In partnership with
Did you know that we have LinkedIn, Instagram and X accounts
that you can follow?
Hey everyone,
A little update before we get into today’s issue: TIW has a refreshed format. We’ve made a few changes to make the newsletter easier to read, quicker to scan, and more useful, while keeping the same AI, tech, and productivity content you’ve come to expect.
You’ll now find a Deep Dive for one story worth spending a little more time on, Quick Hits for the news you can catch in a minute, Tool Snapshots with a little more context on why each tool is actually worth knowing. And of course, AI Made This for the interesting things people are creating with AI.
Same TIW, just a fresh way to bring you the good stuff.
We’d love to know what you think! Is the new format working for you? Is there anything you’d like us to add, change, or cover more of in future issues? Just hit reply and let us know, we’d love to hear from you.
In this issue:
🤿Deep Dive: Claudes new model takes on Science
🤝Our Sponsor: Turn ideas into clearer roadmaps
⚡Quick Hits: The latest in AI, tech, and productivity
⚒Tool Snapshots: Tools for AI, no-code, and productivity
🤖AI Made This: Interesting and inspiring creations made with AI
🤿 DEEP DIVE
Claude’s New Model Isn’t Just Better at Coding. It’s Designing Proteins and Mapping Venus.
Anthropic’s new Claude Fable 5.1 is being pitched as a major upgrade for coding and knowledge work. But some of the more unusual examples have little to do with writing software.
In testing, the same underlying model designed high-affinity protein binders, created a more detailed elevation map of Venus, and sped up computational biology models by as much as 2.5x.
That’s the part of this release that’s easy to miss.
One model, two versions
Claude Fable 5.1 and Claude Mythos 5.1 are actually the same model. The difference is in their safeguards and who can access them.
Fable 5.1 is generally available. Mythos 5.1 has more permissive safeguards for vetted cybersecurity and life sciences professionals and is currently available through trusted access programs.
Anthropic says the model sets a new performance frontier for coding, knowledge work, and long-running problem-solving. It also says lower-effort settings can achieve results similar to or better than Fable 5 at a lower cost.
One example came from investment firm Millennium. According to Anthropic, Fable 5.1 identified the cause of a rare crash in the firm’s internal systems that its engineers and other models had been unable to explain after several years of trying.
But the scientific experiments show a different side of the model.
For molecular design, Mythos 5.1 was given open-source protein design and folding tools. Its designs were then experimentally tested by external organizations. Across three targets, its binding affinities were 10x higher than the best designs submitted to Adaptyv Bio’s protein design competitions. Across 12 targets, nearly 50% of its designs were viable binders, compared with the 10–15% hit rates which are typical today.
Fable 5.1 also used decades-old radar images from NASA’s Magellan mission to train a neural network and produce a new high-resolution elevation map covering a third of Venus. The resulting map shows details down to 2 to 3 kilometers rather than 10 to 20 kilometers, with heights up to 25% more accurate than before.
And in computational biology, Mythos 5.1 rewrote parts of 7 open-source deep learning models to make them run faster. The resulting speed-ups reached 2.5x, while preserving identical outputs, with estimated GPU cost reductions of 30–60% on certain analyses.
There’s an important distinction, though. Access to Mythos 5.1 is restricted because its capabilities also raise biological and cybersecurity risks. Anthropic says its evaluations found the model remains below its next defined biological risk tier and within the lower cyber-risk category of its Frontier Compliance Framework.
Fable 5.1, meanwhile, comes with more precise safeguards. Anthropic says its cyber safeguards produce around 60% fewer interventions per Claude Code session than the previous Fable 5 safeguards, while the model can now be used to identify software vulnerabilities but not generate exploits.
Image credit: Anthropic
The quick take
Fable 5.1 and Mythos 5.1 are the same underlying model with different safeguards and access.
Scientific tests included protein design, mapping Venus, and optimizing computational biology models.
Fable 5.1’s cache reads now cost 75% less, bringing estimated typical workload costs down around 25%.
Mythos 5.1 remains restricted to vetted cybersecurity and life sciences users.
Anthropic says its safety testing found improvements over Mythos 5, although some alignment limitations remain.
🤝 OUR SPONSOR
AI made PMs faster. Multiplayer mode is still broken.
A PM can summarize research, draft a PRD, and mock up a prototype before lunch. The hard part starts when the team has to decide what actually gets built.
Jira Product Discovery gives product teams one place to capture insights, prioritize ideas with consistent frameworks, and build living roadmaps stakeholders can rally around.
And because it’s connected to Jira, the context behind every decision stays with the work—so developers and their agents know not just what to build, but why.
AI helps PMs move faster. Jira Product Discovery helps the whole team build with confidence.
⚡ QUICK HITS
The latest in AI, tech, and productivity worth knowing
Meta’s new voice model knows when to listen - Meta introduced Muse Voice Transcribe, a real-time speech model that handles transcription, speaker identification for 20+ speakers, endpointing, and multilingual code-switching. One interesting trick is “adaptive delay”: the model dynamically decides how long to listen before transcribing each word, trading a little extra waiting time for better accuracy when speech is harder to understand.
OpenAI’s next model comes with a cybersecurity warning - OpenAI says its forthcoming Astra model can discover and exploit previously unknown software vulnerabilities without human guidance, making it the first of its models to cross the company’s “critical cybersecurity threshold.” OpenAI says Astra will be available soon, but access to its most advanced cybersecurity capabilities will be restricted.
⚒ TOOL SNAPSHOTS
Futuristic tools within AI, no-code, and productivity
Track GPU rental prices with a reproducible market index.
Why it’s useful: Gives anyone a consistent USD-per-GPU-hour benchmark based on published on-demand rates from a fixed provider panel.
Work toward personal or professional goals with a lifelike coach.
Why it’s useful: Combines coaching with simulations, videos, and roleplays, making it useful for individual learners as well as enterprise training.
Search across thousands of podcasts like you search the web.
Why it’s useful: Makes podcast conversations easier to discover by searching 130,000+ actively transcribed shows and their growing library of episodes.
🤖 AI MADE THIS
Interesting and inspiring creations made with AI
Created with Midjourney
ℹ️ ABOUT US
The Intelligent Worker helps you to be more productive at work with AI, automation, no-code, and other technologies.
We like real, practical, and tangible use-cases and hate hand-wavy, theoretical, and abstract concepts that don’t drive real-world outcomes.
Our mission is to empower individuals, boost their productivity, and future-proof their careers.
We read all your comments - please provide your feedback!
Did you like today's email?
What more do you want to see in this newsletter?
