All, Here's an unusual post. This email came to me by my LinkedIn account. I might have missed it amongst other LinkedIn traffic.
It's self-explanatory and worth reading. The author asks that I share it as a response to a recent Mifnet thread. I hope that everyone knows that anyone disagreeing with a post is welcome to offer a response. Mifnet is shared more widely than we may assume. DW Hi David, I was sent a note highlighting a message posted to Mifnet attacking some work I recently published. Eventually, you'll anger someone in this business, but that's life. While I am not a member of Mifnet, I wanted to reach out to you to offer a proper response to ensure the discussion is not one-sided. The response is below. The post in question was referencing Delta and Fetcher's AI model. I don't know who posted the critique, and I respect the discretion of Mifnet. Frankly, I don't want to know. I only ask as a professional courtesy that you post my response to defend my integrity, which was called into question. Many thanks for your help and work toward ensuring transparent discussion and banter. Especially these days. Best regards, Courtney RESPONSE======================= What a strange and oddly angry critique. I'm disappointed I haven't received any of the angry emails to discuss before reading it on a forum, but such is the state of banter and discourse today. It was expected, though: I even wrote about it in the section that appears to have been skipped: "That said, this paper isn’t written specifically for you. As you feel the urge to “well, actually,” consider that explaining the concepts of recurring neural networks and transformers to a population that interacts with all AI solely through a browser chat window requires some reduction." Doesn't take machine learning to predict that one. :) Since we're in aviation and we do everything in acronyms, I think the most elegant response is RTFA. The responses to the critique were already offered in the original paper. 1. Delta’s 2024 passenger revenue was $52 billion. Miller absurdly claims that the Delta/Fletcherr “AI pricing” approach could increase revenue by 10% (“and could be higher than 10%”)... From the paper: "The AI company claims revenue improvements of 10% due to pricing optimization from the model. We do not know if Delta is seeing the claimed 10% in their test cases, but we do know the airline is very, very satisfied with the results." 2. Miller describes the Delta/Fletcherr “AI pricing” approach as “a revolution in ticket pricing” but never defines (much less explains) any algorithmic breakthrough versus traditional airline revenue management... From the paper: "This isn’t revolutionary, in and of itself. As we will see, these types of models have been used in airline revenue management departments for decades. What is new is the integration of these models directly into the pricing system." 3. ... He completely ignores the problem that “AI” tools don’t make decisions and can’t support decisions that aren’t following patterns thoroughly documented in the training data. From the paper: "While this is a grossly oversimplified explanation of what Fetcherr’s model is capable of achieving, it’s essential to note that it does not surpass the theoretical capabilities of the old-school paper-and-pencil regression model." 4. After initially pounding the pricing revolution drum, Miller begins to backtrack. Just as a point of clarity on this one: You can't backtrack within the same document. That's not backtracking, that's called adding context and nuance. Anyone who knows me knows how religiously I take context and nuance in analysis. I accept the compliment, despite it being delivered through a logical fallacy. The remainder of the critique point is simply asking for confidential information from airlines that I certainly won't reveal. The only thing I will say on the matter is that many airlines have expressed frustration at not having data available and normalized, including all airlines mentioned in this report. That's not made up, as the poster suggests. It's a very real challenge expressed by the most sophisticated airlines, and it's current. 5. ...Miller makes insincere efforts to discredit critics of extreme personalized pricing, claiming without citations that they said Delta was planning to collect information about personal checking accounts. Certainly not insincere: https://www.wrdw.com/2025/07/23/delta-expand-its-ai-ticket-pricing-alarm-some/. I think this point is critically and dangerously overlooked by the internet poster. Although the comments from the public are without any basis, they were still made. This is critical in appreciating the context from which the intended reader is likely approaching this paper, and how different it is from that of the poster. These were precisely the questions I was receiving from CEOs of aviation and investment companies, leading to the writing of the paper. It is a silly notion, AND people believe it. I'm not certain of the origins of the poster, but it may be a simple matter of not appreciating the proliferation of misinformation in the United States. You can't just wish it away, and addressing it certainly isn't being "insincere." I'm also not certain where the poster read that there was any suggestion that Delta was ever planning to use any personal data. In fact, it was made explicitly clear that Delta was never planning to do any of this. This was a key tenet of the paper. From the paper: "This is not how AI models work and is certainly not how Delta and Fetcherr’s model works" I don't know, like I said, it was a very strange note to read on a Friday morning (x-apple-data-detectors://3). I'm not normally in the business of responding to disengenuous forum posts, but I'm also not in the business of allowing such disengenuous