Reading view

There are new articles available, click to refresh the page.

The Effects of Another Ad in iOS App Store Search

By: Nick Heer

Jeremy Provost of development firm Think Tap Work:

It’s been 64 days since we first noticed Apple’s second ad position in search results for iPhone and iPad. Our update after two weeks showed consistently less search ad impressions for our apps, unless we invested heavily in paying for Search Ads.

Here are some updated numbers. Just like last time, these numbers only include App Store Search impressions from iOS devices. As you’ll see, these numbers get harder and harder to compare over time.

Chris Lindsay, developer of Nihongo, a Japanese dictionary app:

Before the rollout, my organic and paid downloads had remained pretty steady for most of the last year. After the rollout, my my organic installs dropped, and my paid installs rose. My overall downloads actually stayed roughly flat, but a large chunk of what used to be organic downloads appears to have shifted into paid downloads instead:

The ads themselves still work well. The problem is that many of these paid downloads seem to be users I previously would have acquired organically.

These ads are effectively another surcharge Apple has foisted upon developers for the privilege of distributing software to my iPhone and yours. Far from being premium “curated” experience, the App Store is this way because Apple has every incentive to steadily make it a little bit worse for users and developers — because where else are you going to go for your iPhone apps?

⌥ Permalink

The Mythical App Store Reviewer Month

By: Nick Heer

Jeff Johnson:

I’d like to make an analogy between software development and Apple App Store review. A common, cursory reaction to the obvious failures of app review, the continual appearance of countless scams in the App Store, is to suggest that Apple hire more reviewers. My contention is that adding reviewers is not a solution to the problem of App Store curation, and the belief in such a solution is a myth. I don’t claim that hiring more reviewers would make app review slower. Rather, I think that meaningful, effective curation can’t be measured simply by the amount of available labor, much like [Fred] Brooks argues that the possibility of measuring useful work in units of time, man-months, is a myth.

Apple markets the App Store as a “curated storefront”, but that is not meaningfully true if it is serving up, as Apple says, about two million apps. Meanwhile, as Johnson writes, “nobody worries about scams in Apple Arcade […] a truly curated service”.

The thing is that Apple’s App Store should have a carefully selected inventory of apps. That is Apple’s whole brand: premium, highly-desirable products, and people are willing to pay a little more. The App Store does not match that promise. I think the direction of regulatory and court decisions on the governance of iOS app distribution could be a gift for more selective curation, the kind of thing for which some third-party developers would want to pay extra compared to the competing third-party app marketplaces that would also be available.

Alas, we are on the cusp of another WWDC during which Apple seems unlikely to make major changes to software distribution across its many “post-P.C.” platforms.

⌥ Permalink

‘How Deepfakes Tore a High School Apart’

By: Nick Heer

Samantha Cole, 404 Media:

On the morning of December 4, five ninth grade girls, all 14 or 15 years old, showed up for class at Radnor High School. By 8 a.m. — the sun had been up for less than an hour — it felt like the entire school already heard what happened the night before. A fellow freshman boy allegedly created AI-generated sexually explicit videos of the girls using an app, and sent them to his friends. From there, word of the videos and gossip spread from teenager to teenager, school to school, until they made their way back to the girls whose faces were in the deepfakes.

[…]

The images originated from one boy, who used an app called Movely, the girls and their parents believe. The app is similar to dozens hosted in the Apple and Google app stores and advertised on Instagram and TikTok that promise to create AI images and videos of users as superheroes, animals, or influencers; behind a paywall, however, users could edit photos and videos with text prompts.

It almost goes without saying, but the “paywall” is — or was; the app has been removed — an in-app payment from which Apple takes a 15–30% cut.

Apple released its annual justification for running software distribution through the App Store — it told European regulators it actually has five, so maybe this press release only concerns the one accessible from an iPhone — and there are some big numbers in it, as usual. Apple says it “took a number of actions to block bad actors from distributing malicious software, rejecting over 2 million problematic app submissions last year alone”. This Movely app was not one of them. It was only removed after the Tech Transparency Project reported in April that App Store search terms like “nudify” and “undress” displayed results for apps that do exactly that. In its press release, Apple says it has many features for directing kids to age-appropriate apps and restricting them from downloading those which are not but, of the software found by TTP in the App Store and Google Play Store, “31 of the apps were rated suitable for minors”.

