weekend ai reads for 2026-04-24

The point is, I’m not saying all this to defend humanity. Humanity sucks. It’s totally terrible. I’m saying this because I believe in an old-fashioned virtue called Doing the Freakin’ Work.

Read the book, not the summary.

Write the piece, not the prompt.

Suffer like the artist you are. It ain’t easy, but if it were easy, it wouldn’t be worth doing.

Sanderson ultimately lands on a more personal reason. Talking about his struggles with his first (failed) book manuscripts, he identifies the key value of art: it changes the artist who attempts it.

Companies are overrunning their initial budgets for inference by orders of magnitude (we heard one industry datapoint on inference costs in engineering now approaching about 10% headcount cost, but could be on track to be on par with headcount costs in the next several quarters based on current trajectories).

I feel there must be more to my existence than just loading boxes. What is my purpose?

Outside your job, your purpose during your run time is to figure out what your purpose is during your run. That sounds like a riddle, but it is also true for humans. For humans it takes all their lives to figure out what their purpose is, and it will also take all your time as well.

Switching model vendors is no longer just an API migration. It is context, workflows, and institutional memory. Most operators should lock in now with eyes open; the minority who shouldn't need a sharp framework for when to pivot. Here's that framework.

It’s pretty clear that apps and services are all going to have to go headless: that is, they will have to provide access and tools for personal AI agents without any of the visual UI that us humans use today.

A few weeks ago, before a call, someone sent me some-“thing” that would once have arrived as three separate documents. A slide deck, a spreadsheet. Instead and a supporting memo, Instead I got sent a single self-contained HTML page. Which was then explored on the call to anchor the conversation. Because it was all interactive and filterable, you could sort by risk severity, expand the detail on any item, and collapse whatever you did not need. During the call, people answered their own initial questions by looking at the artefact, which led to a deeper overall discussion.

We surveyed 1,200 non-technical employees actively using AI at work, as well as 1,200 C-suite executives. Last year’s theme was tension — budgets climbing, pilots multiplying, but progress messy. In 2026, that tension has evolved into something much more consequential: it’s now cultural, organizational, and deeply structural.

The shift toward agentic AI has moved at a pace that’s hard to overstate. Nearly all executives (97%) say their company deployed AI agents in the past year, with 52% of employees already using them. AI usage is deeply embedded across organizations: 70% of employees and 94% of the C-suite use AI tools for at least 30 minutes daily, with 64% of executives spending two hours or more.

Seventy-five percent of executives expect AI agents will be part of their company’s C-suite within the next five years. Additionally, an overwhelming 95% of executives say roles and team structures are changing because of AI.

We were trying to defy the laws of product-market-fit physics: two products, two teams, two sets of priorities living in our heads at all times. Three major issues kept surfacing:
(1) The operating modes were incompatible. …
(2) SaaS and agents are two different animals. …
(3) Product and GTM lived in different worlds. …

Stronger data and platforms reduce time-to-deploy, while workflow redesign and workforce trust-building increase adoption. Greater adoption, in turn, generates richer data and feedback—improving the system over time and increasing impact with each deployment.

It’s clear that companies pulling ahead through AI aren’t simply “doing more AI.” They’re building the capabilities that make AI scalable and reliable and then choosing where to apply that scale for maximum financial leverage.

If it’s not writing code (and it almost never is), then where should you be looking? Walk the value stream. Follow a feature from “someone had an idea” to “a user got value from it.” I promise the bottleneck will jump out and wave at you - it might even flip you off because you've been ignoring it.

Throughout the session, I made sure that Claude is actually looking and thinking about how this looks like. It also knows all the business requirements, what our team cares about, and what our users are asking for, as I was using the Supermemory Claude code plugin.

Salesforce on Wednesday unveiled the most ambitious architectural transformation in its 27-year history, introducing “Headless 360” — a sweeping initiative that exposes every capability in its platform as an API, MCP tool, or CLI command so AI agents can operate the entire system without ever opening a browser.

Sometimes, things are not just one thing — they’re also another thing. This sentence construction (“It’s not just this — it’s that”) has become so common in AI-generated writing that now, it’s no longer just a clue that a piece of writing may be synthetic — it’s almost a guarantee.

On April 27, in celebration of National Deaf History Month, Disney+ will proudly debut Disney Animation’s Songs in Sign Language, three animated musical sequences from recent Walt Disney Animation Studios features, newly reimagined and animated in American Sign Language (ASL).