This Week's Sponsor:

Copilot Money

The Apple Editor’s Choice Award App for Tracking Your Money. Start Your Free Trial Today

Posts in Linked

Federico’s Latest Automation Academy Lesson: Building a Better Web Clipper with Shortcuts and AI→

A webpage saved with Universal Clipper.

A webpage saved with Universal Clipper.

I share Federico’s frustration over saving links. Every link may be a URL, but their endpoints can be wildly different. If like us, you save links to articles, videos, product information, and more, it’s hard to find a tool that handles every kind of link equally well.

That was the problem Federico set out to solve with Universal Clipper, an advanced shortcut that automatically detects the kind of link that’s passed to it, and saves it to a text file, which he accesses in Obsidian, although any text editor will work.

Universal Clipper integrates with the Obsidian plugin Dataview, too.

Universal Clipper integrates with the Obsidian plugin Dataview, too.

Universal Clipper, which Federico released yesterday as part of his Automation Academy series for Club MacStories Plus and Premier members, is one of his most ambitious shortcuts that draws on multiple third-party apps, services, and command line tools in an automation that works as a standalone shortcut or as a function that can send its results to another shortcut. As Federico explains:

I learned a lot in the process. As I’ve documented on MacStories and the Club lately, I’ve played around with various templates for Dataview queries in Obsidian; I’ve learnedhow to take advantage of the Mac’s Terminal and various CLI utilities to transcribe long YouTube videos and analyze them with Gemini 2.5; I’ve explored new ways to interact with web APIs in Shortcuts; and, most recently, I learned how to properly prompt GPT 4.1 with precise instructions. All of these techniques are coming together in Universal Clipper, my latest, Mac-only shortcut that combines macOS tools, Markdown, web APIs, and AI to clip any kind of webpage from any web browser and save it as a searchable Markdown document in Obsidian.

Although the shortcut may be complex, the best part of Federico’s post is how easy it is to follow. Along the way, you’ll learn a bunch of techniques and approaches to Shortcuts automation that you can adapt for your own shortcuts, too.

Automation Academy is just one of many perks that Club MacStories Plus and Club Premier members enjoy including:

  • Weekly and monthly newsletters 
  • A sophisticated web app with search and filtering tools to navigate eight years of content
  • Customizable RSS feeds
  • Bonus columns
  • An early and ad-free version of our Internet culture and media podcast, MacStories Unwind
  • A vibrant Discord community of smart app and automation fans who trade a wealth of tips and discoveries every day
  • Live Discord audio events after Apple events and at other times of the year

On top of that, Club Premier members get AppStories+, an extended, ad-free version of our flagship podcast that we deliver early every week in high-bitrate audio.

Use the buttons below to learn more and sign up for Club MacStories+ or Club Premier.

Join Club MacStories+:

Join Club Premier:

Permalink

Amazon Adds Links to Purchase Books to Its Kindle App→

As reported by Andrew Liszewski at The Verge, Amazon has updated the Kindle app to add a “Get Book” button, a direct result of last week’s contempt order entered by Judge Gonzalez Rodgers. When tapped, it takes users to the Amazon page for the book in Safari with the Kindle version selected. Before today’s update, you couldn’t purchase a book without going to Safari first.

In a low key statement to The Verge over email, Amazon’s Tim Gillman said:

We regularly make improvements to our apps to help ensure we are providing customers the most convenient experience possible. By selecting ‘Get Book’ within the Kindle for iOS app, customers can now complete their purchase through their mobile web browser.

I expect other companies will follow Amazon and Spotify’s leads in the coming weeks. Although Apple has appealed Judge Gonzalez Rodgers’ contempt order, the Judge declined to stay its enforcement during the appeals process. It’s always possible an appeal could force Amazon and others to undo changes like this, but I think a more likely outcome is that an appellate court allows Apple to charge a fee where Judge Gonzalez Rodgers wasn’t – one that’s lower than the 27% that got Apple into trouble in the first place.

Permalink

Post-Chat UI→

Fascinating analysis by Allen Pike on how, beyond traditional chatbot interactions, the technology behind LLMs can be used in other types of user interfaces and interactions:

While chat is powerful, for most products chatting with the underlying LLM should be more of a debug interface – a fallback mode – and not the primary UX.

So, how is AI making our software more useful, if not via chat? Let’s do a tour.

