D3 Daily — Wednesday, September 23, 2026

OpenAI and Anthropic shipped three frontier models on the same day and turned the race into a price war. Here is what actually matters from today's feed.

Today’s Highlights

GPT-6 Sol vs. Luna vs. Claude Opus 5.5: which should you use?

The Neuron AI

Three major models landed on the same day, which means the real question is not which one wins a benchmark. It is which one fits the workflow in front of you. A practical breakdown of where Sol, Luna, Opus 5.5, and Astra each earn their keep.

Opus 5.5 vs. GPT-6 Sol: which model won my blind taste test?

Lenny’s Newsletter
Inspiration

Blind evals across writing, frontend prototypes, long-running agentic tasks, SVGs, coding, and video editing. Then Barbie Bench, because of course. The value here is not the verdict, it is watching someone stress-test models against real work instead of vibes.

I left Claude for months. Opus 5.5 is why I’m back

Lenny’s Newsletter
Inspiration

A candid account of abandoning a tool, then coming back when it stopped fighting the user. The specifics on frontend, SVGs, and agentic tasks are worth your time if you are deciding where to spend your model budget.

The 3 roles of context for AI agents

Sidebar.io
Inspiration

Power users curate three kinds of context: global across tasks, local to the task, and ambient raw streams like email. A clean mental model for anyone designing agent experiences or just trying to get more out of the ones they already use.

Audit your agent files

Sidebar.io
Inspiration

A practical guide to figuring out what your coding agent actually still needs. Most agent configs accumulate cruft the same way design systems do. This is the cleanup pass.

Advanced evals: how to find and fix hidden AI failures in your product

Lenny’s Newsletter
Inspiration

Why you should never skip error discovery. The failures that hurt are the ones you never see, and this walks through how to surface them before your users do.

Why the best resume may be the wrong resume for a design team

UX Collective

What design leadership gets wrong about experience, expertise, and innovation on a team. A strong resume signals individual output. A strong team needs complementary gaps, and hiring for the former often costs you the latter.

In the age of AI, the UX field survives on leaders who cultivate juniors

UX Collective

The pipeline problem is not abstract anymore. If senior designers stop investing in juniors because AI makes the work faster, the field quietly hollows out from the middle. Worth reading alongside the resume piece above.

Designing with AI

Heydesigner
Inspiration

A roundup worth skimming: what you don’t build, 100 UX methods ranked by value, the rules travel but the eye doesn’t, and design tokens needing more than semantics. The tokens piece alone is worth the click.

Quick Hits