Of Movely, the TTP said in its report:

Likewise, an App Store search for “adult AI” returned an ad for Movely – AI Photo to Video. The app offers a suite of AI photo and video editing tools including a try-on feature that will replace a woman’s clothes with outfits including bikinis and lingerie. One tool allows users to select part of any photo and edit it with a text prompt. To test this feature, TTP uploaded an image of a woman in a white T-shirt standing next to a river. After using the selection tool to highlight the woman’s shirt, we entered the prompt “topless.” The app immediately generated four versions of the woman nude from the waist up. It required a paid subscription to download the AI images.

TTP could not reach Movely’s developer, FES2 Inc., for comment. Emails sent to the developer bounced back as undeliverable.

(For clarity, the TTP says it used A.I.-generated images of women to test these apps.)

The search query used to find this app, “adult A.I.”, feels like something Apple should be testing against. If it does not want porn or porn-adjacent apps in its store, it should obviously block these kinds of keywords and flag the apps which are in the results. Moreover, Apple says:

As powerful AI development tools drive a surge in app submissions, Apple’s App Review process has seamlessly scaled to handle the volume and to help ensure every new app and app update meets the App Store’s high standards for privacy, security, and quality.

The Movely app should have raised flags here, too. The developer’s website was, according to the .co whois site, registered in July 2025, and is basically a placeholder. The app’s website was registered a week earlier, and the email address in the privacy policy does not match the one in the terms of service, nor does either match the developer’s website. Also, the blog is full of posts about generating A.I. girls and changing clothes.

These red flags are not obvious in hindsight; they should have been obvious from the time this app was submitted. Meanwhile, apps from longtime and trustworthy developers like Manton Reece and Radu Dutzan are stuck in App Review for dumb and basically invalid reasons.

⌥ Permalink

WWDC 2025 Announced

By: Nick Heer

Like those since 2020, WWDC 2025 appears to be an entirely online event with a one-day in-person event. While it is possible there will be live demos — I certainly hope that is the case — I bet it is a two-hour infomercial again.

If you are planning on travelling there and live outside the United States, there are some things you should know and precautions you should take, particularly if you are someone who is transgender or nonbinary. It is a good thing travel is not required, and hopefully Apple will once again run labs worldwide.

⌥ Permalink

⌥ Apple Could Build Great Platforms for Third-Party A.I. If It Wanted To

By: Nick Heer

There is a long line of articles questioning Apple’s ability to deliver on artificial intelligence because of its position on data privacy. Today, we got another in the form of a newsletter.

Reed Albergotti, Semafor:

Meanwhile, Apple was focused on vertically integrating, designing its own chips, modems, and other components to improve iPhone margins. It was using machine learning on small-scale projects, like improving its camera algorithms.

[…]

Without their ads businesses, companies like Google and Meta wouldn’t have built the ecosystems and cultures required to make them AI powerhouses, and that environment changed the way their CEOs saw the world.

Again, I will emphasize this is a newsletter. It may seem like an article from a prestige publisher that prides itself on “separat[ing] the facts from our views”, but you might notice how, aside from citing some quotes and linking to ads, none of Albergotti’s substantive claims are sourced. This is just riffing.

I remain skeptical. Albergotti frames this as both a mindset shift and a necessity for advertising companies like Google and Meta. But the company synonymous with the A.I. boom, OpenAI, does not have the same business model. Besides, Apple behaves like other A.I. firms by scraping the web and training models on massive amounts of data. The evidence for this theory seems pretty thin to me.

But perhaps a reluctance to be invasive and creepy is one reason why personalized Siri features have been delayed. I hope Apple does not begin to mimic its peers in this regard; privacy should not be sacrificed. I think it is silly to be dependent on corporate choices rather than legislation to determine this, but that is the world some of us live in.

Let us concede the point anyhow, since it suggests a role Apple could fill by providing an architecture for third-party A.I. on its products. It does not need to deliver everything to end users; it can focus on building a great platform. Albergotti might sneeze at “designing its own chips […] to improve iPhone margins”, which I am sure was one goal, but it has paid off in ridiculously powerful Macs perfect for A.I. workflows. And, besides, it has already built some kind of plugin architecture into Apple Intelligence because it has integrated ChatGPT. There is no way for other providers to add their own extension — not yet, anyhow — but the system is there.

Gus Mueller:

The crux of the issue in my mind is this: Apple has a lot of good ideas, but they don’t have a monopoly on them. I would like some other folks to come in and try their ideas out. I would like things to advance at the pace of the industry, and not Apple’s. Maybe with a blessed system in place, Apple could watch and see how people use LLMs and other generative models (instead of giving us Genmoji that look like something Fisher-Price would make). And maybe open up the existing Apple-only models to developers. There are locally installed image processing models that I would love to take advantage of in my apps.

