Boeing.
For me, the latest saga with Boeing smacks of déjà vu.
Back in October of 2019, I wrote up an analysis on Boeing’s then-current failures with respect to its 737 MAX line of aircraft: the faulty Maneuvering Characteristics Augmentation System (“MCAS”).
And you might be saying to yourself “hey, I had no idea that you’re an expert in aviation technology, Rex!” Surprisingly, I am not, but that’s rather the point. Because sure, while MCAS certainly had its own technical flaws, it wasn’t just those flaws that led to the airplane crashes and subsequent losses of life. Rather, it was a system of shoddy operational controls and failures of risk management at Boeing which enabled MCAS to make its way into live flight scenarios and ultimately cause passenger planes to drop out of the sky.
At its peak, my little report managed to make its way to #3 on Google search results for “Boeing 737 MAX case study” or something of the sort. So of course, someone at Boeing must have read it, implemented the fixes, and now Boeing has a stellar safety record with no further issues. Right?
Sigh.
In case you’ve been willfully blocking out the news for the past few weeks, Boeing’s in a big ‘ol vat of hot water again because a part of the fuselage on one of its passenger planes recently broke off mid-flight and it’s an honest miracle that no one died. Much less technical than the MCAS issue, it seems that here, the “blowout” was due to certain bolts not being properly tightened during assembly of the fuselage. The plane at issue, a 737 MAX 9 operated by Alaska Airlines, had been in service for roughly 8 weeks at time of incident, which points more to a manufacture or design issue than a maintenance issue. After the incident, all 737 MAX 9s in the US (across all airlines) were grounded by the FAA pending inspection. And perhaps thankfully so, as in some cases, those inspections uncovered the same or similar flaws in other 737 MAX 9s.
There are two sets of issues here:
Boeing has a product safety / quality assurance / risk management problem.
The flying public has a Boeing problem.
The first issue is more dry and less interesting for a general audience. I write risk management analyses every now and then, but unless you spend a good amount of time thinking about risk management (as I do), you may not find those pieces to be all that interesting. On the other hand, if you spend a good amount of time thinking about how the world today is slowly circling the drain (as I do), you may be very interested in trying to figure out how, in 2024, we live in a world where commercial aircraft continue to suffer catastrophic failures while flying.
In looking into this, it’s important to acknowledge, at the outset, that manufacturing and selling commercial jetliners is a difficult business. There are pretty steep barriers to entry—both in terms of capital and know-how—and so it’s not the sort of industry in which we’d expect market share to be spread across a large number of players. But the true numbers are telling: the market for commercial jetliners in the United States is fundamentally a duopoly. Moreover, you may not even need a chart to tell you that. I’d venture a guess that more of you out there now take notice to the type of aircraft when booking travel arrangements. In the US, when’s the last time you flew on an aircraft other than a Boeing or an Airbus? Sure, there’s a few Bombardier CRJs, Embraer E-Jets and others in the fleets of various US regional airlines, but by and large, it’s been either Boeing or Airbus, no? But that, in and of itself, isn’t the problem. The problem is the set of incentives that the duopolistic market structure creates for market participants.
Here, we’re going to focus specifically on Boeing, because of all of the major commercial jetliner (i.e., large passenger plane, and not including business jet) producers in the world, Boeing is the only American company. (Airbus is French, Embraer is Brazilian, etc.) And as America’s player in this globally vital industry, Boeing is essentially “too big to fail.”
Now, for those of us around during the 2008 financial crisis, the words “too big to fail” strike an eerie chord. In the financial context, the phrase was used to describe large financial institutions of systemic importance to the US & global financial systems. Many of these institutions were (in my opinion, rightfully) accused of taking irresponsible risks while enjoying the comfort of knowing that, in any calamitous events, they would be “bailed out” by the US government (and, perhaps more importantly, the US taxpayer), lest the financial system crumble. I think that one could reasonably level similar accusations against Boeing today.
See, for all of its market share in the commercial aircraft context, Boeing enjoys even greater dependence on the part of the US government—as well those of many of its allies—in the defense context. I don’t particularly want to get into a discussion about the morality of all this, but, like it or not, Boeing makes a lot of the weaponry that US and US-allied armed services use to keep the proverbial wolves at the proverbial door.
Bringing things back to the commercial aviation context, it makes me uncomfortable to think about just how bad of an incident, or multiple incidents, Boeing would have to be responsible for in order for the issue to truly reach brass tacks, in light of the totality of circumstances regarding, among other things, the aforementioned military dependence. It’s uncomfortable, but perhaps necessary, for us to acknowledge, that we, as a society / country, are quite intractably dependent on a for-profit corporation, and one with questionable safety practices in their manufacturing of passenger-carrying airplanes, no less.
