Anthropic’s first Mythos-class Claude model, Fable 5, hit the world like an atom bomb this week, and that’s barely an exaggeration. But Apple’s rebooted Siri could be the AI moment that actually reaches everyone else.
A modified version of Mythos, the benchmark-shattering Claude model that’s scary-good at cybersecurity and worryingly knowledgeable about bioweapons, Fable 5 comes wrapped in so many safeguards that it reportedly refuses even the most basic chats about biology. (Mythos itself is still restricted to a small subset of approved users.)
Claude Fable also kicked off other controversies, including stringent data-retention policies that made even Microsoft balk, while Claude subscription users learned that the model would be yanked from their plans later this month.
In contrast to the Fable firestorm, Apple’s long-awaited AI relaunch was more like a chill summer cookout. Chastened after its botched Apple Intelligence rollout two years ago, Apple judiciously pumped the brakes, introducing a variety of low-key AI tools that, on the surface, looked relatively tame.
Welcome to another edition of Prompt Mode, your weekly AI newsletter.
I’m your host, Ben Patterson. Each week on Prompt Mode, I’ll be serving up analysis of the AI trends that matter to everyday users like you and me. Stay tuned for practical AI tips, hands-on experiences with the latest AI tools, and–you guessed it–prompts to help you get the most out of your AI assistants.
Thanks for reading, and if you like what you see, just sign up right here.
Yet Apple’s revamped Siri AI app and quietly powerful Apple Intelligence features are (I’d argue) collectively a bigger deal than the Fable brouhaha.
For most of us, Claude Fable — and on an even greater scale, Mythos — are both awe-inspiring and inaccessible, like the atom-smashing CERN particle accelerator. Both Fable and Mythos are capable of agentic AI feats that stagger the mind, yet they’re ridiculously overpowered for everyday use, more likely to trigger a cybersecurity apocalypse than help make our everyday lives better.
Apple’s AI features, on the other hand, are smaller yet far more useful, like Siri’s ability to scour your iCloud data for details about your loved ones, compose and send email messages on the fly, and create calendar events from natural language prompts. On a Mac, Siri will always be a right-click away, ready to analyze data, organize files, and compose text from scratch, while Apple Intelligence will allow apps like Passwords to strengthen weak passwords using behind-the-scenes AI agents.
These quietly powerful AI features will be baked into Apple products in the hands of millions of everyday users — and for many of them, it will mark the first time they’ve used AI to actually do something rather than just chat. Personally, I think that’s a big deal, even bigger than Claude Fable’s Godzilla-like arrival.
Call me crazy, but I’d like to think that Apple AI redux could usher in a new “AI for all” era, where AI is truly working for us rather than the other way around.
More in AI this week
- Claude Fable may be ridiculously overkill for everyday tasks, but I’m still impressed by the utility of smaller Claude models, like the one that created a polished and greatly improved user manual for my new air conditioner. (PCWorld)
- Want more examples of Apple AI features that could be surprisingly useful? Here are 5 of them. (PCWorld)
- The most impressive AI capabilities are often the scariest because of their privacy implications. Apple’s Private Cloud Compute may be the solution. (PCWorld)
- ChatGPT and other AI chatbots have a strange fascination with lighthouse keepers, particularly one named “Elias Thorne.” AI researchers have some theories. (404 Media)
- AI is expensive, especially when you have teams of AI agents gobbling tokens like they’re M&Ms. Yet we may be on the precipice of a price war, with Google, Anthropic, and OpenAI reportedly poised to slash prices as they furiously compete for users. (WSJ)
- One of the scariest AI vulnerabilities is prompt injection, which employ malicious prompts or hidden text to subvert an AI’s instructions and/or guardrails. OpenAI’s new Lockdown Mode is designed to stop prompt injection attacks in their tracks. (PCWorld)
Prompt of the week: The “board of advisors” prompt
Need advice on a mid-tier personal decision, like the best way to spend a rare day off or whether to choose a gym membership over at-home exercise equipment? You could put it to AI, but you might wind up with a long-winded answer that’s not particularly nuanced or helpful.
Instead, try a prompt that turns ChatGPT or another AI chatbot into a “board of advisors,” with each member of the board debating your issue from a different perspective. Once the experts are done picking over your dilemma, they’ll serve up a recommendation that may surprise you with its thoughtfulness.
That’s all for now!
Thanks for reading the latest issue of Prompt Mode. Want more next week? Don’t forget to sign up to start receiving this newsletter in your inbox.
See you next time!
