Month: June 2022

Taste, Figures, Images ⇥ spikeartmagazine.com

Dean Kissick, Spike:

Images drawn by code make the world seem lighter and less binding still. Reality is concealed below signs that point nowhere: there’s no such thing as a Tubby Cat, not in life or in fiction. Rembrandt never painted raccoons. He never saw a raccoon in his life! These images are not simulacra because they don’t represent or imitate anything. The new modes of figuration don’t refer to anything at all. They are pictures from somewhere else. They are garbled whispers of code in the fall. Containing no meaning, more empty than a black square.

In, as they say, the before times — before all the galleries had to close their doors to visitors — I was an occasional exhibiting artist, and this piece hit home. One of the odder things that happened in the past two years has been this rise in the quality and believability of images generated by machine learning. So far, it has mostly been an outlet for shallow lowercase-“s” surrealism and pop culture jokes. There is nothing wrong with either of those things, but it is still early days for it as an artistic medium — and it shows.

AirDrop Helper Makes Use of MacBook Notches ⇥ twitter.com

Ian Keen built a brilliant little utility which transforms the notch area of a recent MacBook Pro or MacBook Air model into an AirDrop collector bin. Keen even made a variant for notchless Macs. It is not yet publicly available, but I would use this in a heartbeat. I AirDrop stuff all the time and the sharing menu feels comparatively clunky.

Struggling to think of a great phrase to describe this kind of first-rate, tip-top effort, but it will come to me.

Nobody Knows How to Watch Movies Anymore ⇥ wired.com

Jason Kehe, Wired:

[…] That’s why simpleminded arguments like “our attention spans are shot” are so rarely, on their own, convincing. You simply attend to different things these days, like TV, or TikTok. (Worse things, some say, less unified, less artistic, but to an alien it looks like complete attention all the same.) In 2022, there’s something uniquely daunting about the prospect of committing to a movie, even for just 90 minutes. So you scroll and scroll and scroll, never quite ready to make a decision, aware, on some level, that you lack the strength to see it through.

Maybe this doesn’t bother you. Movies are a dying art form; TV is ascendant! I suspect, however, it does. The less you watch movies, the more you miss them. You miss the completeness of them, of a story fully told — something TV (or TikTok, neverending) almost never provides. A movie is designed, after all, to be watched all at once, its rhythms and pacing serving the arc of a single emotional journey.

Or at least they used to. Blockbusters are now more than ever shoehorned into never-ending story arcs and expanding universes of characters.

Update: Elamin Abdelmahmoud, Buzzfeed News:

Are you exhausted yet? We seem to have arrived at the nadir of original stories, a cultural moment where many of the TV shows and movies feature names and characters we already know. It does not matter how well we know them — it just matters that the audience is already familiar with the world. We are living through the age of peak intellectual property. Hollywood has learned the safe route, found a reliable pattern: Every time studios push this button, $13 comes out. Why wouldn’t they keep pushing it? But at what cost to originality?

I know plenty of people have complained about the dearth of original cinema for years, but it feels like we have hit a new low point in the risks studios are willing to take. Maybe that is because so many of these movies and shows are inherently expensive to make, with marquee actors and thousands of visual effects shots, so studios demand the closest thing they can get to guaranteed success.

Facebook Plans ‘Discovery Engine’ Feed Change to Compete With TikTok ⇥ theverge.com

Alex Heath, the Verge:

In an internal memo from late April obtained by The Verge, the Meta executive in charge of Facebook, Tom Alison, spelled out the plan: rather than prioritize posts from accounts people follow, Facebook’s main feed will, like TikTok, start heavily recommending posts regardless of where they come from. And years after Messenger and Facebook split up as separate apps, the two will be brought back together, mimicking TikTok’s messaging functionality.

Combined with an increasing emphasis on Reels, the planned changes show how forcibly Meta is responding to the rise of TikTok, which has quickly become a legitimate challenger to its dominance in social media. While Instagram has already morphed to look more like TikTok with its focus on Reels, executives hope that a similar treatment to Facebook will reverse the app’s stagnant growth and potentially lure back young people.