So what’s the check on Boeing? Is it regulation / prosecution? Maybe. But just how hard will governmental actors go on a company so crucial to the national interest? In the wake of the crashes related to MCAS, Boeing agreed to pay $2.5 billion in fines, penalties and other compensation. For reference, though, in the same year as that announcement (2021), Boeing notched gross revenues of just over $62 billion. Was the amount an effective wake-up call for Boeing’s top brass? Maybe. Maybe not. But even if it was, did that change of tune make it all the way down to the factory floor? At least from what we’re seeing now, it seems to not have.
But what about stock price? Boeing is a public company, after all, and its board of directors bears a duty to act in the best interests of the company’s shareholders—a duty which is then passed through to the CEO, the company’s other management, and on and on down to, conceivably, its lowliest floor supervisors. Shouldn’t all of these people—most, if not all, of them being Boeing shareholders themselves—be laser-focused on making sure their planes are safe so their stock price won’t tank on news of, say, a door flying off during flight? Ideally, yes, but in reality, maybe not. Even if we accept the incentive as real, it’s nevertheless incredibly difficult to translate that sort of incentive all the way down the reporting lines within an organization as large as Boeing. And while for a C-level executive, a shift in stock price may have very tangible meaning in terms of personal wealth, for someone turning a wrench or performing a safety inspection on a factory floor, it more likely does not.
But, to back to what I’ve just written, we shouldn’t necessarily accept the incentive as real, because equity markets, and even commercial aviation markets, don’t always behave the way in which we’d imagine they would using, say, common sense.
For example, one would think that Boeing’s stock price would have reacted negatively to news of one of its planes losing a door while traveling through the sky. Call my logic old and folksy, but that just makes sense. And of course, that is what happened…at first. Boeing’s latest safety fiasco did initially have the predictable negative impact on its stock price, but things like stock price incentives (fueled, in part, by commercial behavior) just don’t work the same way in a duopoly as they do in more diffuse markets. And, as of the date of this writing (18 January 2024, or less than 2 weeks after the blow-out incident), Boeing’s stock price has jumped on news of an order of an additional 150 of its MAX jets from India-based Akasa Air. “Lost a door? Just sell some more!” as the ol’ jingle goes. And it’s pretty easy to do when you’re one of the only games in town.
So, how do we fix this? How do we make Boeing actually do the work to put safety first? I’m not sure there’s an answer here besides “hope.”
AI Two-Step.
From November to December of last year, there was a bit of a leadership jumble at OpenAI, the Microsoft-backed company behind ChatGPT. The company’s board ousted Sam Altman, a co-founder of the company and its CEO, for reasons that weren’t all that clear. The move backfired, however, as the company’s employees threatened to resign en masse and its major shareholders were fundamentally in revolt. Ultimately, Sam Altman’s position at OpenAI was restored. Cool story if you’re into corporate governance, but otherwise whatever, right? Sort of.
See, what boards are—at least in theory—supposed to do is prevent companies from doing things that will spell the doom of all humanity (provided that spelling the doom of all humanity is at odds with maximizing shareholder value, of course). And yes, there’s a bit of a quibble here, because what about companies that pollute a lot? But at least in terms of climate change, most major companies have boards, or board committees, designed to address these concerns and balance them with the pursuit of profit. Whether it’s actually been effective, well, that’s a topic for a different Business Thoughts story, but it’s at least an ongoing conversation. But climate change isn’t the only thing involving corporations and the potential doom of all humanity; have you seen The Matrix? How about any of the Terminator movies? That’s right, it’s time to have a serious discussion about rogue AI.
Long before AI even existed in the form in which we know it today, folks were worried about what would happen if the power of AI would eventually outstrip human intelligence (thereby becoming what’s often referred to as “artificial superintelligence” (“ASI”) or “artificial general intelligence” (“AGI”), though nomenclature still varies), and go rogue, acting on its own and potentially against the interests of humanity. Elon Musk has been shouting from the rooftops about this for a while now, and for others, well, have a look at the Wikipedia pages for “the singularity” and “Roko’s Basilisk.”
