The Sad Story of the Long-Promised 3 GHz Power Mac G5 ⇥ 512pixels.net

With just a few days to go until WWDC, Stephen Hackett chronicled the history of the mythical 3 GHz Power Mac G5, previewed at the conference over twenty years ago:

The heat generated by the G5 would become an issue later on, but first we need to talk about Steve Jobs broke a fundamental rule of Apple:

We don’t comment on future products.

The Power Mac G5 was before my time as a Mac user, but I remember the otherworldly impression it made on me. I am no Mac historian, but it still feels to me like it began a distinct era of flagship Macs that were both powerful and relatively accessible — the inflation-adjusted price of the best G5 spec, without build-to-order upgrades, is nearly $2,000 less than the base price of today’s Mac Pro.

Microsoft Redesigns Office Portal to Prioritize Copilot ⇥ reddit.com

I, like millions of people, spend a good chunk of my Mondays through Fridays in Microsoft 365 (formerly Office) applications, many of which I partly or wholly access through a browser. For years, I could just type portal.office.com in my browser and get links to Word, Excel, Outlook — all the classics — plus some organization-specific applications.

Microsoft being Microsoft, it has decided to bury all that. We are in the artificial intelligence era now, baby. Instead of seeing the applications millions of people rely on, Microsoft has decided the default view should be a Copilot chat. The familiar suite of apps is in another page accessible from the sidebar; each app can, at least, be pinned to that same sidebar. This was barely announced and seems to have taken many by surprise.

But, hey, Copilot should at least make me more productive, right? It is Microsoft’s own service — albeit backed by OpenAI’s work — and integrated, in theory, with our organization’s applications and files. Maybe it could function as an application launcher. So I told Copilot to “launch Word”. Copilot:

I can’t directly launch Microsoft Word from here, but I can help you open it on your device. Here’s how you can do it:

On Windows:

  1. Press the Windows key on your keyboard or click the Start menu.

  2. Type “Word” in the search bar.

  3. Click on Microsoft Word from the search results.

On macOS:

  1. Open Finder or press Command + Space to open Spotlight Search.

  2. Type “Word” and press Enter when Microsoft Word appears.

Truly embarrassing stuff to interrupt users’ learned workflow for something worse because it is important for Microsoft to be seen as a cutting-edge innovator. It has made Copilot part of its branding — so much so that it will advertise Copilot even where it is not available, for “consistency”.

Texas Will Soon Be the Ninth State to Pass a Popular ‘Right to Repair’ Law ⇥ techdirt.com

Richard Lawler, the Verge:

Texas moved closer to becoming the next state with a right to repair law on the books, as the state Senate unanimously voted 31 – 0 to finalize HB 2963 this weekend. It would require manufacturers to make spare parts, manuals, and necessary tools available for equipment sold or used in the country’s second most populated state.

Maddie Stone, Grist:

That [if Ohio’s bill becomes law] would mark an important political inflection point for the right-to-repair movement. While most of the states that have passed repair laws so far are Democratic strongholds, bills have been introduced in all 50 as of February. The adoption of a right-to-repair law in deep red Ohio — where Republicans control the state House, Senate, and the governor’s office, and Donald Trump won the last presidential election by more than 10 percentage points — would further underscore the broad, bipartisan popularity of being allowed to fix the stuff you own.

Karl Bode, TechDirt:

This is definite progress, and I hate to be a buzzkill, but it’s worth reiterating that of the 8 states that have passed right to repair reforms so far, not a single one has actually enforced them in any meaningful way. This despite no shortage of bad corporate actors working overtime to kill independent repair shops, make manuals and parts hard to find, use obnoxious DRM, or claim that repairing your own stuff violates warranty.

The problem with corporations insisting upon the safety and reliability that only comes from staying within their own ecosystem of parts and accessories is that they are frequently lying or, at least, exaggerating about the dangers. At least lawmakers across the political spectrum and around the world are beginning to call them on their tall tales.

