I.

A few years ago, I had an annual one-on-one with the Chief Technology Officer of an employer with more than ten thousand staff.

Senior management absolutely fawned over this person — extremely politically savvy, they would say. Amazing at acquiring funding. Really cares about everyone on the team. The platform that paid my salary would not exist without their incisive mind. They're invited by multi-billion dollar companies to speak on stages about the immense success they've had deploying their products across every possible sector of the economy, and magazines would breathlessly extol their astounding virtues by placing them on lists with titles like 'Australia's Top 200 Tech Innovators!'

I had no idea what to make of this. People who I respect had only positive things to say, but I had never heard this person say anything beyond the most tepid of platitudes. Break down silos. Be more Agile. Deliver value. They had no technical background, were frankly a weaker speaker than I am, and a weaker writer, but because of the ringing endorsements I tried to understand why this person has gotten to where they are. While I was broadly aware that many people are promoted to positions they were unfit for, savvy friends saw something in this one. And I dearly wanted to understand because I was beginning to realize that either a huge amount of work is fraudulent nonsense in the style of what the late David Graeber called a "bullshit job" or I was so inexperienced that I was missing something important.

In short, what separated this titan of industry from the mere mortals? What impressed my colleagues so? What made the rest of the organization swoon over this avatar of capital, likely salaried anywhere between A$300K to A$1M, when they themselves were striking over wage negotiations?

This meeting was my chance to figure it out. One hour a year. Sixty minutes to see what their unique intelligence manifested as. I'd present some problems that weren't tractable for me as an individual contributor, which so many others had reported doing, and be blinded by their overwhelming brilliance in people management. They would stare into the soul of my problem and bestow upon me the magic words that would convince my peers to have logs that worked.

We sit down and exchange pleasantries. It starts off fine. They ask the correct questions. I raise some serious technical issues, framed in terms that are comprehensible to non-engineers, and include an outright assessment of the issues using terms like "serious, existential risk" to the work we do. I go so far as to mention that we just had an outage in some data processing that lasted a month without anyone noticing due to the poor quality of our codebase, and they adopt a very concerned expression then ask me to write a document about it for them to read, which I suppose is fair enough.

It is at this point that things begin to derail.

"Would you say that data observability is an issue?", they inquire with a tone that very clearly implies that this is a leading question.

I am immediately deeply worried. For those who are unaware¹, my specialty is building systems that move large amounts of data through companies, organize them in a way that is at least marginally less of a horrific clusterfuck than what random people without specific training will do when left to their own devices, and sometimes assist with statistics. Data observability is the high-level term that captures the ability of a business to go "Instead of downloading the data, it would appear the computers caught fire this morning. Would you like to fix this or pretend it never happened?"

The reason that I'm concerned is that the executive in front of me should not be using that term. They have no idea what it really means, which is fine because they aren't specialized in my area, but I am wondering why someone who requires crayon-tier technical explanations is inquiring about a niche, unsexy element of a platform they don't understand. This would be like my 96-year-old grandfather asking me about Bitcoin mining—impressive if he had arrived at the question organically, but in practice I'm already dialing the bank to report a massive theft.

"...Yes," I reply hesitant, "but I think we can remedy most of that by implementing some basic testing. With some backing, I can accomplish that in a month or two."

"What do you think about Monte Carlo?"

What the fuck is Monte Carlo? I've never heard of that. I mean, I know there's a place called Monte Carlo, and I know of Monte Carlo algorithms, but there's no way this person is talking about either of those. It must be a product.

"I don't know what that is."

"I saw it on LinkedIn last night and I think it will solve our observability issues. Do you think we should get it?"

Oh. Oh no. I don't understand what exactly happened last night on LinkedIn, but I know it is dark and sad and reeks of unfulfilled wants. The executive sat before me has been marketed to. No, worse, they have been Marketed to, with a capital M.

I have a friend who is kept away from late-night telemarketing by her partner, because no matter how ridiculous the appliance is, she will become convinced that she needs it. The siren song of the Shamwow is too alluring. That is being Marketed to, and I hadn't realized until that exact moment that people will make purchases worth hundreds of thousands of dollars with that level of thought².

"I don't think we need that," I croak, dimly aware that one wrong word could result in six months of terrible, pointless work for everyone on the team, "We can solve our problems via some simple discipline."

(The reality is that the work they had done was so poor that no product would work with it.)

We still have fifteen minutes left, and it becomes more and more apparent that the executive desperately wants to buy Monte Carlo, as desperately as my little cousins demand another go on the machines at the arcade. For the next day or so, they will wake and see Monte Carlo wherever they go, and when they close their eyes they will see the Monte Carlo sales team leering in their slumber, their tentacle sales-fingers reaching through monitors and rewiring cortices. Every word I say is potential ammunition in their case for buying this piece of software, which I know nothing about. I am nothing to the C-suite but a device that emits words that will ultimately result in being able to say "Ludic supports buying Monte Carlo".

We reach the end of the conversation, and my face hurts from the forced smile. They are doing their best to relate to some of my earlier sentiments.

"Why do you think some of the engineers are struggling?", they ask.

"I'm sure you understand exactly how it is," I lie, "it just takes a lot of work to perform at a high level, and sometimes people fall behind."

"Yes, absolutely", they laugh, "I study on LinkedIn for up to two hours a day after work sometimes³."

We are not the same.

Dazed, I leave the room and collapse into my chair. Two of the other engineers gather around to ask how it went, as time with an executive here is rare and of immense gossip value.

"We started talking about the things that I wanted, and I thought it was going well, then we somehow ended up on something else entirely. How about you two?"

"No, they just kept asking us about something called Monte Carlo. Bro, what the fuck is Monte Carlo?"

They had asked every data engineer in the department about it. Oy vey.

II.

A huge amount of the economy is driven by people who are, simply put, highly suggestible. That is to say that it is very, very easy to get them excited and willing to spend money.

Consider, for example, what it would take to get you to approach your company's lawyer and suggest software to them, totally unprompted, because you saw an advertisement last night. Scratch that, make it every lawyer at your company as each and every one of them goes "I... have never heard of that". But you just keep going because the next one might tell you that the Shamwow is an awesome product.

The answer, in all likelihood, is that no possible advertisement could get you to behave in such an embarrassing fashion. You would instead think things like "I am not a lawyer", "What the hell is this program and why do I feel fit to judge it?", and "The shame from this conversation will keep me up at night for the next five years."

At the surface level, it sounds like you have a desirable characteristic in senior leadership, and that is true in the sense that you're unlikely to waste company money. Yet statistically you are not in senior leadership at an organization with ten thousand employees, someone who buys software at random and hires Deloitte is, so what gives? Don't companies want people who aren't going to be conned into purchasing nonsense?

It turns out that there are tremendous reasons to want people like this running many organizations, and social mechanisms that move the most easily-impressed of us into positions of power, those reasons just exist at the expense of the company or society in which that person is embedded.

There is a massive industry that is built around gathering people that fit the "thinks LinkedIn is studying" profile into rooms, who also have access to organizational money, and then charging sales teams for permission to get into that room. I was dimly aware that this stuff happens, but it is now impossible from my professional profile to tell that I am one guy doing his best to write good software with a few friends, as opposed to a millionaire, which resulted in the following message:

Dear Ludic,

I saw that you are featured in an upcoming webinar as below:

A Boardroom Guide to AI: Spotting Hype and Managing Costs

Noting this, I would like to bring the below to your attention, as [REDACTED] staged this as a very successful in-person event in June of this year targeting directors.

[REDACTED] is the leading annual event for board directors from publicly traded companies across the United States, attracting directors now for the last 20 years.

Next year's event will take place at [REDACTED]

We expect around 70-80 directors to be in attendance, representing some 200+ boards.

The event is entirely in-person.

If it works for you, I would suggest a short call to discuss the specifics, including positioning Hermit Tech on the agenda (attached) with a commercial opportunity.

Looking forward to hearing from you.

With best regards,
[REDACTED]

After a quick exchange, it became apparent that the deal is as follows: I can wire them money and in exchange be granted access to the fancy room, where I would be allowed to Market to these people. I would turn up in a suit, exaggerate how successful my business is, possibly make some incredibly grotesque comments about women⁴⁵ depending on the clientele so that the male directors know that I'm one of them, and finally we will Do Business.

Money now in exchange for access to credulous people who use words like synergy with a straight face later. I have no doubt that the actual attendees would vary wildly, ranging from a few savvy people, to outright grifters, to the terminally deranged. Even the pleasant and sufficiently skeptical can feel compelled to attend because the truth is that executive compensation and funding is driven by your relationships to other people, but make no mistake, the goal of salespeople with weak products is to find the weakest minds in the audience and lay siege. They are enormously vulnerable — I know many people who fit this profile, and it is disconcerting to see people put the whammy on them. They zone out when Donald Trump is on a nearby television, eyes glazing over, and in private will say "he makes a lot of sense" without being able to repeat anything he said. They buy into things like the prosperity gospel with hardly any prompting, and can more-or-less have absolutely no ability to avoid scams — they'll happily say that they're not technical people but quantum is the future.