Between now and theoretically mainstreaming the metaverse concept, it is interesting to see Meta’s flagship product flail around as it attempts to establish an identity beyond how your parents complain about kids these days. Regardless of your age, does it sound compelling to transform your feed into an infinitely scrolling entertainment and news firehose? Is there room for a bunch of similar networks thirsty for your indiscriminate attention?

Teslas Running Autopilot Involved in at Least 273 Crashes in Past Year ⇥ washingtonpost.com

Here is a little update on that NHTSA investigation into Tesla’s autonomous vehicle systems.

Faiz Siddiqui, Rachel Lerman, and Jeremy B. Merrill, Washington Post:

Tesla vehicles running its Autopilot software have been involved in 273 reported crashes over roughly the past year, according to regulators, far more than previously known and providing concrete evidence regarding the real-world performance of its futuristic features.

It is not particularly notable to have under three hundred crashes in cars made by a single automaker selling hundreds of thousands of vehicles in the U.S. in 2021. But if all these crashes occurred while the cars were running Autopilot — more on that “if” below — that means there was a crash, on average, three out of every four days from June 2021 through June 2022 by the same driver.

Here is the “if”:

The new data set stems from a federal order last summer requiring automakers to report crashes involving driver assistance to assess whether the technology presented safety risks. Tesla‘s vehicles have been found to shut off the advanced driver-assistance system, Autopilot, around one second before impact, according to the regulators.

I am not the type to explore conspiracy theories, but it does not look good for Tesla if Autopilot disengages immediately before a crash. If this summary is true, I hope there is a legitimate explanation for it.

The Money Powering the Servant Economy Is Drying Up ⇥ ft.com

Sarah O’Connor, Financial Times:

For some critics, the growth of this new “servant economy” is a symptom of resurgent economic inequality and an underclass without better options. But there is another factor that has powered its rise: investors have been subsidising consumers by funding companies that often charge less for these services than it costs to provide them.

Now that model is in jeopardy. The big problem is that the money is drying up. A decade of cheap money has given way to high inflation, gloomy growth forecasts and higher interest rates. Investors are beginning to get nervous about piling money into lossmaking companies. Shares in listed companies such as Uber, Lyft and Deliveroo have dropped sharply.

These companies have sustained losses far longer and greater than any boring delivery or taxicab business could. In their yawning crater of disruption lays the careers of taxi drivers.

If a business cannot survive without limitless infusions of investor cash and low-interest loans, is it really a business? Is it competing fairly?

CMA Report on Mobile Ecosystems ⇥ theplatformlaw.blog

Dimitrios Katsifis, writing in the Platform Law Blog, published by a British–Belgian law firm:

On 10 June 2022, the UK Competition and Markets Authority (CMA) published its Final Report on its year-long market study into mobile ecosystems — namely mobile operating systems, app stores, and web browsers. The CMA found that Apple and Google have a tight grip over these increasingly crucial ecosystems, which in turn places them in a very powerful position. As a result, thousands of businesses which rely on these ecosystems to reach their users face restrictions and terms which they have little choice but to accept, while consumers are likely to miss out on new innovations, have less choice, and ultimately face higher prices. In response, the CMA has identified a wide range of potential interventions that could help unlock competition and protect millions of businesses and people.

Even so, ecosystem operators — and Apple in particular — have fiercely opposed any intervention on the part of the CMA, arguing that this would compromise user privacy, security, and safety. […]

Of course there are times when real privacy and security risks conflict with permissive on-platform competition. But the CMA identified several areas where it says Apple and Google overstated those concerns in defence of their platform rules choices. That is a bit of a case of the boy crying “wolf!”. It is difficult to believe platform operators are always making these decisions only in the name of privacy and security when there are conflicts of interest in their own business lines.

