Posts by John Tornow

The brilliant folks at Sandwich have debuted their new app for immersive video on the Vision Pro: Theater (App Store link). It’s so cool.

What was cooler this week was Sandwich live streaming The Talk Show from WWDC in spatial video. I tuned in for the first half of the show, and it was a stunning experience. I can’t wait to see more of this type of thing in the future.

This is a really nice addition to Apple Pay coming in iOS 18: users will be able to pay with Apple Pay on Windows, and in Chrome browsers. Very nice implementation. Excited to implement this for Air Mail later this year.

A thorough and beautiful post from Louie Mantia about the new “dark mode” icons available in iOS 18.

Some (seemingly) good news after the surprise announcement last week that Alamo was closing its doors here in Dallas. Alamo is such a great place, and I really hope it survives.

Casey Newton, writing at Platformer:

It was a moment that has seemed inevitable since November 2022, when ChatGPT launched and catalyzed global interest in how AI can enhance products. In the 18 months since, impatient investors have worried that Apple might be letting the moment pass it by. Savvier observers have noted that this is how Apple has worked for decades now: approaching new technologies deliberately, and on its own time; developing its distinctive take on the product; and releasing it only when polished to the company’s quality standards.

Judging from the preview, Apple Intelligence was created in just this way. The company took time to develop principles around what AI should do on its devices. It landed on a suite of AI features for the operating system, designed to make its devices more valuable by leveraging the massive amount of personalized data on your devices. (Sensitive to the implications of such an invasive technology, Apple also took pains to develop a more private approach to data processing for AI apps as well.)

The privacy-based approach to AI is such a breath of fresh air in the industry.

The headlining feature of Monday’s WWDC Keynote was clearly the introduction of “Apple Intelligence”, Apple’s newly minted name for all-things-AI at the company.

The first mention of “Artificial Intelligence” by Apple was over an hour into the keynote, when Tim Cook called the company’s initiatives “personal intelligence” and “the next big step for Apple.”

I love the framing of a “personal intelligence system”. It strikes a bold contrast between other AI tools that “use world knowledge” versus something that understands you.

Apple didn’t just roll out a bunch of models and developer jargon– they showed real features that use AI to enhance what you can do by yourself with these devices.

A few of the highlights for me were:

• Rewriting emails in Mail looks really nice. It’s very clever how you can choose from different tones when writing. (Not holding out hope here for a “passive aggressive coworker” tone. 🤣)
• Summarizing of emails in the messages list, instead of just showing the generic subject and pre-header for a mail messages is fantastic.
• “Genmoji” to create your own “emoji”-style images for messages is super cool. This is going to get a ton of usage.
• “Clean up” removes background details and people from photos. A year or so later than Google here, but a welcome addition for Photos.
• The whole concept of Private Cloud Compute is very interesting and perfect for Apple’s brand.
• The ChatGPT integration in Siri is a welcome addition, and will certainly be useful. I love how Apple is very clear before sending any data to OpenAI.

As mentioned in the keynote, I really dig how Apple Intelligence is being framed as an AI assistant that “context and meets you where you are.”

WWDC begins today and the annual keynote featuring new software updates for Apple platforms will inevitably bring a ton of new AI features. I’m sure those will be interesting, but my desires for the platforms are focused elsewhere.

Here are a few other items I’d love to see this year. First a few developer-oriented updates:

An Apple Sports API

Credit to Cihat Gündüz for this idea, but it’s a brilliant one. With Apple’s introduction of a very nice sports app earlier this year it seems that this one is within the realm of possibility. Maybe not this year, but eventually would be amazing. I’d still love to build a purpose-built sports scores app with support for a few niche features for folks like me that constantly are watching games in delay after my kids go to bed. ⚾️

System API for AI Models

Developers should not need to ship their own AI models and optimizations to use them on Apple platforms. Let’s get some first-party support for various AI models for different purposes that can be used with a simple abstracted API.

And some end-user features:

Modernize Mail’s Features

Mail.app is an amazing app, but its features still seem very dated. How about Snoozing, Sending Later, a Screener, and “better” handling of newsletters.