credibility attacks to go unanswered. I'd classify this as "missing the forest for the trees." I explicitly set the context for who this paper was written for and that it shouldn't be used by revenue management professionals to say "well, actually" to further their own careers. And yet... But, I will express my disappointment in the disingenuous nature in which the paper was clearly read. It appears to have been taken personally, somehow. Perhaps someone forwarded the paper to the poster with a nasty note? Was it sent to them, suggesting it was an academic paper written for researchers and not non-technical aviation leaders? There must be some angry conversation or context beyond this paper from which the poster is responding. It happens. I get it. There are a few things I would have done to improve this paper, but funnily enough, the poster didn't mention any of them. I reserve those for the many discussions I have with the intended audience. Chances are, this poster knows more than I do about revenue management and pricing. Cool. I'm not positioning myself as an expert. I don't compete with RM professionals. There is no advertisement, because I don't advise airlines on pricing strategies for money. It's literally a free paper put out into the public discourse. The paper is written from the perspective of an inquisitive teacher and provider of context to those asking questions and the aviation public as a whole. Those who do know me know integrity and the desire for "oh wow, I never thought of that" drives my work - never "well, actually." The forum post, unfortunately, entirely disregarded this context. I do include my email on every analysis or presentation published. I received nothing. I get things wrong, and banter and discussion are welcomed - even solicited. I make it a point to present my findings in a non-combative manner, with the intention of finding answers rather than suggesting mine are the only ones that should be considered. I don't expect all responses to match the same tone, but I do expect the responses to include me, particularly if there are any explicit attacks on my credibility. Integrity is NOT something I take lightly. And yet, through all of the responses to points made from well outside the context of the paper, the critique betrayed a key oversight that I had expected and hoped to catch early: "If you’re in data science or airline revenue management, none of this will be new to you. However, you’ll find important context in this paper for how the rest of the industry thinks. This is a bridge between the neural networks in which you’ve been living the past few decades and the neurons of the millions of human beings who work or travel in the commercial airline industry. AI means the same thing to most people: ChatGPT. We believe understanding this misunderstanding is incredibly valuable for present and future data scientists. That said, this paper isn’t written specifically for you. As you feel the urge to “well, actually,” consider that explaining the concepts of recurring neural networks and transformers to a population that interacts with all AI solely through a browser chat window requires some reduction. Then consider that these same people are the ones with all the money." It is very easy to descend into our own world of working with detailed models and become disconnected from the rest of the world. Regardless of the models, it still all comes down to human behavior - weird, irrational human behavior. For the intended audience, the paper delivered a conclusion. However, for the long-time professionals who think they've seen it all and have nothing to learn from how the rest of the world approaches learning a complex topic, a separate conclusion was explicitly offered: Understanding this irrationality IS the point. That point was missed. Next time, just email me. Email address is always in the document. Courtney Visual Approach 214-601-3628 (tel:214-601-3628) -- Visual Approach 214-601-3628 (tel:214-601-3628)
-------------------------------------------------------------------------- Revised: 20250507 You are receiving The Mifnet because you requested to join this list. The Mifnet is largely a labor of love, however the infrastructure isn't exactly cost-free. If you'd care to make a small contribution to the effort, please know that it would be greatly appreciated: https://wardell.us/url/mifbit All posts sent to the list should abide by these policies: 1) List members acknowledge that participation in Mifnet is a privilege--not a right. 2) Posts are always off the record, absent specific permission from the author. 3) The tone of discussions is collegial. 4) Posts are expected to be in reasonably good taste. 5) We discuss ideas and not personalities, and we don't speak ill of other Mifnet members. * The Mifnet WEB SITE is: https://www.mifnet.com/ * To UNSUBSCRIBE from this list at any time please visit: https://lists.mifnet.com/ OR: SEND THIS MESSAGE via email: mifnet-requ...@lists.mifnet.com?subject=leave * Send Mifnet mailing list POSTS/SUBMISSIONS to: mifnet@lists.mifnet.com * You may reach the person managing The Mifnet at: mifnet-ow...@lists.mifnet.com * Please consider the DIGEST version of The Mifnet, which consolidates all list traffic into 1-3 messages daily. See instructions at: https://lists.mifnet.com/ * Manage your personal Mifnet SUBSCRIPTION at: https://lists.mifnet.com/ * For a list of all available Mifnet commands, SEND THIS MESSAGE via email: mifnet-requ...@lists.mifnet.com?subject=help * View The Mifnet LIST POLICIES and PRIVACY POLICY at: https://mifnet.com/index.php/policies * View instructions for Mifnet DELIVERY PROBLEMS at: https://mifnet.com/index.php/delivery-problems * View The Mifnet LIST ARCHIVE at: https://lists.mifnet.com/hyperkitty/list/mifnet@lists.mifnet.com/