Ideally, platform owners should be making decisions about what to allow in their ecosystem because regulators should not be micromanaging.

One other thing; Katsifis:

First, the choice architecture for the ATT prompt — which Apple chose without conducting any user testing — may not maximise user comprehension and thus could unduly influence some users to opt out of data sharing. Among others, the framing of the prompt could result in limited user comprehension, while Apple bars developers from offering any incentives for users to opt in to sharing their data (which in principle is not unlawful under UK privacy legislation).

Permitting tracking and data collection by coercion seems ethically fraught to me. How hard is it to not be creepy? At this point, just wipe the entire digital advertising and data broker industry off the face of the Earth and start from scratch.

Google Says It Bans Gun Ads, but It Will Still Place Them on Websites and in Apps ⇥ propublica.org

Craig Silverman and Ruth Talbot, ProPublica:

For roughly two decades, Google has boasted that it doesn’t accept gun ads, a reflection of its values and culture. But a ProPublica analysis shows that before and after mass shootings in May at a New York grocery store and a Texas elementary school, millions of ads from the some of the nation’s largest firearms makers flowed through Google’s ad systems and onto websites and apps — in some cases without the site or app owners’ knowledge and in violation of their policies.

[…]

In reality, Google has two sets of rules for weapons ads. One is for Google Ads, the ads that run on the company’s own ad network and on properties it owns, such as YouTube or Google.com search results. The other is for ads sold by partners, such as ad exchanges, that place ads using Google’s systems. Ad exchanges enable digital ads to be bought and sold via an automated bidding process. For these partners, Google operates as an “exchange of exchanges” — in which it facilitates the buying and selling of ads on other exchanges — and takes a cut of each ad transaction. Partner exchanges are guided by a set of more permissive rules that allow gun ads to flow through Google’s ad systems.

This looks like way for Google to maintain an arms-length distance from the lucrative markets it nominally claims are too dirty to do business with. I opened a private browser window, went to one of the survivalist websites listed in this article, and it took me exactly one attempt to see an ad for a folding semi-automatic rifle served by Google’s DoubleClick and hosted on a Google server.

Apparent Conflicts of Interest in App Tracking Transparency Rules Targeted by German Antitrust Probe ⇥ techcrunch.com

Natasha Lomas, TechCrunch:

A major privacy feature Apple launched last year, called App Tracking Transparency (ATT) — which requires third-party apps to request permission from iOS users to track their digital activity for ad targeting — is facing another antitrust probe in Europe: Germany’s Federal Cartel Office (FCO) has just announced it’s investigating the framework over concerns that Apple could be breaching competition rules by self-preferencing or creating unfair barriers for other companies.

Last year, France’s antitrust regulator declined to preemptively block Apple from implementing ATT — but said it would be watching how Apple operates in the feature. Poland also opened a probe of the feature at the end of last year.

Germany’s Federal Cartel Office, or Bundeskartellamt, published a version of its press release (PDF) in English.

Shoshana Wodinsky, Gizmodo:

The Bundeskartellamt’s investigation is primarily focused on Apple’s App Tracking Transparency Framework (or ATT, for short). For those unfamiliar, ATT is the set of rules that Apple rolled out as part of iOS 14 that require third-party app developers to ask for permission to track the users that download their apps — and when those users say no, ATT is what shuts off these apps’ access to a slew of valuable user data. You might recognize it from the small window that pops up and asks you whether you want to give an app access to your personal information, and you can respond with “Ask app not to track” or “Allow.”

It appears German authorities are not taking issue with App Tracking Transparency in a general sense but, rather, the conditions Apple has set for whether an app must request permission to track users. Apple says developers must get consent for apps that “[collect] data about end users and [share] it with other companies for purposes of tracking across apps and web sites”. Apple does the former, but it is everything after the “and” it does not do — and German regulators are investigating whether that gives it an advantage.