Instant AirDrop

AirDrop is great, when it works. Which is a rare thing for me. And even when it does work it’s incredibly slow, even for devices that are within inches of each other. I’d love to see AirDrop capable of sending a link or photo between devices within seconds, not minutes.

Working Screen Time

As noted last week, let’s focus on a Screen Time feature that actually does what it says. Or at a minimum, doesn’t allow kids to easily bypass.

Messages++

“Messages” is a great app, but compared to its modern competitors (WhatsApp, et al) it’s too simplistic. More filters, better support for SMS and green-bubble group chats would be welcome additions. How about a way to filter all of the 2FA and confirmation codes out of the conversations flow? Group chats that can be summarized by AI. Delayed sending of messages, so I don’t bother people in other time zones late at night. Drafts syncing. Typing indicators for group chats. Expiring “secure” thread. Lots of opportunity here.

Oh Siri

Siri needs a complete reboot, if not a complete rebrand too. It’s so bad, I almost never use it. And when I do and hope for the best, it rarely succeeds. Siri needs a new (likely AI/LLM-based) engine. It’s time.

Vision Pro Next Steps

I’m not sure where to go with the Vision Pro, but I’m hoping for a very nice 2.0 on visionOS.

→ See also, Michael Tsai’s excellent roundup of WWDC wish lists. I agree with all of Ryan Jones’ ideas for control center:

Hopes for a better Control Center:

1. Big clear single tap audio output
2. Pick home controls
3. Any shortcut
4. Hide less in long presses
5. Rotation lock except video
6. All buttons are customizable
7. No double button in Focus Modes
8. Mini TV Remote at first level

Looking forward to watching the keynote later today. I’ve avoided the spoilers and I’m just looking forward to enjoying the presentation.

Speaking of cool new devices, the Daylight DC-1 looks very cool as well.

Nice new app from Leica (App Store Link) with lens simulation styles. A few that are free, and a few more for an upgrade.

Transform your Apple Watch into a playful, functional handheld device that provides connectivity without distraction, encouraging you to leave your phone at home.

I don’t quite get it, but I love it. More cool hardware devices coming soon..

Joanna Stern, with another excellent piece for the Wall Street Journal. (Apple News+) This time about Apple’s Screen Time, which I’ve been trying to use successfully for years. Glad I’m not alone in this thinking.

Stern’s tweet:

Apple’s Screen Time parental controls are broken, and it feels like an afterthought for the company.

The latest example? Two security researchers have been reporting a bug to Apple since 2021 that lets kids visit blocked sites.

Only after I called did Apple say it would be fixed in the next software update.

These seem like more substantial bugs that can be easily fixed in the “next software update.”

Nick Heer summarizes the problem perfectly:

One could reasonably argue nobody should entirely depend on software to determine how devices are used by themselves or their children, but I do not think many people realistically do. It is part of a combination of factors. Screen Time should at perform the baseline functions it promises. It sucks how common problems are basically ignored until Stern writes about them.

Sean Hollister, writing at The Verge:

Before he brought VR headsets out of the dark ages with Oculus, before he got fired by Facebook following his confusing and controversial political donations, before he went on to found a defense contractor named Anduril that builds lethal drones (!), Palmer Luckey was a Nintendo Game Boy modder.

It’s true. There’s photo evidence — and today, he’s revealing his own take on the ultimate cartridge-playing Nintendo Game Boy, one that you can actually buy for $199.

This thing looks really cool. $199, and shipping by Christmas. I love the resurgence of niche and creative hardware products.

Yesterday, current and former members of OpenAI and Google posted an open letter about the risks of AI and the companies developing it:

We are current and former employees at frontier AI companies, and we believe in the potential of AI technology to deliver unprecedented benefits to humanity.

We also understand the serious risks posed by these technologies. These risks range from the further entrenchment of existing inequalities, to manipulation and misinformation, to the loss of control of autonomous AI systems potentially resulting in human extinction. […]

That escalated quickly.

AI companies possess substantial non-public information about the capabilities and limitations of their systems, the adequacy of their protective measures, and the risk levels of different kinds of harm. However, they currently have only weak obligations to share some of this information with governments, and none with civil society. We do not think they can all be relied upon to share it voluntarily.

