of 1,027 — a tight distribution around roughly a thousand words. Only 3.6% of output (16 stories) is Opinion; the rest is reported news, explainers, and analysis. This isn’t an aggregation feed: the unit of coverage is a full reported piece.
Corpus: September 30 to October 14, 2026 (October 14 partial). Source: Calder Brief publishing corpus.
The full dataset
Every story in the corpus, grouped by section. Counts sum to 443.
| Section | Stories | Share |
|---|---|---|
| Science | 35 | 7.9% |
| Technology | 32 | 7.2% |
| Business | 26 | 5.9% |
| Energy | 22 | 5.0% |
| Culture | 21 | 4.7% |
| Future Flight | 21 | 4.7% |
| Tech Policy | 20 | 4.5% |
| Immersive | 20 | 4.5% |
| Space | 19 | 4.3% |
| How the Internet Works | 19 | 4.3% |
| Robotics | 18 | 4.1% |
| Quantum | 18 | 4.1% |
| Semiconductors | 17 | 3.8% |
| Opinion | 16 | 3.6% |
| Cybersecurity | 16 | 3.6% |
| Gaming | 16 | 3.6% |
| Accessible Tech | 16 | 3.6% |
| Self-Hosted | 16 | 3.6% |
| Smart Home | 15 | 3.4% |
| Repair | 15 | 3.4% |
| Work Tech | 15 | 3.4% |
| Climate Tech | 15 | 3.4% |
| Future of Food | 15 | 3.4% |
Dataset: Calder Brief article corpus, September 30 – October 14, 2026. Computed October 14, 2026. Released under CC BY 4.0 — reuse with attribution to “Calder Brief Data Desk.”
Methodology
The corpus is Calder Brief’s own publishing data: the article records in the site’s generator (_generator/articles*.py), each carrying a slug, title, excerpt, section, publication date, author, and full body text. The analysis ran on October 14, 2026, over 443 articles dated September 30 to October 14, 2026. October 14 is a partial publishing day (the run happened in the morning); the four complete days at the current cadence are October 10–13.
- AI share: a story counts if its title or excerpt matches the case-insensitive regex
\b(ai|artificial intelligence|llm|gpt|claude|anthropic|openai|chatbot|machine learning|neural|deep learning|generative ai|agentic|foundation model|transformer|diffusion model)\b. 132 matches. - Beats: independent keyword rules on title + excerpt. Funding:
funding|raised|series [a-d]|ipo|valuation|investment|acqui|merger|billion|million. Launches:launch|unveil|announc|introduc|debut|ship|release|unbox. Policy:regulat|policy|law|ban |sanction|tariff|antitrust|lawsuit|sue|fined|fcc|ftc|eu |congress|senate|white house. Security:hack|breach|ransomware|zero-day|vulnerab|malware|phish|ddos|leak. Science:study|researcher|discover|telescope|trial|clinical|fda|peer-review. Earnings:earnings|revenue|quarter|q3|q4|profit. A story may match several beats. - Company mentions: word-boundary regex matches on titles only, against a fixed list of ~55 major companies and agencies (Apple, Google, Microsoft, Meta, Nvidia, SpaceX, etc.). Counts are headline appearances, not endorsements.
- Theme words: word-boundary matches on titles only.
- Length: word count of the article body text as stored in the corpus.
Limitations: the corpus is one publication’s output over two weeks, not a sample of “the media” — it measures what Calder Brief covered, which reflects both the news cycle and the site’s section-driven publishing cadence. Keyword rules are deliberately simple and fully disclosed so anyone can reproduce or challenge them.
Sources
- Calder Brief publishing corpus,
_generator/articles.pythrough_generator/articles5.py— 443 article records (slug, title, excerpt, section, date, body), September 30 to October 14, 2026. - Analysis script:
_generator/build_study_coverage_2026.py(reproduces every count on this page).
0 Comments