On its face, this investigation seems ridiculous. The App Tracking Transparency framework is merely intervening in third party apps to confirm whether the user is okay with being tracked in plain language. That sounds fine with me. The digital advertising industry is built on data collected illegitimately, with little consent, and with virtually no regulations. It is long past time for those behaviours to be reeled in.

Apple, as platform owner, gets access to a whole lot more ad targeting information than most individual apps, and because it is all first-party information not shared with anyone else, Apple does not need to ask permission for its ad network to use it. This appears okay: users have agreed to a business relationship with Apple, so of course it gets to use information it knows about users to target ads, and none of this is going to anyone else, so it is not creepy.

However, this advantage is difficult to replicate by any individual app or service. Only other gigantic companies, like Amazon and Facebook, can draw upon a trove of first-party data for ad targeting purposes. I have been arguing for the past year how it looks really bad for Apple to be setting privacy rules on its platform that restrict third-party advertisers while running its own ad network.

The solution here is not to empower others to more easily track users. If Apple is found to be illegally self-preferencing, I would hope it is resolved by allowing the continued operation of App Tracking Transparency while requiring Apple to reduce its first-party advantage.

This gives me a good reason to share my favourite Norm Macdonald joke (contains NSFW language, if that is a concern).

Ahead of Their Time

Marques Brownlee posted an exploration of products that, in his words, were ahead of their time. The main focus are three gadgets, but there are actually five early-to-market products explored in this video:

  • electric cars, generally,

  • Vine,

  • Google Glass,

  • the Samsung Galaxy Camera, and

  • the Motorola Atrix.

These are all product curiosities, but it is this idea of them being “ahead of their time” that I find sort of bizarre on its face. Perhaps the best way to explain what I mean is to explore why all of these products failed, and to look at whether modern-day versions are more successful.

As Brownlee explains, electric cars arrived in the 1800s and were around in various capacities before Tesla launched the Roadster in 2008. GM’s EV-1 was released in the late 1990s with mediocre range on a full charge, and under a lease-only agreement. Less than two thousand were delivered to customers — interestingly, not that much less than the number of Tesla Roadsters sold during its production run — until GM killed the project and scrapped nearly every example. I have a hard time believing the EV-1 would have ever been seen as truly desirable, but its failure was famously complex.

Vine, meanwhile, is far simpler: it died because Twitter, its owner, was spectacularly bad at product strategy. Musical.ly, TikTok’s predecessor, launched in August 2014, TikTok began expanding out of China in September 2016, and Vine was shuttered in October 2016. Vine’s problem was its owner and a lack of vision; it bled users to Instagram, which launched a competing short-form video option, and Twitter made the shutdown worse with a confusing Vine Camera app.

A product being “ahead of its time” implies there is eventually an era where it could have made sense or where its ideas solidify. At the beginning, Brownlee says these products “failed, but just because the world wasn’t ready for them yet” and, at the end of the video, says “Vine walked so TikTok could run”. That seems to be the definition we are working with, and it is what makes the other products Brownlee cited as being “ahead of their time” much more interesting. Here is what he highlights from each:

  • Google Glass was the first augmented reality headset designed to be worn outside of a lab or a gaming context.

  • Samsung Galaxy Camera was a full camera with a smartphone back, so you have all the power of an Android phone with the quality of a compact camera.

  • Motorola Atrix was a dockable phone that effectively transformed into a laptop.

Maybe you have spotted the problem already, but none of these product categories has found present-day success. Google Glass, as Brownlee says, is used in limited industrial and enterprise contexts, and he shows examples of more recent products that follow the form set by the Galaxy Camera and the Atrix. But there is no mainstream successful version of any of these.

There are fundamental reasons Brownlee cites for each, too, that have yet to be resolved. The smartphone-as-laptop concept sounds interesting, but nobody is buying one because the ecosystem is not there yet. Brownlee says Apple would rather sell you many devices rather than one, which is true, but it is also better to have each device format be true to itself. There is no evidence that people find a single device concept compelling in practice. People like their separate phone; they can fiddle with it while using their laptop for meetings.