Canadian Liberal Party Border Bill Also Includes Warrantless Access Provisions ⇥ michaelgeist.ca

Michael Geist:

The government yesterday introduced the Strong Border Act (Bill C-2), legislation that was promoted as establishing new border measure provisions presumably designed to address U.S. concerns regarding the border. Yet buried toward the end of the bill are lawful access provisions that have nothing to do with the border. Those provisions, which raise the prospect of warrantless access to information about Internet subscribers, establish new global production orders of subscriber information, and envision new levels of access to data held by electronic service providers, mark the latest attempt in a longstanding campaign by Canadian law enforcement for lawful access legislation. Stymied by the Supreme Court of Canada (which has ruled that there is a reasonable expectation of privacy in subscriber data) and by repeated failures to present a compelling evidentiary case for warrantless access, law enforcement has instead tried to frame lawful access as essential to address everything from organized crime to cyber-bullying to (now) border safety. […]

As has become frustratingly typical for laws of this type, there are secrecy requirements and warrant workarounds based on urgency. But it is tucked into an unrelated bill that only exists because of provocations from our southern neighbour. This is a completely dishonest runaround to legislate long-desired capabilities.

Update: Jeremy Appel, the Orchard:

I have a Liberal campaign flyer pinned to my office wall, where I place election campaign literature I accumulate from various elections, which reads: Making life more affordable, protecting your rights and standing up to Trump.

This legislation doesn’t do any of these things, and in fact does the opposite of two of the three.

Profound weakness masquerading as strength and confidence.

The Who Cares Era ⇥ dansinker.com

Dan Sinker:

Hanif Abdurraqib, in one of his excellent Instagram mini-essays the other week, wrote about the rise of content that’s designed to be consumed while doing something else. In Hanif’s case, he was writing about Time Machine, his incredible 90 minute deep dive into The Fugees’ seminal album The Score. Released in 2021, Hanif marveled at the budget, time, and effort that went into crafting the two-part 90 minute podcast and how, today, there’s no way it would have happened.

He’s right. Nobody’s funding that kind of work right now, because nobody cares.

(It’s worth pointing out that Hanif wrote this using Stories, a system that erased it 24 hours later. Another victim of the Who Cares Era.)

This does not seem like it is a brand new feeling. It is plausible that one could create a Paul Fairie-esque collection of similar complaints throughout history — of art created only for commerce’s sake; of shallow media intended for brief consumption instead of a lasting impression.

I think it is more useful to de-emphasize the “era” part of Sinker’s criticism, and to instead consider how carelessness creeps in. Sarah Wynn-Williams published “Careless People” earlier this year. I wrote in April about Perplexity’s carelessness of ethics and privacy. Perhaps both of these examples are better described as corrosion or corruption. What Sinker is describing is more like cultural and societal ennui.

Les Orchard:

The user isn’t the customer. And they’re not the product either. The real product is behavioral optimization — metrics on a dashboard. The paying customer is somewhere else entirely, and the “content” is just a means to nudge behavior and juice KPIs.

There have always been measures for success, but the ability to relentlessly optimize for them is a fairly new phenomenon. Plenty of films and plays are created by reverse-engineering the tropes favoured by award committees, but they still need to be good in order to draw an audience.

Update: A lack of care is, in part, downstream of other forces (previously linked).

How the V60 Became the Coffee Brewer of Choice for Many ⇥ youtube.com

I am a daily user of a V60, so this new James Hoffmann video scratched an itch I did not know I had. Not to spoil it, but the V60 does not make a significant appearance until the last four minutes, but the journey is worth the wait.

The discussion of the mid-2000s specialty coffee movement in Canada brought back some memories, too. Vancouver was the most significant place in the country and perhaps one of a few places in the world where this attitude toward coffee was taking over, and Calgary luckily received some of the reflected glory. There are shots of menus from 2006 with by-the-cup brewing for two or three dollars, around the same time Phil & Sebastian was getting started here. There is a note at the bottom of one of the menus from Elysian advertising advance purchases of a forthcoming award-winning coffee at the then-devastating price of ten dollars per cup.

Turns out twenty years is a long time.

Meta and Yandex Apps on Android Have Been Tracking Users in Newly Creepy Ways ⇥ localmess.github.io

Aniketh Girish, et al.:

We disclose a novel tracking method by Meta and Yandex potentially affecting billions of Android users. We found that native Android apps — including Facebook, Instagram, and several Yandex apps including Maps and Browser — silently listen on fixed local ports for tracking purposes.

Dan Goodin, Ars Technica:

The covert tracking — implemented in the Meta Pixel and Yandex Metrica trackers — allows Meta and Yandex to bypass core security and privacy protections provided by both the Android operating system and browsers that run on it. Android sandboxing, for instance, isolates processes to prevent them from interacting with the OS and any other app installed on the device, cutting off access to sensitive data or privileged system resources. Defenses such as state partitioning and storage partitioning, which are built into all major browsers, store site cookies and other data associated with a website in containers that are unique to every top-level website domain to ensure they’re off-limits for every other site.

