trey causey dot com - Blog Posts

You're not taking AI overhangs seriously enough

How you, yes you, can help build a better future with AI.

AI can revive the middle class business model

Imagining a world look like where more people can just do things

Why LLM hallucinations are primarily a UX challenge rather than an algorithmic one, and how design improvements can mitigate their impact on users.

A practical guide to building responsible AI teams that succeed by aligning with business goals, hiring builders over framework-creators, and focusing on empirical risk management.

Beyond doomers and accelerationists: why skeptical optimists who believe AI problems are solvable with effort are essential for building an abundant AI future.

Why business books and productivity advice often commit the error of selecting on the dependent variable, and how to spot this common reasoning flaw.

Why new systems that afford quantitative measurement face disproportionate scrutiny compared to incumbent systems, and why we should hold all systems to the same standards.

The emergence of a new kind of product manager who understands data infrastructure, ML, and how to build products where data is at the core of the value proposition.

Notes from my own job search on what to expect in data science interviews, from technical screens to on-site interviews, and how to prepare for both type A and type B data scientist roles.

Reflections on the emotional toll of producing public-facing work like open source or writing, where criticism flows fast while rewards reveal themselves slowly.

Guidelines for requesting informational coffee chats that respect the other person's time and increase your chances of getting a response.

Essential software engineering skills that data scientists need to collaborate effectively: version control, code review, testing, documentation, and writing modular code.

The natural affinities between data science and product management, plus the challenges data scientists face when making the transition to PM roles.

A better approach to hiring data scientists: treat candidates as intelligent humans, use homework assignments over whiteboard coding, and build a process that leads to equitable outcomes.

A guide to getting started in data science, covering the essential foundations in math, statistics, experiments, machine learning, and tooling that aspiring data scientists need.