But what does this have to do with OpenAI and Sam Altman? Has the basilisk been born? Not quite. But it’s been reported that the company, under Altman’s leadership, had made some breakthroughs in its development of AI, and its board: a) felt that it hadn’t been properly informed; and/or b) was concerned about the nature of the breakthroughs and the security implications thereof. Wonderful. And perhaps even more alarming is this reporting from The Washington Post that for Sam Altman, this isn’t the first time he’s been dismissed for going rogue. So should we be worried that he’s now back in charge at OpenAI? Is this a step on the chain of events that leads to a doomsday AI scenario? Or does this all feel more like wild extrapolation and fear-mongering?
A few observations:
AI may prove to be the single-most important technological development since the internet, maybe the semiconductor. It’s a pandora’s box that cannot be re-sealed.
The current state of AI technology is a far, far cry away from a runaway doomsday scenario. We don’t have to worry about Skynet just yet.
While the scenario of a rogue artificial superintelligence being the undoing of the human race might be possible, there’s an ocean between that and it being plausible, much less probable. In the current discourse, no credible arguments or explanations yet exist.
Nevertheless, as with any vital technology, research and development should be undertaken responsibly, and with oversight. Even if not world-ending, an AI system that causes nuclear accidents or power grid failures would be less than desirable, to say the least.
In any oversight function, it is likewise vital to maintain focus on the goal—that is to say, furthering technological development. Whether in a commercial context, with respect to business competition, or in a nation-state context, with respect to national security, failing to keep pace in development spells risk. Exceedingly restrictive oversight, aiming to manage risk, may inadvertently create it.
The recent OpenAI kerfuffle might have been a case of a corporate board jumping the gun, but it shouldn’t be interpreted as a cautionary tale against robust oversight and measured technological development. And while regulation—and perhaps international law—may come into play at some point, until that time, internal oversight functions will be crucial in ensuring that AI developments don’t yield disastrous consequences. The lesson here isn’t that AI developers need less governance; it’s that they need better governance.
Or maybe just stronger corporate governance protections to more effectively shield themselves from oversight. Or that.
Quick Hits
DOJ Halts JetBlue-Spirit Merger — I wasn’t shocked to hear about this, to be honest. Antitrust analysis is fundamentally rooted in doing what’s best for consumers. In this case, cheap airfare! Basically, if you want to pay a bit less to chance it on Spirit, you should be able to! For the life of me, I’m not sure why anyone would rather fly Spirit than JetBlue, or fly Spirit period, but maybe that’s a privileged thing to say. I do think that if Spirit flights were operated like JetBlue flights, even if prices were higher than what Spirit fares would have been, that’s still a net win for the consumer at large. More good flights to more places. Sometimes, it’s not just about who can do it for less. And didn’t we just finish talking about the importance of quality in the context of air travel? Hmm. We’ll see what happens on appeal. The merger agreement, meanwhile, is set to expire this summer.
Another Indian Assassination Plot — I can’t say I’m surprised by this one, given what I’d written up back in September of last year re: the assassination of Hardeep Singh Nijjar, another Sikh activist, in Canada. This time, the plot took place in the US and was thankfully foiled by law enforcement. The indictment in this case provides evidence of the Indian government’s involvement in the Canada plot as well. It also alludes to the possibility of other assassination plots, essentially a campaign to silence Sikh dissidents. Disturbing stuff, to say the least, and likely a topic explored by Chris Wray, the Director of the FBI, on his trip to India last month.
SPACs Didn’t Work Out So Great — This one’s unsurprising. I’m not saying the traditional IPO process is super wonderful or anything, but in what’s become something of a wrestling match between speed and transparency these days (in so many contexts, really), it’s nice to see a win in the transparency column (even if it took a truckload of investor losses to get there).
RTO Isn’t Going As Planned — Another real shocker here. WFH was sort of a gift of the pandemic. It was a gift for employees, but also companies, if they were smart about it. Rent is expensive. Commuting wastes time. And the list goes on. If anything, this story tells me we might be moving towards more flexible work arrangements again, and this is really only bad news for owners of commercial real estate.
The Apple Watch Ban — I’m confident that Apple has really good lawyers. I’m also relatively confident that those lawyers get involved in the product development process. So, I’m not entirely sure how we got to this point, but I’d bet that Apple saw this fight coming from a long way off and nevertheless decided to proceed. As of this week, Apple’s plan is to import the affected watches with the blood oxygen feature turned off while the patent fight continues to play out in court. It seems like the feature being turned off is merely a matter of software, though, and if Apple were to prevail, the affected features could just be turned on via a system update. As a sidenote, I’m wondering whether someone’ll try to jailbreak the features, and my money’s on probably. If you’re a law nerd, here’s a good blog post about the whole thing.
