Did you recently see an ad for Roku Ads Manager in a newsletter? We’re running a short brand lift survey to understand what’s actually breaking through (and what’s not).

It takes about 20 seconds, the questions are super easy, and your feedback directly helps us improve how we show up in the newsletters you read and love.

If you’ve got a few moments, we’d really appreciate your insight.

Take the survey.

Anthropic dropped Claude Fable 5 today. It's the same underlying model as Claude Mythos 5, the frontier research model they've kept locked behind closed doors, now made available to the public with safety guardrails baked in.

The benchmarks back it up across every category that matters to builders.

In software engineering, Stripe used it to compress months of engineering work into days, migrating a 50-million-line Ruby codebase in 24 hours that would have taken a full team two months by hand.

In knowledge work, it scored highest of any model on Hebbia's senior-level finance benchmark. In vision, it can rebuild a web app's source code from screenshots alone.

In long-context tasks, it stays focused across millions of tokens and actively uses its own notes to improve outputs.

In life sciences, internal teams using Mythos 5 accelerated parts of the process by around ten times, with the model independently executing tasks normally done by a scientist.

The longer and more complex the task, the bigger its lead.

The safeguards are interesting. On a narrow set of topics like cybersecurity, biology, and chemistry, Fable 5 routes to Opus 4.8 instead. Users are told when that happens. It affects less than 5% of sessions on average. A separate version, Mythos 5, is available to a small group of cyber defenders and critical infrastructure providers with some of those guardrails lifted.

This changes what's possible for builders.

We've been testing it all day and it's noticeably better. It does eat through sessions faster than Opus though, so be intentional.

Treat it as a thought partner, not an answer machine. Don't take the first response and run with it. Brainstorm with it, make it play devil's advocate, spin up sub-agents with different perspectives and let them debate. These models still lean sycophantic, so push back.

Have it interview you before you build anything. We use a skill called Grill Me that fires 15 to 30 questions and pulls everything out of your head into a working doc. Tell a fresh setup to run Grill Me on your business. It's how we planned this entire issue.

Then verify its own work. Tell it to check the result as a beginner, a software engineer, and a business owner. Open a Playwright browser and have it click around. That's the difference between a 70% first pass and a 92% one.

AI changed what senior engineering means. Gauntlet is built for engineers who want to lead that change — not watch it. The most effective way to become AI-first. No tuition. No tutorials. Just the work. Apply now.

Cohort is filling up now

_{Must be a US citizen to qualify.}

⏱️ 30–45 min | 🔧 YouCam API account (use code ytmakerthrive for 500 free units), basic coding skills

Why build this? $816 billion in goods get returned every year. Most of it comes down to one thing: people couldn't visualize it on themselves. The YouCam API solves that. It's the same tech used by major fashion brands, now open to any builder with pay-as-you-go pricing. E-commerce tools, creator apps, styling tools, uniform configurators—the surface area is huge.

Steps:

What you get: A virtual fitting room you can drop into anything. Face and proportions stay accurate. Fabric and texture render realistically. Region targeting means you can build mix-and-match tools, not just basic swaps.

Full tutorial:

VC Boom

Score your deck, meet investors who fit, and raise more

Built by an 8-year VC who raised hundreds of millions across 47 startups.

Here's what it does:

Upload your deck and get a score across 7 investor dimensions in under 90 seconds, plus the one fix that moves the needle most. Then it matches you with 100 investors from a network of 47,000+ who are actually active in your space, with a one-line reason each one fits. From there, it drafts personalized cold emails that go out from your own inbox.

Founders using it have raised $95M so far. Free to start, no subscription.