The difference between targeted advertising and spyware is there is no difference.

After Girish, et al., disclosed this behaviour, Meta’s apps ceased tracking users with this method, and Goodin said Yandex will also stop. Meta is still under a consent decree struck in 2019 with a $5 billion penalty after violating a 2012 agreement. Executives at Meta do not care about privacy, rules, laws, or common sense. They will keep doing stuff like this. Ad tech is an indefensible industry run by megalomaniacs who would better serve society if they were made to live in a cave under an ice sheet, though I do not care which one.

A.I. Calvin and Hobbes Strip ⇥ social.lol

This, from Adam Newbold, is a perfect encapsulation of a bunch of ethical problems related to artificial intelligence. The prompt:

Generate an image for a Calvin & Hobbes strip. Four panels. Calvin and Hobbes are walking through the woods, talking to each other, both holding smart phones and looking at them intently the entire time.

Panel 1: Calvin says to Hobbes, “This strip was made entirely with ChatGPT, which should be impossible given the strict intellectual property rights restrictions on Calvin & Hobbes content.”

Panel 2: Hobbes responds to Calvin, “Oh? Then how did it make it?”

Panel 3: Calvin responds to Hobbes, “Some guy just typed this into a box and clicked a button. That’s all it took.”

Panel 4: Hobbes responds to Calvin, “That’s so fucked up.”

This is entirely doable without generative artificial intelligence, but it requires far more skill. The ease of this duplication is maddening. I find this offensive in exactly the way Newbold intended it to be.

More important, I think, is the control exercised over the likenesses of Calvin and Hobbes by the strip’s creator Bill Watterson, as Newbold noted in the strip. Watterson famously rejected all but a handful of licensed merchandising ideas. But the mechanism for how he might protect this is the same as the one used by Disney when it fights parody and reinterpretation of its vast intellectual property, even though the motivations are different. Watterson’s protective quality is admirable, driven by artistic integrity to the extent he has left many millions of dollars’ worth of tchotchkes on the table to retain the spirit of the strips. Disney’s is entirely business motivated, evidenced by the tens of billions of dollars in licensed tchotchkes sold last year alone.

This is not the first “Calvin & Hobbes” strip made with generative A.I., nor does generative A.I. begin and end at self-referential prompts like these. Some assholes have created plugins — more-or-less — to badly emulate Watterson’s unique style in generative A.I. programs. It is awful.

I want to live in a world where we can differentiate between the necessary reinterpretation of intellectual property while respecting the wishes of artists. This is a tricky line, I know. It requires us — individually, but also the organizations responsible for generative A.I. stuff — to think about who is making such a request in good faith, and decide whether we are going to honour that.

One more thing: Watterson is a pretty private person, rarely giving interviews. But, right above this paragraph, I think we can get a sense of how he might feel about this.

A.I. Answer Engines Are Worth Trying ⇥ tidbits.com

Adam Engst, TidBits:

To distinguish these tools [Perplexity and ChatGPT’s web search] from traditional search engines like Google and Bing, I’m calling them “answer engines.” Although they are performing live Web searches for you, the focus is on answering your question rather than displaying the results of the search. I see answer engines as the next step in networked knowledge acquisition because they fundamentally change how we find and absorb information online:

[…]

The key benefit of answer engines is that they typically provide you with exactly the information you want, with no additional effort required. They’re a bit like Wikipedia in this way—you could fact-check statements with the listed sources in a Wikipedia article, and you can often learn more by following links to related topics, but most of the time, you’re happy to read the article and move on.

I think Engst is, in an ideal world, correct — A.I. search is positioned as a way to get straight to the answer. It, theoretically, cuts through a web poisoned by bait for search engines, marketing copy disguised as information, and thin articles suffocated by ads.

So far, I have rarely been satisfied by the results I have received. It has never been satisfactory for someone to say “according to Google” when asked how they know a piece of information, and the deadpan way answer engines frame their responses gives them an unearned authority. I have found it necessary to double check everything. I recently documented ChatGPT’s answer with an invented report with an invented quote that led me to the results I was looking for after some web searches. I have also posted screenshots from Google’s A.I. answer engine on Mastodon. It made up an Excel option and cited it to Microsoft’s documentation; it claimed Noah Wyle co-starred with Alex Horne in “The Pitt”. In fairness, I would like to see that.