To quote Ed Zitron, who later on in this excellent piece quotes me, forming the mythical content promotion ouroboros:

Whatever organization that's burdened you with some sort of half-baked, half-useful piece of shit business app has done so because the people up top don't care if it's good, just that it works, and "works" can be an extremely fuzzy word. It doesn't matter that Microsoft Teams is universally-loathed and regularly threatens to crash every time you load it. A Microsoft salesperson used its monopoly power to cut your boss a deal to either bundle it with a bunch of other mediocre shit or they saw the name "Microsoft" and said "oh boy! I love Microsoft Word!" and pulled out their credit card so fast it left a gouge in their monitor.

Indeed, we've only seen one side of the coin — I get the messages aimed at sponsors, where they trust that I will don the mantle of the wolf and select the choicest morsels from the flock that they have gathered. A friend who wishes to remain anonymous sent me what the prey receives:

This is absolutely shameless. "Ensure you're guaranteed to meet and engage with an elite group in the cybersecurity space". What are the actual questions?

Do you have lots of money? Are you authorized to spend that money? They're just doing lead qualification. That's all this is. I currently run sales and marketing for about twenty hours a week — I know, I know, what have I become? — but I would not be able to ask questions this crass without hiding my face in shame.

The same person sent me the following PDF from a Melbourne-based conference which includes a sponsorship package that they call the "guy at the bar thinks you're cute and wanted to buy you this drink" package.

You can straight-up buy people tickets to attend events, and have a concierge deliver them into the eager maw of your marketing machine. I was shocked to see an absolutely trivial price tag too. A$2.5K? Even my tiny operation spent more than that last week buying hardware for one engineer — it's a rounding error to get the person that chooses what technology you're going to be using for the next five years and whether you're being laid off into my sales kill box. I just have to wave my company credit card and bellow forth a wretched miasma of lies about no-code tools and generative AI, and voilà, some of you are unemployed now!

III.

Up until now, the picture that I've painted is one of credulous people who are easily excited. But, of course, people are multifaceted, and minds can contain complex, self-deluded, and contradictory motivations.

There is a horrendous incestuousness to the sales cycle at large enterprises, certainly in the AI and data space where I work, and it is tied intricately with the way that jobs are distributed in most of the economy⁶. The main way to get a good job offer is to come recommended by someone inside of a company. Software engineering and other specialties of a technical bent (including artists, writers, etc.) have the additional barrier that our skills are testable to some degree, but this is not the case for most senior management and executive roles.

In those roles, it is essentially impossible to work out whether someone's tenure as a manager or executive was useful. Consider Elon Musk, the patron saint of dysregulated man-children — it is surprisingly difficult to get one straight story about the man and the effectiveness of his methods. I've heard stories about how he boldly decided to build a rocket from scratch when providers were inadequate and about how SpaceX's dominance is attributable to him personally stalking the halls and firing incompetents. I've heard other people say that they hide interns from him because he has a habit of firing them on the spot over irrelevant trivia, and an executive in the space sector personally told me that SpaceX succeeds despite his antics due to the work of their COO, Gwynne Shotwell. I don't have any concrete evidence for which narrative is closer to truth, or if he has ever fired an intern after a pop quiz, and that's for someone that is written about more than almost anyone on the planet.

When Johnny McManager rolls into town and assures me that he was instrumental in tweaking the widgets in a massive banking application, and that this definitely drove massive revenue for the business, what recourse do I have? Do I trust the referees he provided? Do I start cold-calling executives at the bank to try and get an assessment? Do they even know and, if they did know, would they admit that a project failed on their watch?

Slava Akmechet writes:

Company metrics have momentum and lag. Nearly all political behavior exploits these two properties. [...] So opportunists don't worry about any of that. The winning strategy is to ignore company metrics completely and move between projects every eighteen months so that nobody notices.

Or, in other words, high-level statements like "I led a successful project" mean nothing. The project may not have been successful, or was judged to be a success for political reasons, or was successful for reasons that had nothing to do with management. This is a recurring theme across many sources. Sean Goedecke, another writer here in Melbourne, writes:

I know it sounds extreme, but I think many engineers do not understand what shipping even is inside a large tech company. What does it mean to ship? It does not mean deploying code or even making a feature available to users. Shipping is a social construct within a company. Concretely, that means that a project is shipped when the important people at your company believe it is shipped. If you deploy your system, but your manager or VP or CEO is very unhappy with it, you did not ship.

And conversely, if you do not deploy your system to a working state but can someone make your VP or CEO happy, then you did ship. This sounds even stranger. How can you ship if the code doesn't work?

It's called lying, and it'll solve all your problems!

Recall that the platform I was working on previously had logging that was broken for months and was idling to the tune of half a million dollars, but it had nonetheless "shipped". And the project before that claimed the full million dollars from the sponsor, deliverable upon shipping, from the funder despite still not actual being in production several years later. This happens constantly, and may be more common than projects actually working out.

Because management in large, dysfunctional (read: typical) companies is a game about promising to ship things to people further up your chain, people are broadly incentivized to say that everything has shipped no matter what has happened unless it is impossible to lie about this easily.

What ends up developing gradually is a network of people who are selected for their ability to support convenient social narratives, and if you're going to be negative at all, you aren't allowed in the club. When someone is asked to be a team player, what is really being said is "shut the fuck up and we'll let you into the club". That is precisely why people use phrases like team player — it isn't hard to pick something less clumsy and upsetting, but then you might not realize it's a threat which is the whole point!

IV. Mind Tricks Don't Work On Me, Only Status

Status absolutely fascinates me. I believe that is drives much more behaviour than even the acquisition of money — many non-introspective people only want money because they think it will bring them respect in the eyes of others. This theme emerges everywhere I read deeply. Keith Johnstone's seminal work on improvised theatre, Impro, opens with a chapter on status dynamics. The psychological difficulty of the status swings afforded by randomness-driven fields such as academia is a core theme in Taleb's The Black Swan.

Did you know Elon Musk just got caught paying people to hit the leaderboard in games so he could lie about being so smart that he's both the CEO of all those companies and somehow crushes people at games that they spend hundreds of hours on?

While I like Snowflake as a piece of software, it is probably not a high priority to move to it at most large companies for various reasons I won't get into here. Fine, I'll get into one of them. It's just a really good data warehouse, you absolute maniacs, it isn't the cure for cancer, why the fuck is it valued at $53B?

Because everyone is buying it, and this has to be driven by non-technical leadership because there aren't enough technical leaders to drive that sort of valuation. Why would non-technicians be so focused on a database of all things, a concept so dull that it is Effective Communication 101 to try and avoid using the term in front of a lay audience? It's because if you buy Snowflake then you're allowed to get onto stages at large venues and talk about how revolutionary Snowflake was for your business, which on the surface looks like a brag about Snowflake, but is actually a brag about the great decisions you've been making and the wealth you can deploy if someone becomes your friend. And the audience is full of people that are now thinking "If I buy Snowflake, I can be on that stage, and everyone will finally recognize my brilliance".

It is a bribe, straight up, and done in such a way that everyone understands that further bribes are available for anyone willing to be enthusiastic about something they don't understand. Matt Stoller has written at some length about how government purchasing is heavily driven by award acquisition, and it all rounds out to "this is discount Illuminati bullshit".

The net result is that a huge number of our leaders are essentially stealing money, but they can't withdraw the money directly, so they have to spend the organization's capital on expensive nonsense to purchase status then convert that status into a better salary somewhere else at a really, really bad exchange rate. It really is embezzling without the charm of efficiency. We'd be better off letting them withdraw $1M instead of forcing them to spend $30M so that your competitor offers them a $1M raise.

And it turns out there's a market for status too! I started getting these messages after changing my title to director.

Hi Ludic,

Would you like to be interviewed by one of our journalists and talk about your work and product? :)

Xraised (www.xraised.com) is a vibrant community of industry leaders and innovators like you. We offer a comprehensive package for £99, which includes an interview with our journalist to be posted on our website, Spotify and Amazon Music, with the option to add a dedicated press release about you and your company. The content is authentic, and the potential audience is 84M readers from North America, Australia, UK and the EU.

Would you like to schedule an interview via this link? [REDACTED]

Contact us at interviews@xraised.com if you have any questions :)

[REDACTED]
Project Manager at Xraised

You can see all the confluence of all the factors above. They're targeting a demographic that exists — unwilling or unable to attract an audience by strength of quality. Desperate enough for attention to pay £99 instead of just doing some email outreach. Dunce enough to think inserting the word "authentic" makes it so, and gullible enough to think that £99 could actually reach even 1% of 84 million people. The most popular thing I've ever written has done around 2M hits, and that was enough that I can fly to almost any country in the world and have someone buy me a beer. Xraised is trying to find people that think you can purchase anything like that forty times over for £99, and those people all have employees.⁷

V.

If that sounds dystopian, it is! But it helps to remember that many of us are, broadly speaking, living in an era of unprecedented wealth, and that is only possible because some things work. Cars do work. You're on the internet right now. Things working means that there are non-fraudulent sectors of the economy, it just takes some serious looking to find the people in those sectors, and unfortunately a willingness to bury the dream of becoming a full-time employee for thirty years and never having to make course adjustments to your career trajectory. Even a great company will have people move on, or grow until it is a big company, and if it doesn't then they aren't hiring anymore so you aren't allowed in.

