In partnership with
🧠 THAT ONE AI - This Week’s Signal
This week:
🧠 Anthropic opens its most powerful model to the public
🌍 Google's real-time translation is genuinely impressive
📱 Apple's Siri AI lands — with caveats
🧰 Tools worth testing
🧠 Anthropic Just Made Its Most Powerful Model Public
For months, Anthropic's Mythos tier existed behind closed doors, available only to around 150 vetted partners through Project Glasswing. This week, a version of it became available to everyone.
Claude Fable 5 is that public release. A few things worth knowing:
Fable is a scoped version of Mythos, with queries touching cybersecurity, biology, and chemistry rerouted to Opus 4.8 instead
On benchmarks, it posts significant gains over both Opus 4.8 and GPT-5.5 across coding, reasoning, and knowledge work
Mythos 5 also releases to Project Glasswing partners with looser restrictions on cybersecurity use and lower costs than the original Mythos Preview
Fable is available across all Claude subscription tiers until June 22, after which it moves to separate usage credits priced at $10 per million input tokens and $50 per million output tokens
Most labs call every release their best model ever. What's different here is that the broader AI community seems to actually agree with Anthropic on this one. The benchmark results are hard to argue with.
The June 22 pricing shift is worth noting if you plan to build on it.
The bigger signal: 👉 The gap between frontier AI and what's publicly accessible just closed significantly. What happens next depends on what people actually build with it.
An engineering fellowship to land your next job
Many engineers feel stalled because the role itself has not evolved. The work looks the same, but the market has moved.
Senior engineering in 2026 demands ownership, faster judgment, and comfort with ambiguity. If your role is not pushing you there, it may be holding you back.
Last cohort, 15 hiring partners sent 31 representatives to evaluate challengers through 246 live interviews. Gauntlet offers a reset. Apply now.
Must be a US citizen to qualify.
🌍 Google's Live Translate Might Be the Most Useful AI Demo This Year
Gemini 3.5 Live Translate does one thing and does it well: it listens to spoken audio and translates it in real time across 70+ languages, staying just a few seconds behind the speaker while preserving tone and pacing.
It is already live in the Google Translate app and rolling out to Google Meet. Watch it in action to get a sense of how smooth the output actually is.
No switching apps, no waiting for a transcript, no flat robotic output. The speaker sounds like themselves in another language.
For anyone running international meetings, working with multilingual teams, or communicating across language barriers regularly, this is the kind of tool that removes friction in a way that actually changes how you work.
The bigger signal: 👉 Real-time translation that preserves voice and pacing is not a novelty. For global teams, it is infrastructure.
📱 Apple Rebuilt Siri. It's Better. It's Also Two Years Behind.
At WWDC 2026, Apple officially rebranded its assistant to Siri AI and delivered the overhaul it promised back in 2024. Two years late, but it's here.
What actually changed:
Siri AI runs on Apple's own models built in collaboration with Google's Gemini, though not the same Gemini version Google ships to its own users
It can read your screen, pull context from Photos and Messages, and take actions across apps without you switching between them
A dedicated Siri AI app functions as a full chatbot hub, keeping conversation history linked privately across your devices
All processing happens on-device or through Private Cloud Compute with no data saved
Free update this fall for iPhone 15 Pro and newer, public beta next month, no EU or China access at launch
Here's the honest read: for someone who has never used a frontier AI model, this will feel like a leap. For anyone who uses Claude or ChatGPT regularly, the demos looked like 2024-level capability. Siri is better. It's just not close to the frontier.
The bigger signal: 👉 Apple's real advantage is not the model. It's two billion devices and a privacy architecture nobody else can match at that scale.
Tired of news that feels like noise?
Every day, 4.5 million readers turn to 1440 for their factual news fix. We sift through 100+ sources to bring you a complete summary of politics, global events, business, and culture — all in a brief 5-minute email. No spin. No slant. Just clarity.
That One AI 🧰 TOOLBOX
A few tools quietly worth exploring:
🌐 Gemini 3.5 Live Translate → Speech-to-speech translation across 70+ languages, live in Google Translate and Google Meet
🎨 ChatGPT Magic Layer → Turn ChatGPT-generated images into fully editable Canva designs in one step
⚡ Zed → A code editor built for speed and AI collaboration, designed for developers who find other editors too slow
🎙️ Demi → Drafts email replies in your voice, books meetings, and keeps important threads from falling through the cracks
🔚 EXIT NODE
Apple shipped a better Siri but is still chasing the frontier. Anthropic opened its most powerful model to the public. Google made speaking across languages feel almost effortless.
The gap between AI research and what regular people can actually use is closing fast.
The question is no longer what AI can do. It is whether the tools people already use will deliver it well enough to matter.
See you next issue.