A powerful camera with a smartphone operating system has also proved to be a weird concept because, in so many words, these are enthusiast problems that have enthusiast solutions. If you really care about camera image quality, you have a dedicated camera. If you want to be able to rapidly share images, it creates an ad hoc Wi-Fi network. Meanwhile, smartphone image quality continues gets better all the time.

Google Glass, meanwhile, was just a really dorky object. Followup products — like Snap’s Spectacles, and the Facebook and Ray-Bans collaboration — are barely more than prototypes. People are not convinced.

This ended up being sort of a retort to Brownlee’s video. That is not necessarily how I intended it; his video did inspire me to think about this a little more. I think we should be cautious in assuming any of the three product formats on his list could eventually succeed with mainstream adoption. Maybe they will, but it is not necessarily because the world was not ready for them yet. But maybe they will not because they are fundamentally flawed ideas — and that is okay. Experimentation is fun. We should not conflate that with early versions of eventually successful products, however, especially when those product categories have not proved successful.

Canadian Government to End Use of COVID Alert App ⇥ theglobeandmail.com

Josh O’Kane, the Globe and Mail:

Ottawa is planning to announce the shutdown of the national COVID-19 contact tracing app this week, months after changes to PCR testing regulations in many provinces had rendered it largely useless across much of Canada.

[…]

The COVID Alert app’s death has been a protracted one. By several metrics, it did not achieve what it was designed to do. Only 6.9 million of Canada’s 38 million residents had downloaded it by last February – which is when Ottawa stopped publishing usage numbers – and Canadians entered only 57,704 codes declaring they had COVID.

COVID Alert is the national exposure notification app built on the troubled joint Apple–Google framework. This is not a success story by any measure, but it is good to see the federal government winding it down to prevent misuse.

Apple Paid the Netherlands €50 Million as It Dragged Its Feet on Legal Compliance ⇥ acm.nl

Apple Developer:

Following productive conversations with the Netherlands Authority for Consumers and Markets (ACM), today we’re introducing additional adjustments to Apple’s plan to comply with the regulator’s order pertaining to dating apps on the App Store in the Netherlands […]

The Dutch Authority for Consumers and Markets is satisfied with these changes:

Apple has changed its unfair conditions, and will now allow different methods of payment in Dutch dating apps. With this concession, Apple will meet the requirements that the Netherlands Authority for Consumers and Markets (ACM) set under European and Dutch competition rules. Until recently, customers of dating apps had only been able to pay using the payment method that Apple imposed. In ACM’s opinion, Apple abused its dominant position with those practices. From now on, dating-app providers are able to let their customers pay in different ways. ACM forced these changes by imposing an order subject to periodic penalty payments. In the end, the sum of all penalty payments totaled 50 million euros.

The changes Apple has made since it announced how it planned to comply with the Dutch regulators’ rules in February have consistently been in the direction of increased ease for developers and a somewhat less petulant response from Apple. I am not sure if there was some technical problem prohibiting Apple from reducing its commission on subscriptions purchased through third-party providers until now, but it looked from the outside like Apple was delaying and minimizing its compliance.

Inside Apple’s iOS 16 Remake of the iPhone’s Iconic Lock Screen ⇥ techradar.com

Lance Ulanoff, TechRadar:

The “magazine look” [Alan] Dye mentioned is more than just the overall composition of Lock Screen elements. It’s that fluff of dog fur or the swell of flowing hair that intersects with the time element and, instead of sitting behind the numbers, layers on top of it. It’s an arresting – and professional – look that’s created automatically. Apple calls it “segmentation.”

Creating this look is something Dye, and his design team dreamed of for years.

“We’ve been wanting to achieve this look, but the segmentation’s gotten so good, that we really feel comfortable putting [in there]. Unless the segmentation is just ridiculously good, it breaks the illusion.”