Even so, these answer engines have been helpful for me as starting points. It can sometimes be difficult to know what search terms to use, even if you can formulate in your head the question or type of answer you are hoping to receive. I have found this particularly helpful with web development questions. They are tools that can support someone on a research journey. But we should remain skeptical that some trained black box of mystery is able to directly answer questions accurately. We must check, and re-check. Our ability to use these products depends on a well-developed skill for research.

Sponsor: Magic Lasso Adblock: Incredibly Private and Secure Safari Web Browsing ⇥ magiclasso.co

Online privacy isn’t just something you should be hoping for – it’s something you should expect. You should ensure your browsing history stays private and is not harvested by ad networks.

Magic Lasso Adblock: No ads, no trackers, no annoyances, no worries

By blocking ad trackers, Magic Lasso Adblock stops you being followed by ads around the web.

As an efficient, high performance and native Safari ad blocker, Magic Lasso blocks all intrusive ads, trackers, and annoyances on your iPhone, iPad, and Mac. And it’s been designed from the ground up to protect your privacy.

Users rely on Magic Lasso Adblock to:

  • Remove ad trackers, annoyances, and background crypto-mining scripts

  • Browse common websites 2.0× faster

  • Block all YouTube ads, including pre-roll video ads

  • Double battery life during heavy web browsing

  • Lower data usage when on the go

With over 5,000 five star reviews; it’s simply the best ad blocker for your iPhone, iPad, and Mac.

And unlike some other ad blockers, Magic Lasso Adblock respects your privacy, doesn’t accept payment from advertisers and is 100% supported by its community of users.

So, join over 350,000 users and download Magic Lasso Adblock today.

On Browsers, A.I., and the Web ⇥ manuelmoreale.com

Manuel Moreale:

Now, while I’m here let me tackle a somewhat related issue I’ve been thinking about lately and that is the idea that AI is going to become the default layer between users and the web. We all yelled and screamed because the web has too many gatekeepers, we all lamented Google search results going to shit, and we all celebrated when new search engines were coming up. Why would I be happy trading a search result page filled with links — even if ranked in a flawed way — for a block of text that gives me an opinionated answer and maybe some links?

A fantastic question. One could easily ask the same about why we would trade the vast surveillance of online advertising for the universal surveillance gestured toward by Perplexity, too. It is all familiar, but more — in ways we might appreciate, and ways we loathe.

Why Take9 Will Not Improve Cybersecurity ⇥ darkreading.com

Bruce Schneier and Arun Vishwanath, Dark Reading:

There’s a new cybersecurity awareness campaign: Take9. The idea is that people — you, me, everyone — should just pause for nine seconds and think more about the link they are planning to click on, the file they are planning to download, or whatever it is they are planning to share.

There’s a website — of course — and a video, well-produced and scary. But the campaign won’t do much to improve cybersecurity. The advice isn’t reasonable, it won’t make either individuals or nations appreciably safer, and it deflects blame from the real causes of our cyberspace insecurities.

We will always struggle to accurately identify malicious emails because that is the whole point of them. When criminals are able to scam those who are professionals in this field — people like a LastPass system engineer, Jim Browning, and Troy Hunt — it is clear we cannot expect an average user to be a so-called human firewall.

Artificial Intelligence Should Bear Responsibility for Its Costs ⇥ drewdevault.com

Drew DeVault:

Now it’s LLMs. If you think these crawlers respect robots.txt then you are several assumptions of good faith removed from reality. These bots crawl everything they can find, robots.txt be damned, including expensive endpoints like git blame, every page of every git log, and every commit in every repo, and they do so using random User-Agents that overlap with end-users and come from tens of thousands of IP addresses – mostly residential, in unrelated subnets, each one making no more than one HTTP request over any time period we tried to measure – actively and maliciously adapting and blending in with end-user traffic and avoiding attempts to characterize their behavior or block their traffic.

As curious and fascinating as I find many applications of generative artificial intelligence, I find it difficult to square with the flagrantly unethical way it has been trained. Server admins have to endure and pay for massive amounts of traffic from well-funded corporations, without compensation, all of which treat robots.txt as something to be worked around. Add to that the kind of copyright infringement that would cost users thousands of dollars per file, and it is clear the whole system is morally bankrupt.