I'll write about it at some length in the future, but my own consultancy currently only works with startups putting together an analytics stack to conform with the requirements of an enterprise sale because, while everything in business comes with some weird incentives, startups selling to enterprise is a case where the client has immense skin in the game and failure is not acceptable. No one pays us consulting rates to fail to ship in startup land (people can fail to ship on their own, thank you very much). Real work is possible! I've seen it!

And I've so far managed to avoid becoming ensnared in the status trap by strength of will, because I am a superior being and very important executive director with the greatest team on the planet, or more humbly the strongest team in the southern hemisphere. We're going to become the strongest team in the northern hemisphere tomorrow by purchasing Monte Carlo, the best platform in the world to prepare you for becoming "AI-Ready"!

I'm very excited to talk to Olivia Miller, who Monte Carlo desperately wants me to know is a human.

What the fuck is going on?

Let’s make one thing clear first: I’m not singling out Google’s handling of problematic and malicious browser extensions because it is worse than Microsoft’s for example. No, Microsoft is probably even worse but I never bothered finding out. That’s because Microsoft Edge doesn’t matter, its market share is too small. Google Chrome on the other hand is used by around 90% of the users world-wide, and one would expect Google to take their responsibility to protect its users very seriously, right? After all, browser extensions are one selling point of Google Chrome, so certainly Google would make sure they are safe?

Unfortunately, my experience reporting numerous malicious or otherwise problematic browser extensions speaks otherwise. Google appears to take the “least effort required” approach towards moderating Chrome Web Store. Their attempts to automate all things moderation do little to deter malicious actors, all while creating considerable issues for authors of legitimate add-ons. Even when reports reach Google’s human moderation team, the actions taken are inconsistent, and Google generally shies away from taking decisive actions against established businesses.

As a result, for a decade my recommendation for Chrome users has been to stay away from Chrome Web Store if possible. Whenever extensions are absolutely necessary, it should be known who is developing them, why, and how the development is being funded. Just installing some extension from Chrome Web Store, including those recommended by Google or “featured,” is very likely to result in your browsing data being sold or worse.

Google employees will certainly disagree with me. Sadly, much of it is organizational blindness. I am certain that you meant it well and that you did many innovative things to make it work. But looking at it from the outside, it’s the result that matters. And for the end users the result is a huge (and rather dangerous) mess.

Some recent examples

Five years ago I discovered that Avast browser extensions were spying on their users. Mozilla and Opera disabled the extension listings immediately after I reported it to them. Google on the other hand took two weeks where they supposedly discussed their policies internally. The result of that discussion was eventually their “no surprises” policy:

Building and maintaining user trust in the Chrome Web Store is paramount, which means we set a high bar for developer transparency. All functionalities of extensions should be clearly disclosed to the user, with no surprises. This means we will remove extensions which appear to deceive or mislead users, enable dishonest behavior, or utilize clickbaity functionality to artificially grow their distribution.

So when dishonest behavior from extensions is reported today, Google should act immediately and decisively, right? Let’s take a look at two examples that came up in the past few months.

In October I wrote about the refoorest extension deceiving its users. I could conclusively prove that Colibri Hero, the company behind refoorest, deceives their users on the number of trees they supposedly plant, incentivizing users into installing with empty promises. In fact, there is strong indication that the company never even donated for planting trees beyond a rather modest one-time donation.

Google got my report and dealt with it. What kind of action did they take? That’s a very good question that Google won’t answer. But refoorest is still available from Chrome Web Store, it is still “featured” and it still advertises the very same completely made up numbers of trees they supposedly planted. Google even advertises for the extension, listing it in the “Editors’ Picks extensions” collection, probably the reason why it gained some users since my report. So much about being honest. For comparison: refoorest used to be available from Firefox Add-ons as well but was already removed when I started my investigation. Opera removed the extension from their add-on store within hours of my report.

But maybe that issue wasn’t serious enough? After all, there is no harm done to users if the company is simply pocketing the money they claim to spend on a good cause. So also in October I wrote about the Karma extension spying on users. Users are not being notified about their browsing data being collected and sold, except for a note buried in their privacy policy. Certainly, that’s identical to the Avast case mentioned before and the extension needs to be taken down to protect users?

Again, Google got my report and dealt with it. And again I fail to see any result of their action. The Karma extension remains available on Chrome Web Store unchanged, it will still notify their server about every web page you visit (see screenshot above). The users still aren’t informed about this. Yet their Chrome Web Store page continues to claim “This developer declares that your data is not being sold to third parties, outside of the approved use cases,” a statement contradicted by their privacy policy. The extension appears to have lost its “Featured” badge at some point but now it is back.

Note: Of course Karma isn’t the only data broker that Google tolerates in Chrome Web Store. I published a guest article today by a researcher who didn’t want to disclose their identity, explaining their experience with BIScience Ltd., a company misleading millions of extension users to collect and sell their browsing data. This post also explains how Google’s “approved use cases” effectively allow pretty much any abuse of users’ data.

Mind you, neither refoorest nor Karma were alone but rather recruited or bought other browser extensions as well. These other browser extensions were turned outright malicious, with stealth functionality to perform affiliate fraud and/or collect users’ browsing history. Google’s reaction was very inconsistent here. While most extensions affiliated with Karma were removed from Chrome Web Store, the extension with the highest user numbers (and performing affiliate fraud without telling their users) was allowed to remain for some reason.

With refoorest, most affiliate extensions were removed or stopped using their Impact Hero SDK. Yet when I checked more than two months after my report two extensions from my original list still appeared to include that hidden affiliate fraud functionality and I found seven new ones that Google apparently didn’t notice.

The reporting process

Now you may be wondering: if I reported these issues, why do I have to guess what Google did in response to my reports? Actually, keeping me in the dark is Google’s official policy:

This is by the way the response I received in November after pointing out the inconsistent treatment of the extensions. A month later the state of affairs was still that some malicious extensions got removed while other extensions with identical functionality were available for users to install, and I have no idea why that is. I’ve heard before that Google employees aren’t allowed to discuss enforcement actions, and your guess is as good as mine as to whom this policy is supposed to protect.

Supposedly, the idea of not commenting on policy enforcement actions is hiding the internal decision making from bad actors, so that they don’t know how to game the process. If that’s the theory however, it isn’t working. In this particular case the bad actors got some feedback, be it through their extensions being removed or due to the adjustments demanded by Google. It’s only me, the reporter of these issues, who needs to be guessing.

But, and this is a positive development, I’ve received a confirmation that both these reports are being worked on. This is more than I usually get from Google which is: silence. And typically also no visible reaction either, at least until a report starts circulating in media publications forcing Google to act on it.

But let’s take a step back and ask ourselves: how does one report Chrome Web Store policy violations? Given how much Google emphasizes their policies, there should be an obvious way?

In fact, there is a support document on reporting issues. And when I started asking around, even Google employees would direct me to it.

If you find something in the Chrome Web Store that violates the Chrome Web Store Terms of Service, or trademark or copyright infringement, let us know.

Sounds good, right? Except that the first option says:

At the bottom left of the window, click Flag Issue.

Ok, that’s clearly the old Chrome Web Store. But we understand of course that they mean the “Flag concern” link which is nowhere near the bottom. And it gives us the following selection:

This doesn’t really seem like the place to report policy violations. Even “Felt suspicious” isn’t right for an issue you can prove. And, unsurprisingly, after choosing this option Google just responds with:

Your abuse report has been submitted successfully.

No way to provide any details. No asking for my contact details in case they have questions. No context whatsoever, merely “felt suspicious.” This is probably fed to some algorithm somewhere which might result in… what actually? Judging by malicious extensions where users have been vocally complaining, often for years: nothing whatsoever. This isn’t the way.

Well, there is another option listed in the document:

If you think an item in the Chrome Web Store violates a copyright or trademark, fill out this form.

Yes, Google seems to care about copyright and trademark violations, but a policy violation isn’t that. If we try the form nevertheless it gives us a promising selection:

Finally! Yes, policy reasons are exactly what we are after, let’s click that. And there comes another choice:

That’s really the only option offered. And I have questions. At the very least those are: in what jurisdiction is child sexual abuse material a non-legal reason to report content? And: since when is that the only policy that Chrome Web Store has?

We can go back and try “Legal Reasons to Report Content” of course but the options available are really legal issues: intellectual properties, court orders or violations of hate speech law. This is another dead end.

It took me a lot of asking around to learn that the real (and well-hidden) way to report Chrome Web Store policy violations is Chrome Web Store One Stop Support. I mean: I get it that Google must be getting lots of non-sense reports. And they probably want to limit that flood somehow. But making legitimate reports almost impossible can’t really be the way.

In 2019 Google launched the Developer Data Protection Reward Program (DDPRP) meant to address privacy violations in Chrome extensions. Its participation conditions were rather narrow for my taste, pretty much no issue would qualify for the program. But at least it was a reliable way to report issues which might even get forwarded internally. Unfortunately, Google discontinued this program in August 2024.

It’s not that I am very convinced of DDPRP’s performance. I’ve used that program twice. First time I reported Keepa’s data exfiltration. DDPRP paid me an award for the report but, from what I could tell, allowed the extension to continue unchanged. The second report was about the malicious PDF Toolbox extension. The report was deemed out of scope for the program but forwarded internally. The extension was then removed quickly, but that might have been due to the media coverage. The benefit of the program was really: it was a documented way of reaching a human being at Google that would look at a problematic extension.