If you have not installed the iOS 16 beta, trust me on this: this segmentation effect is really really good. I am not saying that all photos I have tried work perfectly, but only a handful of images that seem like they should work have not; most portraits create a fabulous and compelling lock screen. What an upgrade.

Ulanoff:

Dye told us that if the system doesn’t think the photo will look great, it won’t suggest it, a point of care and attention that helps guide the user towards more visually arresting Lock Screens.

This is the only thing that worries me. I am sure Apple has trained its machine learning workflows on a wide array of images, people, animals and so forth. Yet I have a hard time shaking Maciej Cegłowski’s description of machine learning as “money laundering for bias”. Apple’s changes have been thoughtful, as demonstrated in this interview with Dye and Craig Federighi, but I always have that worry about subjective choices made by machines.

The Talk Show Live From WWDC 2022 ⇥ daringfireball.net

John Gruber has published this year’s episode of the Talk Show Live from WWDC. Topics covered include the redesigned and rebranded System Settings application in MacOS Ventura, CarPlay, developer sentiment, and the fun of using computers. As always, Apple’s executives generally stick to the company’s talking points, but there are revealing moments to better understand Apple’s thinking.

It is worth watching the whole video. It has been a tough couple of years, and things like this in-person recording give me hope that maybe the worst of this pandemic is behind us. I hope it feels the same for those who were there, John and his guests, and you.

I Cannot Believe I Am Defending Mark Zuckerberg ⇥ 9to5mac.com

I do not understand this ridiculous post by Chance Miller of 9to5Mac. It is all about how Meta’s PR team digitally removed the manufacturer’s logo from Mark Zuckerberg’s laptop in a promotional photo, likely a MacBook of some kind. It is no big deal and perhaps just a silly post.

But then Miller gets strange referencing that famous photo of Zuckerberg’s desk showing a MacBook Pro with a taped camera and microphone:

An unrelated side-note here: Suckerberg puts tape over the webcam and microphone of his MacBook. The CEO of Facebook, the company with no regard for user privacy, is worried about Apple spying on him through his Mac’s webcam and microphone. And he still uses a MacBook today!

I want to focus on the second sentence: Zuckerberg is almost certainly not worried about Apple spying on his company’s activities, but about anyone spying. The NSA has repeatedly issued guidance (PDF) — as recently as this week — advising people in higher-risk contexts to cover cameras and microphones.

Is it a little ironic that both the NSA and the CEO of Meta closely guard their privacy and security? Sure, but they are also compensating for different risk levels. The average person probably does not need to worry about someone using their microphone to monitor their conversations. The American intelligence apparatus and CEOs are atypical contexts.

I find Miller’s article both painfully cringeworthy and just flat-out wrong. And calling him “Suckerberg” is petty and childish.

iPadOS 16 Adds Memory Swap to M1 iPads ⇥ macrumors.com

Sami Fathi, MacRumors:

Apple is adding virtual memory swap to iPadOS 16, allowing apps on the recent iPad Pro and iPad Air models to use free and available storage as extra memory for demanding workloads.

This sounds to me almost like old fashioned swap memory. If that is truly what is happening, I could not be more excited — and I do not even have an iPad that supports it. I have been one small voice in a chorus of people begging for features that justify the iPad Pro name. Allowing apps to truly remain in memory is a big step forward in transitioning iPadOS from its smartphone origins to something more, as Apple puts it, “desktop class”.

This, as much as the new-ish window management system, are almost convincing enough for me to think about buying a new iPad. Give it a few months and we shall see whether I bite the bullet.

Why it is limited to the highest-end new iPads is perhaps not the biggest mystery, but there does appear to be a minor story within this story. Fathi:

Virtual memory swap and the higher memory ceiling join a list of features coming with iPadOS 16 that are exclusive to the M1 iPad Pro and iPad Air , including Stage Manager. In fact, Apple told Digital Trends that Stage Manager relies on this virtual RAM swap.

It seems Apple probably did offer that rationale, but you will find no such quote in that Digital Trends article today. An unfortunate attribution error and an unsurprising explanation.