Do not get me wrong — existing intellectual property law is in desperate need of reform. Big, powerful corporations have screwed us all over by extending copyright terms. In Canada, the number of works in the public domain will be stagnant for the next eighteen years after we signed onto the Canada–United States–Mexico Agreement. But what artificial intelligence training is proposing is a worst-of-both-worlds situation, in which some big businesses get to retain a tight grip on artists’ works, and others get to assume anything remotely public is theirs to seize.

Sky Extends A.I. Automation to Your Entire Mac ⇥ macstories.net

Federico Viticci, MacStories:

For the past two weeks, I’ve been able to use Sky, the new app from the people behind Shortcuts who left Apple two years ago. As soon as I saw a demo, I felt the same way I did about Editorial, Workflow, and Shortcuts: I knew Sky was going to fundamentally change how I think about my macOS workflow and the role of automation in my everyday tasks.

Only this time, because of AI and LLMs, Sky is more intuitive than all those apps and requires a different approach, as I will explain in this exclusive preview story ahead of a full review of the app later this year.

Matthew Cassinelli has also been using an early version of Sky:

Sky bridges the gap between old-school scripting, modern automation, and new-age LLM technology, built with a deep love for working on the Mac as a platform.

This feels like the so-far-unfulfilled promise of Apple Intelligence — but more. The ways I want to automate iOS are limited. But the kinds of things I want help with on my Mac are boundless. Viticci shares the example of automatically sorting a disorganized folder in Finder, and that is absolutely something I want to do easier than I currently can. Yes, I could cobble together something with AppleScript or an Automator workflow, but it would be so much nicer if I could just tell my computer to do something in the most natural language I understand. This is fascinating.

Judge Dismisses 2021 Rumble Antitrust Suit Against Google on Statute of Limitations Grounds ⇥ reuters.com

Mike Scarcella, Reuters:

Alphabet’s Google has persuaded a federal judge in California to reject a lawsuit from video platform Rumble accusing the technology giant of illegally monopolizing the online video-sharing market.

In a ruling on Wednesday, U.S. District Judge Haywood Gilliam Jr said Rumble’s 2021 lawsuit seeking more than $2 billion in damages was untimely filed outside the four-year statute of limitations for antitrust claims.

Rumble is dishonest and irritating, but I thought its case in which it argued Google engages in self-preferencing could be interesting. It seems to rank YouTube videos more highly than those from other sources. This can be explained by YouTube’s overwhelming popularity — it consistently ranks in the top ten web services according to Cloudflare — yet I can see anyone’s discomfort in taking Google’s word for it, since it has misrepresented its ranking criteria.

This is an unsatisfying outcome, but it seems Rumble has another suit it is still litigating.

Google is Burying the Web Alive ⇥ nymag.com

John Herrman, New York magazine:

But I also don’t want to assume Google knows exactly how this stuff will play out for Google, much less what it will actually mean for millions of websites, and their visitors, if Google stops sending as many people beyond its results pages. Google’s push into productizing generative AI is substantially fear-driven, faith-based, and informed by the actions of competitors that are far less invested in and dependent on the vast collection of behaviors — websites full of content authentic and inauthentic, volunteer and commercial, social and antisocial, archival and up-to-date — that make up what’s left of the web and have far less to lose. […]

Very nearly since it launched, Google has attempted to answer users’ questions as immediately as possible. It had the “I’m Feeling Lucky” button since it was still a stanford.edu subdomain, and it has since steadily changed the results page to more directly respond to queries. But this seems entirely different — a way to benefit from Google’s decades-long ingestion of the web and giving almost nothing back. Or, perhaps, giving back something ultimately worse: invented answers users cannot trust, and will struggle to check because sources are intermingled and buried.

The CIA’s 2010s Covert Communication Websites ⇥ ourbigbook.com

Ciro Santilli:

This article is about covert agent communication channel websites used by the CIA in many countries from the late 2000s until the early 2010s, when they were uncovered by counter intelligence of the targeted countries circa 2010-2013.

This is a pretty clever scheme in theory, but seems to have been pretty sloppy in practice. That is, many of the sites seem to share enough elements allowing an enterprising person to link the seemingly unrelated sites — even, as it turns out, years later and after they have been pulled offline. That apparently resulted in the deaths of, according to Foreign Policy, dozens of people.