Chrome Web Store and their spam issue

In theory, there should be no spam on Chrome Web Store. The policy is quite clear on that:

We don’t allow any developer, related developer accounts, or their affiliates to submit multiple extensions that provide duplicate experiences or functionality on the Chrome Web Store.

Unfortunately, this policy’s enforcement is lax at best. Back in June 2023 I wrote about a malicious cluster of Chrome extensions. I listed 108 extensions belonging to this cluster, pointing out their spamming in particular:

Well, 13 almost identical video downloaders, 9 almost identical volume boosters, 9 almost identical translation extensions, 5 almost identical screen recorders are definitely not providing value.

I’ve also documented the outright malicious extensions in this cluster, pointing out that other extensions are likely to turn malicious as well once they have sufficient users. And how did Google respond? The malicious extensions have been removed, yes. But other than that, 96 extensions from my original list remained active in January 2025, and there were of course more extensions that my original report didn’t list. For whatever reason, Google chose not to enforce their anti-spam policy against them.

And that’s merely one example. My most recent blog post documented 920 extensions using tricks to spam Chrome Web Store, most of them belonging to a few large extension clusters. As it turned out, Google was made aware of this particular trick a year before my blog post already. And again, for some reason Google chose not to act.

Can extension reviews be trusted?

So when you search for extensions in Chrome Web Store, many results will likely come from one of the spam clusters. But the choice to install a particular extension is typically based on reviews. Can at least these reviews be trusted? Concerning moderation of reviews Google says:

Google doesn’t verify the authenticity of reviews and ratings, but reviews that violate our terms of service will be removed.

And the important part in the terms of service is:

Your reviews should reflect the experience you’ve had with the content or service you’re reviewing. Do not post fake or inaccurate reviews, the same review multiple times, reviews for the same content from multiple accounts, reviews to mislead other users or manipulate the rating, or reviews on behalf of others. Do not misrepresent your identity or your affiliation to the content you’re reviewing.

Now you may be wondering how well these rules are being enforced. The obviously fake review on the Karma extension is still there, three months after being posted. Not that it matters, with their continuous stream of incoming five star reviews.

A month ago I reported an extension to Google that, despite having merely 10,000 users, received 19 five star reviews on a single day in September – and only a single (negative) review since then. I pointed out that it is a consistent pattern across all extensions of this account, e.g. another extension (merely 30 users) received 9 five star reviews on the same day. It really doesn’t get any more obvious than that. Yet all these reviews are still online.

And it isn’t only fake reviews. The refoorest extension incentivizes reviews which violates Google’s anti-spam policy (emphasis mine):

Developers must not attempt to manipulate the placement of any extensions in the Chrome Web Store. This includes, but is not limited to, inflating product ratings, reviews, or install counts by illegitimate means, such as fraudulent or incentivized downloads, reviews and ratings.

It has been three months, and they are still allowed to continue. The extension gets a massive amount of overwhelmingly positive reviews, users get their fake trees, everybody is happy. Well, other than the people trying to make sense of these meaningless reviews.

With reviews being so easy to game, it looks like lots of extensions are doing it. Sometimes it shows as a clearly inflated review count, sometimes it’s the overwhelmingly positive or meaningless content. At this point, any user ratings with the average above 4 stars likely have been messed with.

The “featured” extensions

But at least the “Featured” badge is meaningful, right? It certainly sounds like somebody at Google reviewed the extension and considered it worthy of carrying the badge. At least Google’s announcement indeed suggests a manual review:

Chrome team members manually evaluate each extension before it receives the badge, paying special attention to the following:

1. Adherence to Chrome Web Store’s best practices guidelines, including providing an enjoyable and intuitive experience, using the latest platform APIs and respecting the privacy of end-users.
2. A store listing page that is clear and helpful for users, with quality images and a detailed description.

Yet looking through 920 spammy extensions I reported recently, most of them carry the “Featured” badge. Yes, even the endless copies of video downloaders, volume boosters, AI assistants, translators and such. If there is an actual manual review of these extensions as Google claims, it cannot really be thorough.

To provide a more tangible example, Chrome Web Store currently has Blaze VPN, Safum VPN and Snap VPN extensions carry the “Featured” badge. These extensions (along with Ishaan VPN which has barely any users) belong to the PDF Toolbox cluster which produced malicious extensions in the past. A cursory code inspection reveals that all four are identical and in fact clones of Nucleus VPN which was removed from Chrome Web Store in 2021. And they also don’t even work, no connections succeed. The extension not working is something users of Nucleus VPN complained about already, a fact that the extension compensated with fake reviews.

So it looks like the main criteria for awarding the “Featured” badge are the things which can be easily verified automatically: user count, Manifest V3, claims to respect privacy (not even the privacy policy, merely that the right checkbox was checked), a Chrome Web Store listing with all the necessary promotional images. Given how many such extensions are plainly broken, the requirements on the user interface and generally extension quality don’t seem to be too high. And providing unique functionality definitely isn’t on the list of criteria.

In other words: if you are a Chrome user, the “Featured” badge is completely meaningless. It is no guarantee that the extension isn’t malicious, not even an indication. In fact, authors of malicious extensions will invest some extra effort to get this badge. That’s because the website algorithm seems to weigh the badge considerably towards the extension’s ranking.

How did Google get into this mess?

Google Chrome first introduced browser extensions in 2011. At that point the dominant browser extensions ecosystem was Mozilla’s, having been around for 12 years already. Mozilla’s extensions suffered from a number of issues that Chrome developers noticed of course: essentially unrestricted privileges necessitated very thorough reviews before extensions could be published on Mozilla Add-ons website, due to high damage potential of the extensions (both intentional and unintentional). And since these reviews relied largely on volunteers, they often took a long time, with the publication delays being very frustrating to add-on developers.

Disclaimer: I was a reviewer on Mozilla Add-ons myself between 2015 and 2017.

Google Chrome was meant to address all these issues. It pioneered sandboxed extensions which allowed limiting extension privileges. And Chrome Web Store focused on automated reviews from the very start, relying on heuristics to detect problematic behavior in extensions, so that manual reviews would only be necessary occasionally and after the extension was already published. Eventually, market pressure forced Mozilla to adopt largely the same approaches.

Google’s over-reliance on automated tools caused issues from the very start, and it certainly didn’t get any better with the increased popularity of the browser. Mozilla accumulated a set of rules to make manual reviews possible, e.g. all code should be contained in the extension, so no downloading of extension code from web servers. Also, reviewers had to be provided with an unobfuscated and unminified version of the source code. Google didn’t consider any of this necessary for their automated review systems. So when automated review failed, manual review was often very hard or even impossible.

It’s only with the introduction of Manifest V3 now that Chrome finally prohibits remote hosted code. And it took until 2018 to prohibit code obfuscation, while Google’s reviewers still have to reverse minification for manual reviews. Mind you, we are talking about policies that were already long established at Mozilla when Google entered the market in 2011.

And extension sandboxing, while without doubt useful, didn’t really solve the issue of malicious extensions. I already wrote about one issue back in 2016:

The problem is: useful extensions will usually request this kind of “give me the keys to the kingdom” permission.

Essentially, this renders permission prompts useless. Users cannot possibly tell whether an extension has valid reasons to request extensive privileges. So legitimate extensions have to constantly deal with users who are confused about why the extension needs to “read and change all your data on all websites.” At the same time, users are trained to accept such prompts without thinking twice.

And then malicious add-ons come along, requesting extensive privileges under a pretense. Monetization companies put out guides for extension developers on how they can request more privileges for their extensions while fending off complains from users and Google alike. There is a lot of this going on in Chrome Web Store, and Manifest V3 couldn’t change anything about it.

So what we have now is:

1. Automated review tools that malicious actors willing to invest some effort can work around.
2. Lots of extensions with the potential for doing considerable damage, yet little way of telling which ones have good reasons for that and which ones abuse their privileges.
3. Manual reviews being very expensive due to historical decisions.
4. Massively inflated extension count due to unchecked spam.

Number 3 and 4 in particular seem to further trap Google in the “it needs to be automated” mindset. Yet adding more automated layers isn’t going to solve the issue when there are companies which can put a hundred employees on devising new tricks to avoid triggering detection. Yes, malicious extensions are big business.

What could Google do?

If Google were interested in making Chrome Web Store a safer place, I don’t think there is a way around investing considerable (manual) effort into cleaning up the place. Taking down a single extension won’t really hurt the malicious actors, they have hundreds of other extensions in the pipeline. Tracing the relationships between extensions on the other hand and taking down the entire cluster – that would change things.

As the saying goes, the best time to do this was a decade ago. The second best time is right now, when Chrome Web Store with its somewhat less than 150,000 extensions is certainly large but not yet large enough to make manual investigations impossible. Besides, there is probably little point in investigating abandoned extensions (latest release more than two years ago) which make up almost 60% of Chrome Web Store.

But so far Google’s actions have been entirely reactive, typically limited to extensions which already caused considerable damage. I don’t know whether they actually want to stay on top of this. From the business point of view there is probably little reason for that. After all, Google Chrome no longer has to compete for market share, having essentially won against the competition. Even with Chrome extensions not being usable, Chrome will likely stay the dominant browser.

