AvBrief Grows Again Using Human Intelligence

I love Square. In my other life, Square’s payment processing system is the foundation of our entire inventory, tax compliance, and point of sale systems. It’s simple, yet comprehensive. It’s intuitive but so incredibly smart and logical, and it produces readable reports that let me track every drop of wine and its profit and loss.

So it was with interest that I read that AI is going to be doing a lot of that magic in the future and that Square founder Jack Dorsey says he can do it better with 40% fewer people. He’s already made billions of dollars off that announcement as shares went up 24% for the parent company called Blocks. Well, good for him, I guess, but where’s my cut? My little business and the millions of others that have funded his company for the last 15 years will not see a dime of that bonanza in the form of lower processing fees or added features. It’s doubtful we who dutifully collect the 2.75% on credit card transactions and send it to San Francisco will notice much change, if any.

But several people will get much richer while 4,000 former employees wonder how they’ll pay their bills after Blocks’ admittedly generous severance package runs out in the summer. Those of you who know the story of AvBrief know why I empathize with those workers, most of whom I suggest were at the top of their game when they got the email from Dorsey. As firings go, I like theirs better than mine, but all firings suck.

The irony here is that I have no intention of leaving Square on the premise that the firings at Blocks were somehow unjustified. That’s none of my business, and my business needs Square. But it also doesn’t mean that we here at AvBrief are going along with the flow. Today we’re announcing yet another expansion of our coverage with the introduction of Instrument Aviator. The addition of Elaine Kauh, a former Belvoir colleague of Larry’s and mine, will help fill a massive void in aviation media since the loss of the specialty titles she wrote for.

She is among more than 50 contributors who have joined AvBrief in the last seven months who provide us with the kind of informed, balanced, and accurate information we all need as pilots. Instrument Aviator will be adding more writers of Elaine’s caliber and will be the authoritative source for IFR pilots who want to stay at the top of their game.

As mankind wakes up to the extraordinary impact of artificial intelligence, we here at AvBrief continue to soldier on in resistance to what we’re assured is the inevitable. But I don’t believe we are doomed to subservience by our silicon overlords. The evolution of the tech in the scant seven months since we made our declaration of independence from the tyranny of electronic brains has me more convinced than ever that bucking this trend is the right thing to do.

Back in those innocent times on Aug. 3, 2025, when I drew the proverbial red line and declared AvBrief an AI-free zone, my main beef was that AI-produced content was boring, formulaic, and riddled with factual errors. Those are still major factors in our ban, but the threat has evolved.

By monitoring other publications which, according to my AI detector, generously ladle AI content into stories under the bylines of human writers, I’ve come across a kind of death spiral for credibility and accuracy that I can only see being avoided by a blanket ban on AI use here.

The preparation of a real news story that employs all the skills, ethics, and values that underpin journalism is a multi-stage process that relies on the critical thinking skills of the journalist to ensure those standards are upheld. The conceptualization, research, and finally creation of the story employs constant use of checks and cross-checks to verify and validate the facts being presented.

It takes a unique blend of curiosity, tempered by skepticism and even some cynicism, to produce an account of an event or conversation that correctly reflects the content and, perhaps more importantly, the context of that newsworthy set of circumstances. Journalists witness or quote those who witness important things that society in general should be aware of. Indeed, deciding what’s important in the first place is the fundamental skill of a journalist.

You can’t tell me that a computer can duplicate those uniquely human skills, at least not yet. And that’s where it gets complex.

AI is essentially an aggregator of data that it finds on the internet. It may cross-reference facts found there that support the mission its human client has sent it on, but it accepts what it finds as the gospel truth even when it’s not. Then those facts are sent back to the internet where they are available for any AI machine to reiterate as facts over and over again.

One of AI’s most common uses in journalism is the digestion of long and complex documents into a neat summary. Hundreds of pages of text laboriously created by legislators, historians, technicians, and educators are reduced to a bullet-point encapsulation that, from what I’ve seen, is usually a random assortment of passages from the original document spewed into a document that utterly fails to explain what those hundreds of pages actually say.

That’s bad enough, but where it gets more dangerous is when human operators entrust those massive energy-sucking, water-wasting data centers to analyze events and conversations and give opinions on their impact to society. Opinion writing is the apex form of journalism in that it relies on accurate, balanced data for fodder in expressing the value, danger, or malevolence of the events or conversations it describes. When AI attempts this kind of analysis the inevitably flawed conclusions drawn by the machines get thrown back into the pool of data available for future use as fundamental truths.

But man, does AI ever do all that stuff fast. AI is essentially a publication productivity tool that allows far fewer people to fill much more space in much less time than was ever possible a few years ago. Of course, that means publishers can spend a lot less money on newsrooms, and I won’t rule out the possibility that humans will be absent from the process in some publications in the not-too-distant future.

It may have already happened in the form of a weekly community newspaper in a town near Vancouver. No person is willing to take responsibility for but it appears faithfully every week, spreading misinformation dressed up with false and misleading AI-generated photos and videos.

Now, I don’t have anything against saving money, but I do reject the notion that pursuit of my profession should abandon our very reason for existence. Journalists witness history in the making and tell others what they’ve seen. The way it’s shaping up, we’re not only getting computers to do our reading and research for us, we’re letting them decide what’s important in that research and determine how we feel about it. It defines madness in my view.

So, the rule here at AvBrief is simple. Every character of every word that is presented here was put there by a human action, mostly through the interaction of fingers and keyboard. That ensures that human influence governs everything that appears and the humans responsible for its appearance are accountable.

And that’s why we have added more than four dozen human contributors to our roster. I haven’t met some of those contributors, but I’m sure they share the qualities of those I have had the pleasure of collaborating with. Without exception, those contributors are such accomplished, dedicated subject matter experts that every one of them makes me question my life choices.

They are the lifeblood of this collection of publications, and they’ve only just started to show us what they can do and how much they can add to our increasingly diverse and rich publications. Journalism is in good hands in our little corner of the information revolution and it fills me with pride that these fine folks are entrusting us to be the method of distribution for their invaluable contributions to aviation.

But that doesn’t mean any of us (least of all me) is perfect. From time to time you’ll see a correction in AvBrief and you should be very glad when you do. It means that despite our attempts to be accurate and thorough in our presentation of the news, we screwed up and are not only taking responsibility for the error, we are anxious to make sure you know we steered you wrong.

AI will never do that, just like it will never end a sentence with a preposition as I’ve done numerous times in this screed. It will never accept that it’s wrong on anything because, after all, everything it knows it learned from the internet.