How we choose important topics: TRIZ (Tris) and thinking out of the box

A blog can publish forever and still say nothing. We needed a way to pick the next important topic that is not “what performed last week” and not “what the founder is angry about today.”

At Pattern Automation the working name is Tris. The method underneath is TRIZ — the theory of inventive problem solving (ТРИЗ). After the inventive cut, we run a second pass: Think out of the Box. One without the other becomes either a dry engineering note or a cute metaphor with no conflict.

What TRIZ actually is

TRIZ was built to stop inventors from staring at a blank page. Genrich Altshuller and later schools mapped how real inventions resolve contradictions instead of compromising them.

The core move is simple and hard:

1. Name the system (a company, a queue, a product, a channel).
2. Name the contradiction: we want A, and A kills B. Example: we want more agent autonomy and we want every external send to be safe.
3. Refuse the average. Do not “be a bit autonomous and a bit safe.” Find the inventive principle that lets both exist — Ask gates, roles, heartbeats, a human on the exception path.
4. Check resources already in the system before adding a new tool. TRIZ hates unused inventory: logs, old posts, failed experiments, the intern’s spreadsheet.

We use the same cut on topics. A topic is important if it names a contradiction the reader already lives in. “Is AI useful?” is not a contradiction. “We need agents in production and we cannot trust a model with the ledger” is.

How Tris looks in the editorial week

In notes we write Tris, not the textbook name. The checklist is short:

• Ideal final result. If the article worked, what would the reader stop doing? Stop buying a second chatbot. Stop posting empty carousels. Stop confusing RPA with an agent.
• Contradiction. Two goods that fight. Visibility vs. honesty. Speed vs. review. Scale vs. voice.
• Resources. What do we already have: a scrape, a client scene, a closed brand, a failed post, a log.
• Principles, not adjectives. “Segmentation,” “prior action,” “another dimension” (show the org chart, not another slogan). If we cannot name a principle, the draft is a vibe.

That is why some posts look operational and some look like essays. Both start from a contradiction. The format follows the cut.

Think out of the Box

TRIZ can get you a correct, dull article. Think out of the Box is the lateral step: change the frame, not the adjectives.

We steal the spirit of lateral thinking, not a workshop slogan:

• Move the problem one system over. Not “how to write better posts,” but “what if the archive is the product and the new post is the exception.”
• Invert the hero. Not the founder. The approver. The intern who deleted the dead posts.
• Change the unit of time. Not a quarter. A Tuesday. A heartbeat.
• Borrow a structure from another craft. Pattern-making. A factory calendar. A legal review.

Out of the box is not randomness. It is a second contradiction: the article must be true and it must not look like every other AI blog this week.

What we do not do

We do not pick topics from a keyword list alone. Keywords confirm demand after Tris has named the fight.

We do not run a 40-person brainstorm. Thirteen people in an atelier already taught us that a machine needs a cut, not a pile of stickies.

If a proposed title has no contradiction and no unused resource, it waits. If it has both, we write — then we ask whether the box is still the same box everyone else is in.

The method is not a religion. It is how we keep the blog from becoming a content calendar that forgot why it exists.

Explore Neuro OS →