In fact, Google has significant incentives to keep a particular class of extensions low, so one might even suspect intention behind allowing Chrome Web Store to be flooded with shady and outright malicious ad blockers.

Everything put into the building that is unnecessary, every cubic foot that is used for purely ornamental purposes beyond that needed to express its use and to make it harmonize with others of its class, is a waste — is, to put it in plain English, perverting someone’s money — George Hill, commercial real estate expert, 1904

A skyscraper is a machine that makes the land pay — Cass Gilbert, architect, 1900.

The most common style for skyscrapers in the US (and probably the world) is the glass box — a structural skeleton of steel or concrete, with a skin of non-load bearing curtain wall made of glass and metal (typically aluminum), and without much in the way of decoration or ornament. If you look at the list of tall buildings recently completed in the US, for instance, buildings with glass curtain wall dominate the list; the buildings that don’t have it mostly have exteriors of exposed concrete framing with large glass windows; it provides a similar aesthetic.

But skyscrapers weren’t always built in this style. If you go back to the early 20th century, you see skyscrapers clad with brick and stone, and much more pattern, detailing, and ornamentation. Buildings like the American Radiator Building (1924), the Jeweler’s Building (1927), the Carbide and Carbon Building (1929), the Chrysler Building (1930), and the General Electric Building (1931) have beautiful stonework, elaborate spires and crowns, and lots of other architectural detail.

Why did this transition happen? A common theory is that it was essentially a plot by modernist architects, who had a particular theory of what made a building “good” — “honest” buildings without excessive decoration or ornament — and who successfully foisted that style on an unwilling public, who preferred the old style. This is more or less the thesis of Tom Wolfe’s From Bauhaus to Our House, a 34,000 word polemic decrying the rise of modern architecture:

Every great law firm in New York moves without a sputter of protest into a glass-box office building with concrete slab floors and seven-foot-ten-inch-high concrete slab ceilings and plasterboard walls and pygmy corridors — and then hires a decorator and gives him a budget of hundreds of thousands of dollars to turn these mean cubes and grids into a horizontal fantasy of a Restoration townhouse. I have seen the carpenters and cabinetmakers and search-and-acquire girls hauling in more cornices, covings, pilasters, carved moldings, and recessed domes, more linenfold paneling, more (fireless) fireplaces with festoons of fruit carved in mahogany on the mantels, more chandeliers, sconces, girandoles, chestnut leather sofas, and chiming clocks than Wren, Inigo Jones, the brothers Adam, Lord Burlington, and the Dilettanti, working in concert, could have dreamed of.

Without a peep they move in! — even though the glass box appalls them all…

And why? They can't tell you. They look up at the barefaced buildings they have bought, those great hulking structures they hate so thoroughly, and they can't figure it out themselves. It makes their heads hurt.

Scott Alexander echoes this theory in his review of the book. Samuel Hughes also gestures at this theory in this Works in Progress piece, where he argues that ornament on buildings didn’t die because it got too expensive, but rather because of the rise of modernism:

In the first half of the twentieth century, Western artistic culture was transformed by a complex family of movements that we call modernism, a trend that extends far beyond architecture into the literature of Joyce and Pound, the painting of Picasso and Matisse, and the music of Schoenberg and Stravinsky. Between the 1920s and the 1950s, modernist approaches to architecture were adopted for virtually all public buildings and many private ones. Most architectural modernists mistrusted ornament and largely excluded it from their designs. The immense and sophisticated industries that had served the architectural aspirations of the nineteenth century withered in full flower. The fascinating and mysterious story of how this happened cannot be told here. But it is a story of cultural choice, not of technological destiny. It was within our collective power to choose differently. It still is.

Changing architectural trends were undoubtedly a factor in the rise of the glass box aesthetic, but this explanation misses a huge part of why this transition happened. Notably, this theory completely omits the role of the real estate developer, who has a greater influence than anyone else in how a building comes together. Skyscrapers are designed by architects, but it’s the developer who conceives of the project, arranges the funding, hires the design team, and ultimately decides what the building will be. To me, the question isn’t “Why did architects embrace modernism? It’s “Why did real estate developers embrace modernism?”

Ultimately, it was economics (or at least perceived economics) that drove developers to embrace this style. Glass curtain wall buildings were cheaper to erect than their masonry predecessors, and they allowed developers to squeeze more rentable space from the same building footprint. Ornate, detailed exteriors were increasingly seen as something tenants didn’t particularly care about, making it harder to justify spending money on them. And once this style had taken hold, rational risk aversion encouraged developers and builders to keep using it.

The role of the real estate developer

In the US, most skyscrapers are built as speculative buildings: they’re built to be rented or sold to someone else. Even large corporate headquarters tend to rent out much of their space to other companies: when Sears built the Sears Tower (now Willis Tower), the company only occupied a little over half the building, and rented out the rest. A skyscraper is thus primarily a business venture: the builder invests money in constructing it, in the hopes of earning a return on that investment by renting it out or selling it. Typically, the builder of a skyscraper will be a real estate developer, someone in the business of acquiring parcels of land and constructing buildings on them.

The basis of a speculative building project is the pro forma: a financial model of the building and its profitability, based on its various projected costs (land, construction costs, financing and so on) compared to the revenues earned from rents. When considering a new building project, the developer will play around with this model — how many apartment units to construct and of what type, whether they should be market rate or luxury, whether any adjacent parcels are needed or not, whether the project should include office space or retail, how much parking is required and where on the site it can fit — until he gets something that seems to work.

The pro forma creates a box that the developer has to work within, and ultimately dictates what sort of building can be built. If costs are too high, or revenues are too low, or the cap rate (net income divided by value of the building) isn’t what it needs to be, then the developer won’t be able to raise the funds from investors to get the building built. The form and function of the building is thus a reflection of what the developer thinks is likely to be profitable. As Jerry Adler notes in High Rise, a chronicle of the construction of a New York skyscraper:

…What [is] most distinctively American about the skyscraper is its purpose, which is to make money. A building, if it is successful, takes on a kind of teleological necessity once it is built, as if no other possible building could have been contemplated in that particular location. We see the arrangement of shapes the architects refer to as the massing, the heart-stopping shimmer in the reflections of the clouds as the wind sucks at the black glass of the windows, the checkerboards of granite on the lobby floor, and to us that is the self-evident end and purpose of all that vast labor. But properly regarded, the physical building is merely the three-dimensional projection of a set of documents known as the pro forma, the calculations that demonstrate how a given piece of real estate can make money. Thus architect, engineer, ironworker all dance to the tune called by the developer.

Of course, for better or for worse, the developer can’t simply build anything that he wants. The developer’s pro forma-shaped-box is constrained by local building rules and regulations: zoning rules that dictate allowable building heights and floor area ratios (the ratio of building floor space to overall land area), code requirements that determine what sort of building systems need to be used (and thus how much the building costs to build), aesthetic requirements implemented by local planning commissions, feedback from design review boards and local residents, and so on.

These constraints can have very large impacts on the size and shape of buildings: the distinctive layer-cake style of early 20th century New York skyscrapers, for instance, was due to zoning rules that required setbacks to prevent buildings from occupying the entirety of their sites. In other cases, securing local approval is so difficult that it ultimately dictates the exterior form of the building, as noted by one Chicago developer in How Real Estate Developers Think:

“…The developer spends a lot of money designing this fancy picture or model,’’ says Emmerman, ‘‘so that they can take it to the alderman, councilman, or mayor, and they either fall in love with it or they bastardize it first and then fall in love with it.’’ Even then, elected politicians will not promote a project until the broader community has seen it and come out in support of it — or at least not come out in complete opposition to it — so the next step is to seek input from the neighbors. ‘‘So then you spend months dealing with the neighborhood and ‘can you make it shorter, can you put the parking underground, can you put in flower boxes,’ and still, we have not talked about how this building functions.’’ By the time the developer commits to proceeding with the detailed design, after many months of promoting the architect’s rendering or model, the image of the building’s exterior shell has become fixed in the minds of the public.

“…Once you have figured out the typical floor plate based on all of these constraints,’’ says Emmerman, ‘‘you try to fit the units into the space that is left over.’’

It’s the developers job to navigate this mass of constraints, and to come up with a plan for a building that will be compelling to renters, acceptable to local residents and building authorities, and earn an acceptable return. The developer is highly incentivized to get this correct: real estate is a competitive market, and a new building project must compete with many other buildings on the market. Developers will thus typically pay very close attention to the local real estate market: growth rates, population demographics, what sort of projects are successful and what aren’t, what features are compelling to tenants, and so on. In some cases developers will go so far as to bring their sales staff to early stage design meetings, and have them weigh in on the design of the building itself, to ensure they’re building something that buyers want and staff can sell.

Developers and architects

To design the building, the developer hires an architect, who will lead the design effort and manage the design team (which includes specialties like structural engineers, mechanical engineers, and so on). The architect is the developer’s agent, and is duty-bound by professional ethics to act in their interest.

Which is not to say that the architect simply does exactly what the developer wants. An architect is hired for their professional expertise on building design, including a building’s aesthetics. A good design that not only meets the developer’s programmatic requirements (number of units, amount of rentable space) but is aesthetically pleasing is one of the services the architect is providing. If an architect advocates for some particular design choice — a finish material, a building shape, a decorative feature — the developer will take that seriously. They hired the architect in part for their aesthetic taste, after all.