Apple Gets Its Annual Fraud Prevention Headlines ⇥ apple.com

Apple issued a news release today touting the safety of the App Store, dutifully covered without context by outlets like 9to5Mac, AppleInsider, and MacRumors. This has become an annual tradition in trying to convince people — specifically, developers and regulators — of the wisdom of allowing native software to be distributed for iOS only through the App Store. Apple published similar stats in 2021, 2022, 2023, and 2024, reflecting the company’s efforts in each preceding year. Each contains similar figures; for example:

  • In its new report, Apple says it “terminated more than 146,000 developer accounts over fraud concerns” in 2024, an increase from 118,000 in 2023, which itself was a decrease from 428,000 in 2022. Apple said the decrease between 2022 and 2023 was “thanks to continued improvements to prevent the creation of potentially fraudulent accounts in the first place”. Does the increase in 2024 reflect poorer initial anti-fraud controls, or an increase in fraud attempts? Is it possible to know either way?

  • Apple says it deactivated “nearly 129 million customer accounts” in 2024, a significant decrease from deactivating 374 million the year prior. However, it blocked 711 million account creations in 2024, which is several times greater than the 153 million blocked in the year before. Compare to 2022, when it disabled 282 million accounts and prevented the creation of 198 million potentially fraudulent accounts. In 2021, the same numbers were 170 million and 118 million; in 2020, 244 million and 424 million. These numbers are all over the place.

  • A new statistic Apple is publishing this year is “illicit app distribution”. It says that, in the past month, it “stopped nearly 4.6 million attempts to install or launch apps distributed illicitly outside the App Store or approved third-party marketplaces”. These are not necessarily fraudulent, pirated, or otherwise untoward apps. This statistic is basically a reflection of the control maintained by Apple over iOS regardless of user intentions.

There are plenty of numbers just like these in Apple’s press release. They all look impressive in large part because just about any statistic would be at Apple’s scale. Apple is also undeniably using the App Store to act as a fraud reduction filter, with mixed results. I do not expect a 100% success rate, but I still do not know how much can be gleaned from context-free numbers.

Lawyers Keep Failing Clients By Relying on A.I. ⇥ theguardian.com

Nicholas Chrastil, the Guardian:

State officials have praised Butler Snow for its experience in defending prison cases – and specifically William Lunsford, head of the constitutional and civil rights litigation practice group at the firm. But now the firm is facing sanctions by the federal judge overseeing Johnson’s case after an attorney at the firm, working with Lunsford, cited cases generated by artificial intelligence – which turned out not to exist.

It is one of a growing number of instances in which attorneys around the country have faced consequences for including false, AI-generated information in official legal filings. A database attempting to track the prevalence of the cases has identified 106 instances around the globe in which courts have found “AI hallucinations” in court documents.

The database is now up to 120 cases, including some fairly high-profile ones like that against Timothy Burke.

Here is a little behind-the-scenes from this weekend’s piece about “nimble fingers” and Apple’s supply chain. The claim, as framed by Tripp Mickle, in the New York Times, is that “[y]oung Chinese women have small fingers, and that has made them a valuable contributor to iPhone production because they are more nimble at installing screws and other miniature parts”. This sounded suspicious to me because I thought about it for five seconds. There are other countries where small objects are carefully assembled by hand, for example, and attributing a characteristic like “small fingers” to hundreds of millions of “young Chinese women” seems reductive, to put it mildly. But this assumption had to come from somewhere, especially since Patrick McGee also mentioned it.

So I used both DuckDuckGo and Google to search for relevant keywords within a date range of the last fifteen years and excluding the past month or so. I could not quickly find anything of relevance; both thought I was looking for smartphones for use with small hands. So I thought this might be a good time to try ChatGPT. It immediately returned a quote from a 2014 report from an international labour organization, but did not tell me the title of the report or give me a link. I asked it for the title. ChatGPT responded it was actually a 2012 report that mentioned “nimble fingers” of young women being valuable, and gave me the title. But when I found copies of the report, there was no such quote or anything remotely relevant. I did, however, get the phrase “nimble fingers”, which sent me down the correct search path to finding articles documenting this longstanding prejudice.

Whether because of time crunch or laziness, it baffles me how law firms charging as much as they do have repeatedly failed to verify the claims generated by artificial intelligence tools.