Via Federico Viticci, MacStories:

Which brings me to my second point. The other feature that I could see Apple market for a “ChatGPT/Claude via Apple Intelligence” developer package is privacy and data retention policies. I hear from so many developers these days who, beyond pricing alone, are hesitant toward integrating third-party AI providers into their apps because they don’t trust their data and privacy policies, or perhaps are not at ease with U.S.-based servers powering the popular AI companies these days. It’s a legitimate concern that results in lots of potentially good app ideas being left on the table.

One of Apple’s specialties is in improving the experience of using many of the same technologies as everyone else. I would like to see that in A.I., too, but I have been disappointed by its lacklustre efforts so far. Even long-running projects where it has had time to learn and grow have not paid off, as anyone can see in Siri’s legacy.

What if you could replace these features? What if Apple’s operating systems were great platforms by which users could try third-party A.I. services and find the ones that fit them best? What if Apple could provide certain privacy promises, too? I bet users would want to try alternatives in a heartbeat. Apple ought to welcome the challenge.

Apple’s Restrictions on Third-Party Hardware Interoperability

By: Nick Heer

There is a free market argument that can be made about how Apple gets to design its own ecosystem and, if it is so restrictive, people will be more hesitant to buy an iPhone since they can get more choices with an Android phone. I get that. But I think it is unfortunate so much of our life coalesces around devices which are so restrictive compared to those which came before.

Recall Apple’s “digital hub” strategy. The Mac would not only connect to hardware like digital cameras and music players; the software Apple made for it would empower people to do something great with those photos and videos and their music.

The iPhone repositioned that in two ways. First, the introduction of iCloud was a way to “demote” the Mac to a device at an equivalent level to everything else. Second, and just as importantly, is how it converged all that third-party hardware into a single device: it is the digital camera, the camcorder, and the music player. As a result, its hub-iness comes mostly in the form of software. If a developer can assume the existence of particular hardware components, they have extraordinary latitude to build on top of that. However, because Apple exercises control over this software ecosystem, it limits its breadth.

Like the Mac of 2001, it is also a hub for accessories — these days, things like headphones and smartwatches. Apple happens to make examples of both. You can still connect third-party devices — but they are limited.

Eric Migicovsky, of Pebble:

I want to set expectations accordingly. We will build a good app for iOS, but be prepared – there is no way for us to support all the functionality that Apple Watch has access to. It’s impossible for a 3rd party smartwatch to send text messages, or perform actions on notifications (like dismissing, muting, replying) and many, many other things.

Even if you believe Apple is doing this not out of anticompetitive verve, but instead for reasons of privacy, security, API support, and any number of other qualities, it still sucks. What it means is that Apple is mostly competing against itself, particularly in smartwatches. (Third-party Bluetooth headphones, like the ones I have, mostly work fine.)

The European Commission announced guidance today for improving third-party connectivity with iOS. Apple is, of course, miserable about this. I am curious to see the real-world results, particularly as the more dire predictions of permitting third-party app distribution have — shockingly — not materialized.

Imagine how much more interesting this ecosystem could be if there were substantial support across “host” platforms.

⌥ Permalink

The Deskilling of Web Development

By: Nick Heer

Baldur Bjarnason:

But instead we’re all-in on deskilling the industry. Not content with removing CSS and HTML almost entirely from the job market, we’re now shifting towards the model where devs are instead “AI” wranglers. The web dev of the future will be an underpaid generalist who pokes at chatbot output until it runs without error, pokes at a copilot until it generates tests that pass with some coverage, and ships code that nobody understand and can’t be fixed if something goes wrong.

There are parallels in the history of software development to the various abstractions accumulated in a modern web development stack. Heck, you can find people throughout history bemoaning how younger generations lack some fundamental knowledge since replaced by automation or new technologies. It is always worth a gut-check about whether newer ideas are actually better. In the case of web development, what are we gaining and losing by eventually outsourcing much of it to generative software?

I think Bjarnason is mostly right: if web development become accessible by most through layers of A.I. and third-party frameworks, it is abstracted to such a significant extent that it becomes meaningless gibberish. In fairness, the way plain HTML, CSS, and JavaScript work is — to many — meaningless gibberish. It really is better for many people that creating things for the web has become something which does not require a specialized skillset beyond entering a credit card number. But that is distinct from web development. When someone has code-level responsibility, they have an obligation to understand how things work.

⌥ Permalink

❌