But this sort of advocacy works only to a point. An architect typically can’t just go in whatever aesthetic direction they want, with the developer forced to follow along. For one, they can’t stray outside the box built by the developer’s pro forma and by what will be approved by the local authorities. This limits what sort of options an architect has to choose from: they can’t simply pick a facade material that’s twice as expensive, or a building shape that doesn’t have sufficient square footage, just because they like the look of it.

For another, architecture is a very competitive field (there are around 120,000 registered architects in the US), and an architect who makes unreasonable demands or can’t keep projects within budget is an architect who will have a hard time getting future work. Often a developer will solicit proposals from several design teams, choosing the team they like the best or the one that has the lowest fee. Developers tend to return to architects whom they work well with, who understand that a building is ultimately a question of economics, who know what sort of things are reasonable to ask for and what can’t be accommodated. Most buildings aren’t designed by starchitects like Frank Gehry or Zaha Hadid who are trying to create unique works of art, but by client-friendly architects who will hew to the constraints they’re given and try to deliver a reasonably-priced building that meets the design requirements.

One such firm was Emery Roth and Sons, founded by Emery Roth in 1903. While not a household name, Roth and Sons’ influence on the New York skyline was enormous. Between 1950 and 1970, the firm designed 70 office buildings, accounting “for half the [New York] office acreage created since the war”:

…Four consecutive blockfronts on the west side of Park Avenue between Forty-eighth and Fifty-second streets were redeveloped according to the plans of Emery Roth and Sons. These skyscrapers were built to house some of America’s mightiest corporations, and above their portals were emblazoned the names of Bankers Trust Company, Colgate-Palmolive, International Telephone and Telegraph and Manufacturers Hanover Trust. The new, glittering, glass-walled faces of Park Avenue, as well as of Madison and Third Avenues, were almost entirely Roth-fashioned. As testament to this fact, a new word was coined — ”Rothscrapers” — to denote the glass and metal towers designed in the Roths’ modern idiom. They had such a dramatic effect on the New York skyline that Ada Louise Huxtable, former architecture critic of the New York Times, saw them “as responsible for the face of modern New York as Sixtus V was for baroque Rome.”

Emery Roth and Sons’ success was not because of their aesthetics – their designs were highly criticized by architecture critics. (Frank Lloyd Wright allegedly refused to drive down the streets that were dominated by Rothscrapers.) It was due to the simple fact that they had a reputation for delivering low-cost buildings with very large amounts of rentable space. In his book on Emery Roth, Steven Ruttenbaum notes that “Like no other architects in the post-war era, the Roths developed the most economical method to construct tall office buildings”:

What builders paid for, the Roths believed, was not art but utility —maximum rentable space, high-speed elevators, flexibility of plan and comfortable interior climate… One builder stated, “The Roths give you maximum utility without losing an inch of rentable space.” Another observer even went so far as to describe the Roth operation as the “business of designing business buildings for business men who put up business buildings for other business men.”

Meredith Clausen similarly describes the Roth’s in her history of the Pan Am building:

A large measure of the firm’s success was due, as Richard Roth was later bluntly to explain in his essay, “High-Rise Down to Earth,” to their aim “not to create masterpieces” but to provide buildings that worked efficiently and economically for their clients: buildings, that is, that met the programmatic requirements of the client and insured a profit.

These sorts of developer friendly architects abound. In Chicago, architect slash developer James Loewenberg was similarly infamous for filling the skyline with aesthetically forgettable but very profitable buildings. In the early 20th century, Shreve and Lamb (later Shreve, Lamb and Harmon), the firm which designed the Empire State Building, was likewise known for being practical, unsentimental, and keeping their eye on the bottom line, as John Tauranac notes in his history of the Empire State Building:

"The day that an architect could sit before his drawing board and make pretty sketches of decidedly uneconomic monuments to himself has gone," said Lamb. "His scorn of things 'practical' has been replaced by an intense earnestness to make practical necessities the armature upon which he molds the form of his idea…he must know how to plan his building so that it will 'work' economically and produce the revenue for which his clients have made their investment."

With the design of the Empire State Building itself, Lamb would state that “whatever ‘style’ it may be is the result of a logical and simple answer to the problems set by the economic and technical demands of this unprecedented program.”

Developers, especially those successful enough to be able to finance skyscraper construction, are also savvy enough to understand the architecture market: what firms are best at what sort of work, which architects are easy to work with and which aren’t, and what they’re getting when they sign on with one. If a high-profile architect is hired, it’s because the developer thinks that their style, or the cachet of their name, will bring value to the project. But even famous and opinionated architects aren’t necessarily given carte blanche to design buildings however they like.

An example of this is 1540 Broadway, the New York skyscraper whose design and construction was chronicled by the book High Rise. The developer hired Skidmore, Owings and Merrill (SOM) to create the design. SOM was (and is) one of the most famous architectural firms in the world, responsible for designing many of the world’s tallest buildings. At the time, SOM was notable for having fervently embraced architectural modernism, so much so that the three cofounders were sometimes called “the three blind Mies,” after modernism founder Mies van der Rohe.

But impeccable architectural credentials and strong aesthetic preferences didn’t mean that SOM was in the driver’s seat. The developer, Bruce Eichner, required constant redesigns to get the building where it needed to be finance-wise. Major architectural features were constantly removed or reworked: for instance, SOM wanted to have three-foot-deep recesses in the building as a major architectural feature, which would be supported by an expensive Vierendeel truss. But this was rejected as too expensive, and SOM was ultimately forced to accept much shallower, nine-inch recesses and a more conventional steel moment frame. The building’s shape was changed to remove articulation and reduce the amount of expensive cladding required. Interior finishes were changed based on feedback Eichner got from a real estate broker on what tenants wanted.

This sort of client feedback is common with high-profile architects. Minoru Yamasaki was the architect who designed the World Trade Center, but his feet were held to the fire the entire time by his client, the Port Authority. From City in the Sky:

Tozzoli also weighed in. The trade center, he said, was going to be beautiful. That was why Yamasaki had been hired. But this was not going to be some esoteric architectural exercise. The commanding force here — besides the bottom line — was going to be “the Program,” the nonnegotiable set of specifications for the job. It totaled 10 million square feet of net rentable space. The message was eminently clear. This was not going to be a project in which Yamasaki came up with a design and then presented the results to an awed client. He was facing the prospect of an endless interrogation and negotiating and bickering with the Port Authority. His plans would be criticized at every step and changed if they did not meet the test of practicality as defined by Tozzoli and Levy and the others. And it would soon be especially obvious that there was no escaping Levy, a man who could be so nasty he could, and often would, make Yamasaki quiver silently with humiliation or anger. Levy was going to be Tozzoli’s whip hand on the trade center. The cursing, tantrum-prone Levy was going to make sure that the little group of architects in the Detroit suburbs designed the kind of buildings that the Port Authority damn well wanted.

Similarly, famous modern architect Philip Johnson designed Pennzoil Place, which won many accolades for its design (it was described as the “Building of the Decade” by architecture critic Ada Louise Huxtable), but its design was ultimately the result of feedback from developer Gerald Hines and Pennzoil chairman Hugh Liedtke1. From Raising the Bar, a biography of Gerald Hines:

The first thing Hugh Liedtke said was, “I don’t want one of those upturned cigar boxes!” said Burgee. “He said, ‘I won’t have any of that’... Hugh started saying “I want something that when people see it, they say, ‘That’s Pennzoil.’ ...Then he said to me, ‘I hate aluminum, can’t be aluminum, these damn aluminum and glass and mirrored glass buildings.’”

…Johnson and Burgee’s initial ideas were unsuccessful — either too conventional or too similar to the Seagram Building that Johnson had designed in New York. Both Hines and Liedtke soundly rejected all of their ideas. Then Hines, without realizing it, came up with the answer. “Gerry said, ‘we’ve gotta have two tenants,’ said Burgee… “And Philip Johnson said, ‘Well, why not two buildings?’”

“Hugh said he didn’t want an upturned cigar box,” Burgee said. “Well, except for the little cut in the building, the buildings we’d designed were cigar boxes from one side. So we said, ‘We’ve got to do something with the top so it isn’t just a flat, cut-off cigar box.’” So they cut the top off of each building at an angle, leaving two dramatic pointed peaks.

“I remember showing the idea to Gerry for the first time, and he said, ‘Oh, no, no, that’ll never do,’” Burgee said. “‘A point on a building?’ he said. ‘You can’t rent that point.’ ...”And he said, ‘we have to take it to Hugh and see what he thinks,’ said Burgee. “But make the points removable,’ So we made the model so that the tops pulled off so that Hugh could see the building flat and also with points.”

“[Liedtke] saw it, and he was quiet, and he thought about it and said, ‘I like it,’” said Burgee. Then the architects showed Liedtke the model without the points. So we took the tops off, and Gerry said, “Mr. Liedtke, this is going to cost X hundreds of thousands of dollars’ or something. Hugh didn’t answer. He looked at us, Philip and me, and he said, ‘Put ‘em back,’ and we put ‘em back on. And he said, ‘That’s it’ and left. Those points never came off again, and that was the end of the meeting.”

