Artanis' Substack

Evals Without Engineering

How to stop playing whack-a-mole with your AI

Apr 24 • Sam Miller

Your AI is only as good as your Policy

You need to be both i) consistent and ii) unambiguous

Apr 15 • Sam Miller and Olly Styles

1

March 2026

Alignment will be all that matters

The main skill in building AI will be writing Policy

Mar 17 • Sam Miller and Olly Styles

3

1

AI is Still a Bubble

But coding agents are the real deal

Mar 3 • Sam Miller

3

1

1

October 2025

Stop Evaluating Unstructured Text

Find the Hidden Structure Instead

Oct 2, 2025 • Sam Miller

August 2025

The Domain Expert <> AI Engineer Relationship

The Bedrock of AI Project Success

Aug 15, 2025 • Andrew Manderson

1

July 2025

Why “Bikeshedding” kills so many AI projects

And what to do about it

Jul 30, 2025 • Olly Styles

2

The Easy Way vs The Hard Way in AI

The Easy Way is quick, but rarely works in practice

Jul 24, 2025 • Sam Miller

4

1

90% accuracy often brings no value

Finding out what “good enough” looks like is the key

Jul 16, 2025 • Sam Miller

1

Traction Won’t Help You Raise

But Belief in Yourself Will

Jul 8, 2025 • Sam Miller

2

2

March 2025

Don't Write Prompts. Write Policies.

Breaking free from prompt engineering

Mar 21, 2025 • Olly Styles and Sam Miller

2

How Fundraising Actually Works

Stop publicising your raise!

Mar 12, 2025 • Sam Miller

12

1