The Curious Case of the 13-Inch MacBook Pro ⇥ 512pixels.net

Stephen Hackett:

Some have said that the 13-inch MacBook Pro has stuck around to hold the $1,299 price point, or to give people who want a “Pro” machine an option that doesn’t come with the $1,999 starting price point the 14-inch MacBook Pro has. It’s also (apparently) popular with corporations who buy in bulk.

During the WWDC keynote, John Ternus said the MacBook Air is the world’s best-selling laptop. In second place? Apparently, the 13-inch MacBook Pro — a strange middle-ground laptop. I wonder how many of these things are bought by people or companies who believe the MacBook Air is insufficiently capable based on its branding.

Mark Gurman reported today for Bloomberg that Apple is working on a 15-inch MacBook Air. That sounds like an appealing replacement to me for the 13-inch Pro, and would bring the lineup closer to the classic Jobs quadrant. Then again, Gurman also reported Apple is working on a 12-inch laptop. It is an interesting time to be customer of Apple’s, that is for sure.

NHTSA to More Closely Scrutinize Tesla’s Autopilot Feature ⇥ nytimes.com

After writing about my CarPlay concerns, I had intended to link to this new report yesterday by Cade Metz at the New York Times as a sort of companion:

Autopilot has been on public roads since 2015. General Motors introduced Super Cruise in 2017, and Ford Motor brought out BlueCruise last year. But publicly available data that reliably measures the safety of these technologies is scant. American drivers — whether using these systems or sharing the road with them — are effectively guinea pigs in an experiment whose results have not yet been revealed.

[…]

A year ago, the National Highway Traffic Safety Administration, the government’s auto safety regulator, ordered companies to report potentially serious crashes involving advanced driver-assistance systems along the lines of Autopilot within a day of learning about them. The order said the agency would make the reports public, but it has not yet done so.

The safety agency declined to comment on what information it had collected so far but said in a statement that the data would be released “in the near future.”

I got distracted and forgot all about this link, which actually worked out for the best, because Neal E. Boudette reports today for the Times about some developments in the NHTSA’s investigation of Tesla specifically:

The agency, the National Highway Traffic Safety Administration, said Thursday that it was upgrading its preliminary evaluation of Autopilot to an engineering analysis, a more intensive level of scrutiny that is required before a recall can be ordered.

The analysis will look at whether Autopilot fails to prevent drivers from diverting their attention from the road and engaging in other predictable and risky behavior while using the system.

This sounds like it is limited in scope to understanding if Autopilot should be doing more to prevent owners from treating the system as anything more than a driver assist feature with a misleading name. That is a good start. But I have to wonder how much of drivers’ overconfidence in the system is reflective of the company’s marketing rather than the more subdued way it is presented in the car. It is hard to roll back Elon Musk’s claims that fully autonomous transportation is one to two years out every year for the past decade. I am not even sure it matters whether people believe him. Hearing that repeated time and time again by often sycophantic media outlets must seep into the public consciousness.

Google’s Fabricated Summaries and This Website

The thing about talking about the way Google works now is that some backstory is needed — and, unfortunately, I find this information very dry. You might as well. And, so, I apologize for the next few paragraphs, but I promise it leads somewhere. You just need to trust me a little.

On a search results page where Google shows ten blue links, it can sometimes be difficult to see any difference between its presentation now and how that same page would have looked ten or twenty years ago. Sometimes you will see many more ads and occasionally there are features like “People Also Ask”. But if you are simply looking at web page results, it is hard to see a difference: there is a title, a URL path, a date, and a summary of the page.

A search results page gives the impression of being a semi-neutral index of webpages related to a query, ranked mostly by their quality and relevance, almost like if you were to ask a librarian for books on a specific topic and they returned with ten volumes of substantial authority. These days, it is anything but.

