What I believe.
A running list of ideas I have conviction in.
Creativity can be engineered.
New ideas are really combinations of old ideas. The right AI agent harness - combined with a vetted, high-signal repository of artifacts - can have its own "taste". The fact that LLMs are probabilistic instead of deterministic supports this.
Sprints + docs are compounding mechanisms.
This was true before AI - and they became 10x more valuable with AI. Organizations that sprint, crystallize learnings, and iterate with AI - will lap those that don't.
Focus always wins.
If you're splitting bandwidth between multiple problem areas, you won't be able to compete with someone obsessing in one of them. Attack one thing at a time. (I recognize the irony of a "full stack" operator saying this. Can't hide from the truth.)
Optimize for usefulness.
If a piece of information isn't useful for what you're actively working on - discard it. There's no shortage of information. There's a massive shortage of time.
Be aware of the macro - but don't fixate on it.
Is your paid media performance down because of wars, or the economy, or AI sentiment? Maybe. But that isn't useful. Focus on the levers under your control. There's always a way to win.
Brands that can absorb the highest CAC will win.
Those focused purely on frontend performance - with zero levers in their unit economics or SKUs - are in for a bad time.
Ad creative is a local maximum.
It's capped at the subset of the TAM that your offers unlock. The intersection of your offers and TAM are the global maximum.
Mass AI burnout is coming.
It will touch every market. AI will go from a selling point - to something that causes aversion and hurts conversion. AI components of an organization will need to blend seamlessly with the human parts. Your clients/customers shouldn't notice the difference, or it will hurt your outcomes.