There are plenty of useful, practical examples in the story showing how natural language understanding and processing can be embedded in different features of modern apps. My favorite example is search, as Pike writes:

Another UI convention being reinvented is the search field.

It used to be that finding your flight details in your email required typing something exact, like “air canada confirmation”, and hoping that’s actually the phrasing in the email you’re thinking of.

Now, you should be able to type “what are the flight details for the offsite?” and find what you want.

Having used Shortwave and its AI-powered search for the past few months, I couldn’t agree more. The moment you get used to searching without exact queries or specific operators, there’s no going back.

Experience this once, and products with an old-school text-match search field feel broken. You should be able to just find “tax receipts from registered charities” in your email app, “the file where the login UI is defined” in your IDE, and “my upcoming vacations” in your calendar.

Interestingly, Pike mentions Command-K bars as another interface pattern that can benefit from LLM-infused interactions. I knew that sounded familiar – I covered the topic in mid-November 2022, and I still think it’s a shame that Apple hasn’t natively implemented these anywhere in their apps, especially now that commands can be fuzzier (just consider what Raycast is doing). Funnily enough, that post was published just two weeks before the public debut of ChatGPT on November 30, 2022. That feels like forever ago now.

Permalink

Sundar Pichai Testifies That He Hopes Gemini Will Be Integrated into iPhones This Fall→

Ever since Apple announced its deal to integrate ChatGPT into Siri, there have been hints that the company wanted to make deals with other AI providers, too. Alphabet CEO Sundar Pichai has added fuel to the rumors with testimony given today in the remedy phase of the search antitrust case brought against it by the U.S. Department of Justice.

In response to questions by a DOJ prosecutor, Pichai testified that he hoped Google Gemini would be added to iPhones this year. According to a Bloomberg story co-authored by Mark Gurman, Davey Alba, and Leah Nylen:

Pichai said he held a series of conversations with Apple Chief Executive Officer Tim Cook across 2024 and he hopes to have a deal done by the middle of this year.

This news isn’t surprising, but it is welcome. Despite Google’s early stumbles with Bard, its successor, Gemini, has improved by leaps and bounds in recent months and has the advantage of being integrated with many of Google’s other products that have a huge user base. What will be interesting to see is whether Gemini is integrated as an alternative fallback for Siri requests or whether Apple and Google ink a broader deal that integrates Gemini into other aspects of iOS.

Permalink

Sycophancy in GPT-4o→

OpenAI found itself in the middle of another controversy earlier this week, only this time it wasn’t about publishers or regulation, but about its core product – ChatGPT. Specifically, after rolling out an update to the default 4o model with improved personality, users started noticing that ChatGPT was adopting highly sycophantic behavior: it weirdly agreed with users on all kinds of prompts, even about topics that would typically warrant some justified pushback from a digital assistant. (Simon Willison and Ethan Mollick have a good roundup of the examples as well as the change in the system prompt that may have caused this.) OpenAI had to roll back the update and explain what happened on the company’s blog:

We have rolled back last week’s GPT‑4o update in ChatGPT so people are now using an earlier version with more balanced behavior. The update we removed was overly flattering or agreeable—often described as sycophantic.

We are actively testing new fixes to address the issue. We’re revising how we collect and incorporate feedback to heavily weight long-term user satisfaction and we’re introducing more personalization features, giving users greater control over how ChatGPT behaves.

And:

We also believe users should have more control over how ChatGPT behaves and, to the extent that it is safe and feasible, make adjustments if they don’t agree with the default behavior.

Today, users can give the model specific instructions to shape its behavior with features like custom instructions. We’re also building new, easier ways for users to do this. For example, users will be able to give real-time feedback to directly influence their interactions and choose from multiple default personalities.

“Easier ways” for users to adjust ChatGPT’s behavior sound to me like a user-friendly toggle or slider to adjust ChatGPT’s personality (Grok has something similar, albeit unhinged), which I think would be a reasonable addition to the product. I’ve long argued that Siri should come with an adjustable personality similar to CARROT Weather, which lets you tweak whether you want the app to be “evil” or “professional” with a slider. I increasingly feel like that sort of option would make a lot of sense for modern LLMs, too.

Permalink

How Does This Keep Happening?→

Today, Blue Prince, a critically acclaimed videogame appeared on Apple’s App Store. The trouble was, it wasn’t offered for sale by its developer, Dogubomb, or its publisher, Raw Fury. The real Blue Prince is only available on the Xbox, PlayStation, and PC.

