SpatzAI: Building a Category Before the Category Exists

I recently saw a post about how Taiwan became the centre of the world’s semiconductor industry.

The simple version is this: Taiwan did not begin with a semiconductor industry. It began with a plan, selected people, serious training, and time to learn.

In 1974, Taiwan sent 19 hand-picked engineers to RCA in America to learn how to build chips. They had no mature industry, no deep experience, and no category to slot neatly into. What they had was a foundational bet.

Years later, that learning compounded into TSMC, now one of the most important companies in the world, producing some 70% of the world’s advanced chips.

I think there is a lesson here for us at SpatzAI.

SpatzAI is not entering a mature category either. It sits in a gap between minor workplace friction, conflict resolution, psychological safety, anonymous incident reporting and organizational politics.

Most systems activate too late, after behaviour has escalated, repeated, hardened, or become political.

SpatzAI is trying to work earlier, before the damage is done.

So, we figure the first step is not mass adoption. The first step is training a small number of disciplined teams (let’s say 4 teams of 5) to recognize that cautioning, objecting to, and resolving micro-conflicts early is a valid way to address behavior they find objectionable.

These first teams are not just test users. They are potential category builders, embarking on a voyage of discovery, an experiment that could change the way teamwork is self-governed, globally.

They will help define the missing space between everyday disagreements and formal safety intervention.

That space could be called “real-time micro-conflict moderating“.

The aim is not to blame or punish infringing team members. It is to find the fault or flaw in the behavior early, fix the interaction, repair and resolve their minor spat, and learn from the patterns created moving forward.

Over time, SpatzChat™ and the Spatz (Team + AI) Review platform can capture the data and evaluate the following:

• What mainly triggers cautions, (we think it is dogmatic and absolute thinking and language).

• What prevents escalation, (we think an early acknowledgement, simple apology or an acceptable apology)

• What repairs trust (we think a robust discussion of converging and diverging opinions and fair accountability. Trust in the system first and everyone abiding by the ref’s decision, regardless of the outcome).

• Which patterns predict future team success or failure (we think it is related to how quick they can resolve their behavioral differences).

• How long it takes to resolve a spat, (we think the earlier the better but also fair and just accountability).

And that is how we think we can form a new category in teamwork over time.

Not from a slogan, but from a structured philosophy and selected teams that understand and agree to give Spatz a try. Using disciplined practice, useful data, and compounded learning by team members and the Spatz LLM.

SpatzAI will not be for every team. We believe it will be for teams disciplined enough to be wishing to be held to account through transparency, course-correction using a real-time intervention, and willing to take advice from a well trained team and AI.