There is an entire industry — mostly comprised of charlatans — dedicated to boosting the ranking of clients’ pages within these results. Sometimes, this can be for nominally ethical reasons; a local restaurant’s own webpage should arguably rank more highly than its Facebook page or Skip the Dishes listing, for example. But this work, no matter how honest its intent, is still manipulation, and it has created a somewhat adversarial relationship between Google and search engine optimization experts.

One thing Google has begun to do to combat misbehaviour is that it often generates page titles and descriptions itself instead of using the ones provided by the webpage. Its criteria for doing so is vague like much of Google’s documentation; the company often says this is because it is trying to avoid tipping off those who may abuse its systems. Any time there is a machine performing this kind of task, there is a risk of introducing untraceable inaccuracies. And here is where this story becomes about the little website you are reading now.

I found my 2017 review of the then-new base model iPad today as I was looking for something related to WWDC announcements this year, after I searched Google for “ipad ram site:pxlnv.com”. The page title appeared in the result as written on the page, but the description was clipped from a section of the article where I specifically describe RAM. And then I noticed something extra weird. Here is a reproduction of that result as it appeared on Google (emphasis theirs):

https://pxlnv.com › blog › 2017-ipad

The 2017 iPad – Pixel Envy

Jul 17, 2017 — Both sizes of iPad Pro also pack 4 GB of RAM. Perhaps the most noticeable difference, though, is that the iPad Pro line is where Apple’s …

Pros and cons: Very fast custom processors ⋅ Pretty great value ⋅ Greater storage options ⋅ View full list

The description has been pulled from what Google believes is a more relevant section of this review. More notably, though, is how Google has generated a “pros and cons” list for this post — and it is wrong.

In the review, I used the phrase “very fast custom processors” only in the context of discussing the iPad Pro line (emphasis added):

[…] Both sizes of iPad Pro represent Apple’s ideals of what tablet-based computing should look like: responsive, big, wide colour displays, very fast custom processors, lots of RAM, and Apple Pencil support.

The phrase “greater storage options” appears twice in the review: once in the context of the iPad Pro, and once as a speculative enhancement of the base iPad (I should edit my writing better):

[…] The Pro models also come with more speakers, newer Touch ID sensors, better cameras — at the expense of a bump — have Apple’s Smart Connector for accessories, and are available with greater storage options. […]

[…]

[…] At some point, the base iPad could conceivably get a newer Touch ID sensor, better cameras, and greater storage options. […]

Both of those are claimed by Google as things I said were qualities of the 2017 base model iPad, but that is not the case for either. (The third phrase, “pretty great value”, is cited correctly in context.) I did not make a list of “pros and cons” anywhere in my review; neither word appears anywhere in its text. But most upsetting is that Google does not make it apparent anywhere on this results page that it is responsible for this description, not me.

If a person quoted me as saying I thought two benefits of the 2017 base model iPad were its “greater storage options” and “very fast custom processors” based on this review, you would question their reading comprehension. If this citation were in something as simple as a high school paper, the student would be docked marks for taking statements out of context. Yet Google is able to do this in an entirely automated way without clearly identifying it as such, and it is supposed to be okay? No, thanks.

It appears this “pros and cons” list is generated for several reviews, though I have not found documentation for it. I have not noticed it until recently, though it is possible it was rolled out as part of last year’s batch of product review ranking updates. I am not sure exactly what causes it to appear for some reviews and not others and, in general, it often generates a fair summary of key points. But that does not compensate for the times it gets things wrong in an invisible way.

I do not monitor my Google search results, nor do I take any particular steps to optimize for them. I have set up my website with Google Search Console to check for broken links and the like. That tool is supposed to help me, as an administrator, manage the way Google indexes my website. I can find nowhere to report this inaccurate summary.

It is not so bothersome that I will spend any more time working on it. The most concerning aspect of this — the thing that got me to write this piece — is how it is not clear from these results that Google is responsible for the contents of these titles and descriptions, nor that they may differ from those specified by the author. This is dishonest and I feel cheated as a search user.