More generally, developers may or may not have strong aesthetic preferences (some care deeply, some not at all), but either way they tend to exercise a firm hand in the design process, putting money where it's useful and cutting it out where its not, and will often be very opinionated about what will work for a given project, based on the market and what tenants are believed to want.

Why developers chose glass curtain walls

So why did developers embrace the glass box aesthetic? Unsurprisingly, it comes down to economics.

Prior to World War II, most skyscrapers had walls of stone or masonry. This was partly a holdover from the days when buildings were built with load-bearing exterior walls. The development of the steel frame at the end of the 19th century allowed buildings with non-load-bearing exterior walls, but codes were slower to change than technology, and still often required a thick layer of stone, brick, or terra cotta on the exterior. But by the late 1940s and early 1950s, building codes were beginning to omit these requirements.

As codes were changing, other technological developments were making thin exterior walls made of metal and glass practical. Air conditioning was becoming common, making it possible to cool buildings with large expanses of windows that might otherwise get too hot in the summer. Aluminum production had risen enormously during the war due to the needs of the aircraft industry, driving down its cost, and after the war manufacturers hunted for new markets that could absorb some of this capacity. Aluminum building cladding and window framing soon became an important new market. In 1948, aluminum-framed windows were just 5% of the market: by 1956, that had risen to 20%. The new headquarters of aluminum manufacturer Alcoa, built in 1952, was described as “one big showcase of aluminum,” with aluminum “walls, roofs, ceilings, doors, hardware, baseboards” and much more. From Tom Leslie’s history of Chicago skycrapers:

The building’s “most important feature…from the standpoint of future large-volume use of aluminum in skyscrapers,” though, was its cladding system of stamped aluminum panels with integral windows, each just an eighth of an inch thick and factory-assembled, avoiding complex assembly on-site.

Glass, too, was getting cheaper. In the 1950s, British glass manufacturer Pilkington invented the float glass process, which produced glass by casting it on a pool of molten tin. The resulting surface was nearly defect-free, eliminating the costly grinding process flat glass had previously required and greatly reducing the cost of high-quality window glass. Between the late 1950s and the 2010s the cost of flat glass fell by 75% in inflation-adjusted terms.

These developments made it possible to replace thick, heavy walls of stone and brick with thin, light curtain walls made of glass and aluminum. And while this style of construction was indeed championed by modernist architects, it also had decisive advantages from a real estate developer’s perspective.

The most obvious was cost. A glass and metal curtain wall wasn’t necessarily all that much cheaper than one made of brick in terms of the materials themselves, but it was much thinner and lighter. Its thinness meant that for two equally sized floor plates, the curtain wall framed one would have more rentable square feet than the stone or brick one. This more than made up for the fact that the thin curtain walls had worse insulation and were more expensive to heat and cool than masonry walls.

Similarly, the fact that a glass and metal curtain wall was so light meant that less structural framing was required to support it. Beams and columns could be made thinner, foundation sizes reduced. Not only did this reduce the cost of the building itself, but it reduced the cost of constructing it: all else being equal, a lighter structure is easier and cheaper to erect. Curtain wall could also be produced as large, prefabricated panels, allowing for very rapid installation on-site and eliminating expensive craft trades like masons.2 With curtain wall, a building no longer needed scaffolding around the edge of the building for masons to work on: the curtain wall panels could simply be swung into place. From The Ecologies of the Building Envelope:

Facades for office towers composed of prefabricated panels could be literally installed within days, and were done so with fanfare to incite a growing clientele… Emery Roth and Sons’ 1954 National Distillers Building, at 99 Park Avenue, featured a prefabricated curtain wall “comprising 1800 two-story-high panels.” It was erected in “six-and-a-half working days, a remarkable feat compared with the eight weeks or more that a conventional masonry facade would have required.” This building was originally specified to be clad in masonry. However, the executives of the Tishman realty and Construction Company “had been impressed with the facade of [Alcoa’s] Pittsburgh headquarters” and decided to switch to aluminum. In the same year, Emery Roth and Sons was commissioned to clad 460 Park Avenue, and astonishingly did so in a single day.

The cost advantages of curtain wall still exist today: if you use RSMeans’ pricing tool on a high-rise apartment building, curtain wall will come out cheaper than either precast concrete cladding or brick (though it will be more expensive than the even cheaper EIFS).

There were also functional advantages of curtain wall that made it attractive to tenants. Large expanses of glass let in lots of natural light and maximized the amount of floor space that had nice views, two things that tenants valued highly (and were willing to pay for). Continuous expanses of glass also made it very easy to rearrange interior office partitions, as Richard Roth noted in an article in Progressive Architecture:

Architecturally strong and impressive as the Radio City masses may be, they do not, with their window-pier-window design, allow for this interior planning flexibility. Therefore — the new exterior design to suit interior plan spacing. It is as simple as that! Window-mullion-window-mullion (445 Park Avenue — Kahn and Jacobs): strip building with continuous windows and continuous masonry spandrels (575 Madison Avenue — Emery Roth and Sons): metal and glass “skin” buildings (Lever House — Skidmore Owings and Merrill): and the metal and metal with the windows losing their accent in the dominance of the metal (99 Park Avenue — Emery Roth and Sons).

Developers and ornamentation

But the popularity of curtain wall is only half of what needs to be explained. A preference for glass and metal cladding, wouldn’t, on its own, eliminate ornament and decoration on buildings. Ornamentation and glass curtain wall aren’t mutually exclusive.

The elimination of ornament also comes down to economics. As we noted earlier, developers place a great deal of weight on what tenants want, and what they’re willing to pay for. And the reality is that while tenants care a great deal about what the inside of the building is like, they pay much less attention to the exterior of the building. Some developers, such as James Loewenberg, state this explicitly. From How Real Estate Developers Think:

“The person who looks to buy or rent a unit in a high-rise,” says Loewenberg, “only cares about three things: the location of the building, the layout of their unit, and the view from their unit. They don’t care as much about the physical appearance of the building and it is my contention that they never really look above the third floor…”

Loewenberg recalls being roundly criticized by another critic for a project with a parking garage finished in exposed concrete. “He asked me, ‘Well, why are you putting marble thresholds in the apartments and why didn’t you spend the money on better finishes for your concrete garage?’ I guess it was heresy when I said, ‘Because people don’t care about what a garage looks like and in fact they would rather see a marble or granite countertop than a magnificent exterior on a garage.’ That critic,” says Loewenberg, “was trained as an architect, so he looks at the outside of the building and thinks that no shortcuts should be taken and that you should only do beautiful architecture. I, on the other hand, would rather put my money where it counts, and I happen to know that it is the finishes that sell.”

Other developers aren’t quite so explicit that they don’t care about exteriors, but nevertheless echo these sentiments. In his book Powerhouse Principles, developer Jorge Perez emphasizes that as a developer you should “spend your money on work that can be seen”:

Obviously the building has to meet standards and codes and everything else, but what sells it is bathrooms, kitchens, flooring, landscaping, and amenities, those things that are highly visible and that people consistently note as higher priorities… We have found, particularly in the for-sale product, that changes you made to bathrooms and kitchens were of extreme importance. People will pay you an additional amount of money for those expenditures. If I spent $2 and I did it correctly, they would pay me $4 for the more beautiful kitchen.

Perez cares a great deal about the design of his projects, but his focus is on things where people spend their time: in their units, in the amenities, in the lobby, and so on. The interior designers he uses merit multiple mentions in his book. And while he also mentions a few high-end architects he’s worked with (Arquitectonica, Robert Stern), his buildings nevertheless tend to end up as basic boxes. From How Real Estate Developers Think:

While [Perez] hires award-winning architects to create his building’s exteriors, like Jim Loewenberg he also knows that it is where people live that matters, and that is why he focuses on the interior design of the public areas and the amenities.

If exterior details and ornamentation are not something a tenant is willing to pay for, they’re not something a developer is inclined to spend money on. And simpler, less ornamented buildings are cheaper to build. Ornaments and details cost money. Articulation and interesting facade shapes increase the amount of (expensive) exterior cladding without a corresponding increase in rentable square footage, and add joints and connections that become potential points of failure.

When a developer does add ornament, it’s often in relatively limited ways in an attempt to get the most bang for their buck, ways that don’t change the size or shape of the building: making the windows a slightly different color, adding luxurious touches to one small, highly visible corner of the building, and so on. One developer described this process to me as “applying doilies.”

An example of this is Aqua Tower in Chicago, designed and built by James Loewenberg. Aqua won awards for its design, which features an undulating shape due to its curved, differently-sized balconies that extend beyond the edge of the building.

Loewenberg notes that this design was incredibly simple and inexpensive to create:

‘‘it is the simplest box we have ever built, and in fact, the hardest part of the job was convincing the contractors that other than the cost of the materials to extend the balcony slabs this would be no different from building the basic box.’’ Ultimately, Loewenberg was successful: ‘‘I finally browbeat the contractors enough to talk them into it.’’

…In the end, the only additional costs incurred to realize such a unique exterior design were the material costs of the extra concrete at the extended balconies and the labor costs of the two workers installing the edge form on each floor… “The total cost for all of it was probably a couple of hundred thousand dollars, which is nothing on a building of this size.’’ Aqua is eighty-two stories and contains nearly two million square feet of hotel space, apartments, and condominiums, and its total cost was about $500 million, so a couple of hundred thousand dollars to get that magnificent effect truly was peanuts.