So long as there is no effective government oversight of these corporations, current and former employees are among the few people who can hold them accountable to the public.

Holding out hope that the government, especially here in the US, is going to sweep in and do a great job regulating this industry seems like a fools errand. What in our recent history shows that the government would be able to do this?

I agree with Casey Newton, on Mastodon:

There’s yet another open letter from the AI safety crowd. If they want more people to take them seriously, they need to get more specific

Rand Fishkin writing at the SparkToro Blog with a bombshell leak of policies from Google on how its search algorithm works:

On Sunday, May 5th, I received an email from a person claiming to have access to a massive leak of API documentation from inside Google’s Search division. The email further claimed that these leaked documents were confirmed as authentic by ex-Google employees, and that those ex-employees and others had shared additional, private information about Google’s search operations.

Many of their claims directly contradict public statements made by Googlers over the years, in particular the company’s repeated denial that click-centric user signals are employed, denial that subdomains are considered separately in rankings, denials of a sandbox for newer websites, denials that a domain’s age is collected or considered, and more.

These documents are really sending the SEO industry into a tailspin this week. This is a fascinating look into one of the most closely guarded secrets in tech.

Mike King also received the documents and has a great breakdown on iPullRank:

I have reviewed the API reference docs and contextualized them with some other previous Google leaks and the DOJ antitrust testimony. […]

You’d be tempted to broadly call these “ranking factors,” but that would be imprecise. Many, even most, of them are ranking factors, but many are not. What I’ll do here is contextualize some of the most interesting ranking systems and features (at least, those I was able to find in the first few hours of reviewing this massive leak) based on my extensive research and things that Google has told/lied to us about over the years.

“Lied” is harsh, but it’s the only accurate word to use here. While I don’t necessarily fault Google’s public representatives for protecting their proprietary information, I do take issue with their efforts to actively discredit people in the marketing, tech, and journalism worlds who have presented reproducible discoveries.

At first, there was silence from Google on this leak. Yesterday, The Verge received email confirmation with a statement:

“We would caution against making inaccurate assumptions about Search based on out-of-context, outdated, or incomplete information,” Google spokesperson Davis Thompson told The Verge in an email. “We’ve shared extensive information about how Search works and the types of factors that our systems weigh, while also working to protect the integrity of our results from manipulation.”

Sara Fischer writing for Axios:

The Atlantic, one of the oldest magazines in the U.S., and Vox Media, one of the nation’s largest digital media holding companies, have both inked separate licensing and product deals with ChatGPT parent OpenAI. […]

The deals give OpenAI added momentum in its quest for credible content to train its algorithms and inform its chatbots — and could also protect the Microsoft-backed company further from future copyright liability. […]

Deal terms weren’t disclosed, but it’s safe to assume both publishers are being compensated for their content — that’s how previous deals between publishers and OpenAI have been structured.

The land grab for licensed content to feed into AI continues.

Also interesting:

Internally, Vox will leverage OpenAI’s tech for its first party data platform, Forte, to bolster advertising creative optimization and audience targeting capabilities.

The Vox Media Union’s response:

Today, members of the Vox Media Union, Thrillist Union, and The Dodo Union were informed without warning that Vox Media entered into a “strategic content and product partnership” with OpenAl. As both journalists and workers, we have serious concerns about this partnership, which we believe could adversely impact members of our union, not to mention the well-documented ethical and environmental concerns surrounding the use of generative Al. We demand that Vox Media engage with us on this issue transparently - and address our many unanswered questions about this partnership — instead of continuing to fail to include our voices in decisions like these. We know that Al is already having a monumental impact on our work, and we demand a seat at the
table in discussions about its future at Vox Media.

Nilay Patel, Editor in Chief of The Verge, responds on Threads:

I don’t have a lot to say about this – our newsroom is independent of the company’s business dealings as it’s always been. We’ll figure out some disclosure language and do a disclosure when it’s appropriate, we are pretty good at those ;)

[…]

To me it’s the same as anything – there is a firewall between editorial and the commercial side of the business. They don’t get to tell us what to do, and we don’t get involved in how they generate revenue