Christopher Day

Head of Data & AI Governance @ Marriott Vacations Worldwide
Founder Daybreak Labs
My thoughts are my own

[Christopher Day @ChristopherSDay]

Maverick is live.

An open-source AI agent that works for hours, not minutes. A swarm of specialist sub-agents that plan, run in parallel, and verify their own work.

→ Your models, your machine (12 providers)
→ A hard budget cap — no surprise bills
→ A safety layer screening every step
MIT licensed. No telemetry. No paid tier.
pipx install 'maverick-agent[installer]'

https://t.co/JNUc3BpVY3

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

my invite must have gotten lost in the mail 🤷‍♂️

[Anthropic @AnthropicAI]

We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5.

We'll begin restoring access tomorrow, and will share an update soon.

We’re grateful to our users for their patience, and to everyone who worked with us on redeploying the models.

[Breanna Morello @BreannaMorello]

🚨SCOTUS RULING🚨

SCOTUS ruled that when police get your detailed cell phone location history from Google (through a "geofence" warrant), it counts as a Fourth Amendment search.

This means you have a reasonable expectation of privacy in that data.

[Daybreak Labs @Daybreakailabs]

Your company learns, why shouldn’t your platform?

[Christopher Day @ChristopherSDay]

@pmarca It’s truly sad how this has turned out. We stifled progress. I know I will get hate but this feels like the “2 weeks” when C19 happened.

[MTS @MTSlive]

SITUATION DETECTED: The Department of Energy has issued a conditional $17.5B loan commitment to finance 10 large-scale Westinghouse AP1000 nuclear reactors across the US.

[Daybreak Labs @Daybreakailabs]

Drop in the shoebox — W-2, 1099-INT/DIV/B/NEC, 1098, 1095-A.

A governed swarm extracts every figure, completes the 1040, and flags what an agent shouldn't sign off alone. Every line on a signed record.

https://t.co/DwbFxtTGkF

[Christopher Day @ChristopherSDay]

Big things!

[Daybreak Labs @Daybreakailabs]

"Build a 13-week cash-flow model for Q3 — pull the GL, flag variances over $25k, draft the board memo."

One prompt. A governed agent swarm. A reconciled deliverable, held for human sign-off.

https://t.co/DwbFxtTGkF

[Daybreak Labs @Daybreakailabs]

A first look at Lightwork.

Put AI agents to work on regulated operations: govern every action, prove it on a signed record, improve under proof.

https://t.co/DwbFxtTGkF

[Christopher Day @ChristopherSDay]

Lightwork 👀👀👀

[Daybreak Labs @Daybreakailabs]

A first look at Lightwork.

Put AI agents to work on regulated operations: govern every action, prove it on a signed record, improve under proof.

https://t.co/DwbFxtTGkF

[Daybreak Labs @Daybreakailabs]

AI agents are capable. Deploying them inside a regulated enterprise is not.

Lightwork is the governance platform and proof layer that changes that.

https://t.co/S0My8fz9E2

[International Cyber Digest @IntCyberDigest]

🚨‼️ BREAKING: The threat actor who breached Novo Nordisk, the company behind Ozempic, has leaked 264 GB of data. They are also mocking Novo Nordisk, claiming the company was using passwords like "novo123" for critical systems and that its security team sucks.

The breach includes source code, proprietary information on both marketed and pipeline drugs, clinical trial records, data on employees, doctors, and patients, and manufacturing details. The threat actor also claims it obtained private internal AI models from Novo's systems.

They allegedly breached Novo in March via a GitHub access token that let it clone the company's repositories and find additional credentials. The intrusion ran for over two months and yielded roughly 1.3 terabytes across more than 700,000 files.

[No context Hazz ⛩ @HazzimIO]

No comments.

[The Associated Press @AP]

BREAKING: The Supreme Court sides with a Texas man who says it’s not a crime for marijuana users to have guns. https://t.co/UFLlxRZiBm

[Claude @claudeai]

New in Claude Code: Artifacts.

Interactive pages built from your session, like a PR walkthrough or a living project dashboard, shared with your team at a private link.

Available in beta on Team and Enterprise plans.

[Noam Shazeer @NoamShazeer]

I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there.

It was a difficult decision to move on. I’m incredibly proud of the amazing team at Google and everything we’ve built together. It has been an honor and a pleasure to work with all of you.

[Anthropic @AnthropicAI]

We compared Claude Code success rates between occupations.

On our toughest measure of success—requiring verifiable evidence that a goal was completed, like committed code—every field was within 7 percentage points of software engineering.

[Christopher Day @ChristopherSDay]

@chrisjenx @0xgaut

[Anthropic @AnthropicAI]

Our latest economic research introduces a framework for tracking Claude Code as it scales.

Who is using Claude Code, and what are they using it for? How is the value of tasks changing? And how much does domain expertise shape whether a session succeeds?
https://t.co/IjjwQvrESo