What appeared on the App Store, and has since been removed, was an opportunistic scam as Jay Peters explained for The Verge:

Before it was removed, I easily found one iOS copy of the game just by searching Blue Prince on the App Store – it was the first search result. The icon looked like it would be the icon for a hypothetical mobile version of the game, the listing had screenshots that looked like they were indeed from Blue Prince, and the description for the game matched the description on Steam.

The port was available long enough for Blue Prince’s developer and publisher to post about it on Bluesky and, according to Peters, for the fake to reach #8 in the App Store Entertainment category. I feel for anyone who bought the game assuming it was legit given Peters’ experience:

I also quickly ran into a major bug: when I tried to walk through one of the doors from the Entrance Hall, I fell through the floor.

This isn’t the first time this sort of thing has happened. As Peters points out it happened to Palworld and Wordle too. Other popular games that have appeared on the App Store as janky scam ports include Cuphead, a version of Balatro that appeared before its official release on iOS, and Unpacking.

This seems like the sort of thing that could be fixed through automation. Scammers want users to find these games, so they can make a quick buck. As a result, the name of the game is often identical to what you’d find on the Steam, Xbox, or PlayStation stores. It strikes me that a combination of automated searching for the top games on each store, combined with an analysis of how quickly a game is moving up the charts would catch a lot of this sort of thing, flagging it for reviewers who could take a closer look.

By the way, if you haven’t tried Blue Prince, you should. It’s an amazing game and early contender for game of the year. You can learn more about the game and find links to where to buy it here. Also, Brendon Bigley, my NPC co-host, has an excellent written and video review of Blue Prince on Wavelengths.

Permalink

Contabulation→

Rumors have been flying for a while about a planned redesign for iOS 19. One of the rumors is that iOS tab bars will support search bars, which led Ben McCarthy, the developer of Obscura, to write a terrific breakdown of how tab bars should be used:

If search is the primary form of navigation, as in Safari, Maps, or Callsheet, it should be at the bottom. If a search bar is just used for filtering content already on screen, then it can make more sense to leave it at the top, as scrolling is probably the more natural way to find what you’re looking for (the Settings app is a good example of this). So I’m delighted at the rumours that iOS 19’s Tab Bars can adapt into Search Bars when needed. I think it’ll be [a] big improvement and allow for more flexible navigation patterns with less code.

But Ben didn’t just provide pointers on how tab bars should be used. They also explained that tab bars:

  • should support actions and context menus,
  • accommodate more than five tabs,
  • and allow for user-generated tabs, something that is common on macOS.

It’s a great post, well worth studying as we wait to see whether and how far Apple will go in modifying the tab bar. As Ben notes, the tab bar has been around since the beginning of the iPhone, has changed very little, and is due for a redesign. I agree.

Permalink

Whisky Shuts Down Project That Enabled Windows Gaming on Mac→

Not long ago, Isaac Marovitz, the developer behind Whisky, the open source WINE front-end that made it easy to play Windows games on a Mac, announced the project had come to an end. Whisky is how Niléane got Cities: Skylines 2 running on an M2 MacBook Air in 2023, and the project was well-regarded in the gaming community for its ease of use. In shutting down the project, Marovitz encouraged Whisky users to move to CrossOver, a paid app by CodeWeavers.

In an interview with Ars Technica’s Kevin Purdy, Marovitz said:

“I am 18, yes, and attending Northeastern University, so it’s always a balancing act between my school work and dev work,” Isaac Marovitz wrote to Ars. The Whisky project has “been more or less in this state for a few months, I posted the notice mostly to clarify and formally announce it,” Marovitz said, having received “a lot of questions” about the project status.

As Purdy explained for Ars Technica, Marovitz also became concerned that his free project threatened CrossOver, and by extension, WINE itself. Last week, CodeWeavers’ CEO wrote about the shutdown, to acknowledge Marovitz’s work and commend his desire to protect the WINE project.

It’s always a shame to see a project as popular and polished as Whisky discontinued. Some gamers may not like that CrossOver is a paid product, but I’m glad that there’s an alternative for those who want it.

To me though, the popularity and fragility of projects like Whisky highlight that a better solution would be for Apple to open its Game Porting Toolkit to users. The Game Porting Toolkit is built on CrossOver’s open source code. However, unlike the CrossOver app sold to gamers, Apple’s Game Porting Toolkit is meant for developers who want to move a game from Windows to Mac. It’s not impossible for gamers to use, but it’s not easy either. I’m not the first to suggest this, and Valve has demonstrated both the technical and commercial viability of such an approach with Proton, but as WWDC approaches, a user-facing Game Porting Toolkit is near the top of my macOS 16 wish list.