The financialization of architecture

Of course, this desire to eliminate ornament which doesn’t contribute to the bottom line didn’t emerge suddenly following World War II: it had been a trend for decades prior to that. In his history of the Empire State Building, John Tauranac notes that in the early 20th century, real estate investors were becoming less and less inclined to fund “spurious” ornamentation that wouldn’t yield greater profits. This paragraph sounds like it’s describing modernist glass skyscrapers of the 1950s, but it's actually about Art Deco skyscrapers of the 1920s:

…The demands of tenants probably more than the desire of avant-garde architects was responsible for the popularity of the modern style, of this functionalism. Tenants wanted the light and space that modern buildings afforded, and were willing to accept modern design to achieve their needs. Suddenly designs that only a few years before might have seemed radical had won a more secure place. It wasn’t that ornament cost so much more, but even if the syndicate had been willing to pay for the little extra for it, there were more fingers in the fiscal pie than those of the four primary underwriters. The Metropolitan Life Insurance Company was a conservative investor whose goal, naturally enough, was to secure the greatest possible return upon invested capital. Underwriters were willing to tie up their investment over a period of many years, but less willing to pay for what was deemed redundant decoration. Hood told of one banker who rejected a design that was all gussied up. When the architect returned with something more practical, the revised plan was financed. Hood said that architects and developers were catching on to the fact that mortgage money would more readily go to office buildings in the modern, functionalist style than those designed in a more traditional style, a trend that would continue…

This trend towards “financialization,” of viewing a building increasingly in economic terms and only adding features that could be financially justified, continued over the course of the 20th century. Following World War II, developers embraced modernism because it provided an excuse to construct buildings less expensively, notes Steven Ruttenbaum:

For builders, it was getting increasingly more expensive to erect commercial structures and to hire craftsmen to decorate them with special flourishes. It is not that ornament became prohibitively expensive, it merely entailed more of an initial investment than builders were prepared to risk. They were, of course, in business to make money, not to “squander” it on things that had no relation to their buildings’ utility. And after World War II, it became increasingly easy for them to accept austere facades in light of the growing acceptance of the International Style. In the late thirties, the great European masters of modernism arrived in America as refugees from the Nazis, and they were fortunate in securing some of the nation’s most prestigious academic positions. It was only a short time until the economical style they espoused became accepted by speculative builders.

In Skyscraper Dreams, Tom Schachtman likewise notes the bland designs of Emery Roth and Sons were in large part due to client feedback:

…when they did try flights with more fancy, as often as no Richard and Julian were asked to scale back the more extravagant touches. Years later, when asked why he did not design more interesting buildings, Richard was heard to mutter that if he’d had more innovative clients he’d have produced more innovative designs. This was more humble than necessary: the designs were better than the critics suggested. And perhaps the forelock-tugging inherent in the statement was subtly encouraged by the Roths’ developer clients, to reinforce awareness of the minor place of the architect in the builder’s grand scheme. In a period in which beauty was no longer perceived a great sales tool, artistry in buildings was not assiduously cultivated.

Similarly, in his analysis of the architectural profession in the 1980s, Robert Gutman notes that architects were increasingly pressured by clients to economically justify the aesthetic elements of their designs, in part because clients were increasingly organizations rather than individuals:

The most significant characteristic of the organization clients is their disposition to view architectural production from purely a rational and instrumental perspective. This means that the organization clients regard buildings as capital assets, which should be managed like every other potential source of productivity, income, and profit… All features of buildings come to be judged by these criteria, including the esthetic dimension, which traditionally was considered as outside of the realm of this system of calculation.

The prevalence of the instrumental mentality in the building and design process represents the latest stage in the historical transformation of the architect-patron relationship, a process that began in the middle of the nineteenth century.

Even non-profits aren’t immune from this sort of pressure:

A fact that often surprises architects when they discover it for the first time is that the instrumental view of architecture has also taken over non-profit institutions, which formerly took a more casual view of their physical plant and regarded buildings primarily as works of architectural art. This new mentality regarding buildings is especially noticeable among university and cultural organizations in the non-profit sector. With the expansion in the number of universities, museums, symphony halls, and theaters, competition for funding is rapidly becoming more intense. Curators, presidents, and other managers of these institutions are under increasing pressure to justify their building needs in terms similar to those that apply to the profit sector.

This tendency to omit decoration that buyers don’t seem to value isn’t limited to modernist, International Style skyscrapers. Consider, for instance, this single family home, which is fairly representative of new home construction in the US.

This home obviously isn’t anything resembling “modern” architecture, and the front of the house has a great deal of ornamentation. Stone cladding, decorative shutters, faux lintels over the windows, and so on. On the sides, however, there’s virtually nothing: no stone cladding, no decoration on the windows, almost no ornament whatsoever. (In many houses, even the decorative muntins on the windows will be omitted on the sides and back of the house.) Builders apparently believe that buyers don’t particularly value ornament other than on the front of the house, and so they omit it everywhere else. And since it’s very rare for owners to go and add any sort of ornament back in, it seems like the builders are correct.

Part of this is a problem of legibility. For some building features, like granite countertops, stainless steel appliances, or washer/dryer hookups, developers can quantify how much they’ll contribute to additional rents, and how much it's worth spending on them. But for exterior ornamentation, or simply an attractive exterior design, it’s much more difficult to quantify the benefits of it. If it can’t be included as an element in a financial model of the building, it's hard for a developer to justify spending money on it.

More generally, there seems to be a certain amount of market inefficiency in the construction of new buildings, for reasons I don’t quite understand. Some developers have told me that the value of good design is the sort of thing that accrues over the long term. A developer putting up a building and then holding on to it might be able to see that value (though convincing outside investors is another story), but for a merchant builder that’s going to turn around and sell it, it won’t be reflected in the price they’ll be able to get for it.

Developers and risk aversion

For these reasons, the unadorned glass box aesthetic became popular. And once a construction system becomes popular, numerous forces conspire to keep it in place. For one, a huge number of decisions in real estate are based on “comparables”: comparing your building to a similar building targeting a similar part of the market. This ultimately pushes new buildings to be similar to what’s already being built. Investors are risk-averse, and don’t want to fund a project that’s too different from what already exists and what the market expects. Developers don’t want to make risky design decisions, because bold choices are perceived to be polarizing, and are more likely to drive away tenants than they are to attract them. And for a developer, the downside of a project going poorly is much greater than a project doing exceptionally well, which incentivizes going with the tried and true. The more common a style becomes, the riskier it is for a developer or investor to deviate from it.

A similar logic drives architectural and engineering decisions. Architects and engineers want to use systems they’re familiar with and have used before: they already have the details and calculations for them (saving them time and effort), and the systems have been vetted and demonstrated to work. An architect who specifies a glass curtain wall that he’s used many times before can be confident that the system will perform as expected, that the waterproofing will work as designed, and so on. If instead he specifies a novel system, he’s both making more work for himself and taking a major risk in doing so.

On the construction side, contractors similarly prefer to use the tried and true rather than something novel with significant downside risk. On Aqua Tower, the contractor was extremely reluctant to build the novel curved balconies, not because they were expensive but because they were different. And if a construction system stops being used, the skilled labor that knows how to install it eventually disappears from the labor pool. Architects and developers have told me that one hurdle for specifying ornate masonry or stonework is that in many cases there’s a really limited number of skilled practitioners that can do that sort of high-quality work. It’s much easier, cheaper, and less risky for a developer to simply use the systems that local contractors are already familiar with.

Is the glass box a market failure?

So ultimately, the glass box aesthetic became popular not only because modernist architects liked it, but because it made buildings cheaper to build, and because developers gradually realized that ornate, highly-detailed exteriors weren’t particularly compelling to renters or to their investors. And once that aesthetic was in place, risk aversion and cost considerations kept it in place.

Of course, none of this means that it’s impossible to build anything other than a plain glass skyscraper. Many developers care deeply about good exterior design, believe more classical design adds value and makes their project more competitive, and get funding to build those sorts of projects. Architects like Robert Stern and Peter Pennoyer have designed and built several skyscrapers with brick and stone cladding, and interesting detailing. But these sorts of efforts are to some extent swimming against the current, which nudges developers towards a simpler, more modernist aesthetic.

One could argue that there’s a sort of market failure at work here: because architecture is ultimately funded by the people who occupy a building but viewed by people outside of it, there are externalities (in the form of benefits of attractive exteriors) that aren’t being appropriately priced in. An “efficient” market for architecture, which used some sort of mechanism to properly weigh the preferences of everyone who has to look at the building, might be expected to produce more beautiful buildings (for whatever your definition of beautiful is).

Unfortunately for proponents of this theory, we already have such a mechanism: as discussed above, building design is heavily influenced by design review boards, planning commissions, and other forms of community input. Our current system bends over backwards to consider local preferences and give weight to these sorts of externalities. This doesn’t seem to have had much of an impact on how “beautiful” buildings are, or the dominance of the glass box aesthetic: all it does is make them harder to build.

LLMs shouldn't help you do less thinking, they should help you do more thinking. They give you higher leverage. Will that cause you to be satisfied with doing less, or driven to do more?

— Alex Komoroske, Bits and bobs

Tags: llms, ai, generative-ai, alex-komoroske