Permalink

How Could Apple Use Open-Source AI Models?→

Yesterday, Wayne Ma, reporting for The Information, published an outstanding story detailing the internal turmoil at Apple that led to the delay of the highly anticipated Siri AI features last month. From the article:

In November 2022, OpenAI released ChatGPT to a thunderous response from the tech industry and public. Within Giannandrea’s AI team, however, senior leaders didn’t respond with a sense of urgency, according to former engineers who were on the team at the time.

The reaction was different inside Federighi’s software engineering group. Senior leaders of the Intelligent Systems team immediately began sharing papers about LLMs and openly talking about how they could be used to improve the iPhone, said multiple former Apple employees.

Excitement began to build within the software engineering group after members of the Intelligent Systems team presented demos to Federighi showcasing what could be achieved on iPhones with AI. Using OpenAI’s models, the demos showed how AI could understand content on a user’s phone screen and enable more conversational speech for navigating apps and performing other tasks.

Assuming the details in this report are correct, I truly can’t imagine how one could possibly see the debut of ChatGPT two years ago and not feel a sense of urgency. Fortunately, other teams at Apple did, and it sounds like they’re the folks who have now been put in charge of the next generation of Siri and AI.

There are plenty of other details worth reading in the full story (especially the parts about what Rockwell’s team wanted to accomplish with Siri and AI on the Vision Pro), but one tidbit in particular stood out to me: Federighi has now given the green light to rely on third-party, open-source LLMs to build the next wave of AI features.

Federighi has already shaken things up. In a departure from previous policy, he has instructed Siri’s machine-learning engineers to do whatever it takes to build the best AI features, even if it means using open-source models from other companies in its software products as opposed to Apple’s own models, according to a person familiar with the matter.

“Using” open-source models from other companies doesn’t necessarily mean shipping consumer features in iOS powered by external LLMs. I’ve seen some people interpret this paragraph as Apple preparing to release a local Siri powered by Llama 4 or DeepSeek, and I think we should pay more attention to that “build the best AI features” (emphasis mine) line.

My read of this part is that Federighi might have instructed his team to use distillation to better train Apple’s in-house models as a way to accelerate the development of the delayed Siri features and put them back on the company’s roadmap. Given Tim Cook’s public appreciation for DeepSeek and this morning’s New York Times report that the delayed features may come this fall, I wouldn’t be shocked to learn that Federighi told Siri’s ML team to distill DeepSeek R1’s reasoning knowledge into a new variant of their ∼3 billion parameter foundation model that runs on-device. Doing that wouldn’t mean that iOS 19’s Apple Intelligence would be “powered by DeepSeek”; it would just be a faster way for Apple to catch up without throwing away the foundational model they unveiled last year (which, supposedly, had a ~30% error rate).

In thinking about this possibility, I got curious and decided to check out the original paper that Apple published last year with details on how they trained the two versions of AFM (Apple Foundation Model): AFM-server and AFM-on-device. The latter would be the smaller, ~3 billion model that gets downloaded on-device with Apple Intelligence. I’ll let you guess what Apple did to improve the performance of the smaller model:

For the on-device model, we found that knowledge distillation (Hinton et al., 2015) and structural pruning are effective ways to improve model performance and training efficiency. These two methods are complementary to each other and work in different ways. More specifically, before training AFM-on-device, we initialize it from a pruned 6.4B model (trained from scratch using the same recipe as AFM-server), using pruning masks that are learned through a method similar to what is described in (Wang et al., 2020; Xia et al., 2023).

Or, more simply:

AFM-server core training is conducted from scratch, while AFM-on-device is distilled and pruned from a larger model.

If the distilled version of AFM-on-device that was tested until a few weeks ago produced a wrong output one third of the time, perhaps it would be a good idea to perform distillation again based on knowledge from other smarter and larger models? Say, using 250 Nvidia GB300 NVL72 servers?

(One last fun fact: per their paper, Apple trained AFM-server on 8192 TPUv4 chips for 6.3 trillion tokens; that setup still wouldn’t be as powerful as “only” 250 modern Nvidia servers today.)

Permalink
OSZAR »