The Death Cross:  Diversity, Corporate Governance and AI.

A random walk in the past – Diversity, death by 1,000 cuts – Corporate governance, just keep borrowing – AI, Actual Implementation – Elon Musk, a misunderstood guy.  

Summary: Diversity, Governance, and AI will cause revolutionary changes over the next 20 years. It’s likely though that the early years will bring catastrophic losses and bankruptcies before AI adds to productivity in an appreciable way. We do not believe that most of the companies that exist in tech today will lead the way.

The death knoll of multinationals.

A random walk through the past.

I have witnessed two tech transitions in my professional lifetime. The first was the “computerization” of the oil industry. The second, the tech collapse of 2000. Both had similar characteristics.

In the late 80’s everybody had to be into computers. It was the future. Most of the engineering systems at that time were analog, but management was desperate to seem techie. 

The company I worked for, a contractor to Aramco, was into wireline engineering. We used tools connected to wireline that went to the bottom of oilwells. As you can imagine, with all the heat and pressure of an oilwell, you can’t simply snap your fingers and digitize. Even if the surface systems worked, no one had yet invented chips that operated properly at 250 degrees. 

So, we (ingeniously) invented digital readouts. Like a digital clock they were analog signals that displayed numbers. 

The transition to real digital was long and expensive, 

As you can probably imagine, everything from the surface equipment to downhole sensors had to be changed. Not modified, replaced. It was a total hell-scape. Even if downhole and surface worked, since the data was in bit form and not an analog readout, it was hard to know if it was accurate. 

We once charged Aramco $45,000 for a data set they had to confirm using surface gauges. Thank Allah we weren’t in the US. The driller would have torn up the invoice while he laughed. 

In the end the change revolutionized the oil industry, adding 30 million barrels per day of worldwide oil production. Oil was found hidden in analog data and drilling efficiency was increased by an order of magnitude.

But it wasn’t cheap, and it wasn’t easy.

The 2000 transition was worse.

Tech was a bunch of overhyped bullshit stories.

Having lived through 1990, I was ready for 2000 and shorted the markets accordingly. A trained monkey could see that technology was light years behind what the Silicon Valley geniuses promised.

Internet speeds were slow, and even then, the web was barely connected. It was near impossible to upload a few kilobytes and before it was finished, you got an upload error. 

It didn’t matter though because money was flying into tech like… well… like it is today in AI. Everybody had to be in the game. More about that later. 

Diversity, death by 1000 cuts.

Rather than rewrite the wheel, Complex Systems Won’t Survive the Competence Crisis, by Harold Robertson is great article on the subject. (1)

Robertson’s thesis, which I have witnessed myself in real time, is that diversity hires are gumming up systems and causing widespread inefficiency in companies. 

To be sure, everyone should have a chance. In the corporate world, the last thing a CEO would want is a great hire held back by a random personal attribute. But in systems engineering, having to work with an unqualified diversity hire is like slamming into a brick wall. 

The boss’s son.

In Landon Fillmore, we have seen this more often than not. As Rodney Dangerfield once said, “It was actually my grandfather that started the business because my dad and myself are both idiots.” 

Nothing makes us walk away from a business faster than the words “and son’s” emblazoned on the sign. But is hiring an unqualified person so your company can appear diverse on your brochure any different than slapping your kids name on the sign to make yourself look like a super parent?

 The short answer is no.

The timebomb in the cubicle down the hall.

Every manager has had to work with at least a few bad hires, but when they are systematically placed in positions across the company, it’s like putting landmines in a field you intend to cross. 

I have seen this in large companies more often than not. It’s bad enough when companies are growing rapidly and you simply need asses in chairs, but when you are forced to hire based on anything other than competence, everybody suffers. Employees are forced to work around the incompetent employee, fix their mistakes or worse, sit with them in zoom meetings. 

And it’s not just low-level employees.

I once sat in on a meeting where the CEO had the entire staff, thousands of employees, log in to see her give a meeting from a family picnic on a Tuesday at 2PM during the busiest time of year. She filmed it on her iPhone, live. This was at a time when the company was begging employees to work overtime. 

In the meeting, she showed her employees her “wonderful” (fat) husband, her “wonderful” (spoiled) children, and her wonderful back yard. She started by stepping out of a BMW SUV. It was the stupidest meeting I’ve ever seen. 

Diversity is a profit margin destroyer.

We have experienced it in Landon Fillmore. No one is exempt from it. In bureaucratic systems, 75% of the productivity comes from 20% of the employees. Adding incompetence to the 80 percenters will exasperate the situation.

Corporate Governance, just keep borrowing.

There are two numbers on a balance sheet that we look for first. They are the quick ratio and net tangible assets. 

Anybody that’s dealt with financial statements knows they’re total bullshit. You can make any company with sales look profitable. If you want to see real fake news, just look at an income statement. Earnings are estimates even after the year is in the books, 

But you can’t lie about your quick ratio. This is the ratio of your current assets, (cash, and accounts receivable) and your current liabilities, (accounts payable, the current part of long-term debt and unearned income.) The second part, shareholders equity is more complex. I’d encourage you to study it on your own. 

Things that keep Warren Buffet up at night.

We really don’t care what pile of bullshit appears on a financial statement. Our only concern is, “what are you doing with the money you claim you earn?” 

In the last ten years it’s been largely pissed away. How they piss it away is miles beyond the scope of a blog post. Unless you have a working knowledge in financial statement analysis, you probably won’t see it. 

That said, there are few things more existentially dubious than taking out a loan to buy your own stock back. (Using all of your income to do it is second.)

When will we know how bad this is?

Wait until the next stock market crash. During these events you get a normalization of accounting systems. The crash gives CEO’s the ass cover they need to bury all their bad investments, absurd assumptions, and loony practices. 

Then, they reprice their “inventive options grants” to go back to their business of lying, cheating, and stealing. As the great financial educator 

The lesson is simple.

When it comes to governance, never trust a financial statement. You should think of it like corporate breast implants. It diverts your eyes from what you should be paying attention to. 

As the great Financial Analyst Leopold Bernstein once said. “Accounting is a social science.” Nuff said? 

AI, Actual Implementation.

At Landon Fillmore, we interact with a lot of software engineers. The guys in this business are unique. First and foremost, the top-level guys are very smart. Our contacts come from Germany, Holland, Spain, Mexico, and Russia. 

The “diversity” comes from the fact that they are 1%-ers. The represent the extreme right of the bell curve and are extremely rare in any country. 

Most of the engineers we know do not work in AI. So how do they know it? They all use it, extensively. 

Code is like a programs DNA.

Code is literally strings of numbers and letters that generate meaning. 

Software engineers and coders use AI to write code. A popular site is ChatGPT. They simply ask Chat to write a line of code and miraculously, it appears. But here’s the rub.

AI is a search feature. As such, it’s a hierarchy of indexes. For it to be effective, it must navigate the universe of data on a subject, determine what is most relevant and then determine if the sources are credible. 

Have you ever used Siri?

Siri is, in my opinion, a totally useless piece of garbage. Why? Because it simply references internet search pages 90% of which are just online sellers hocking stuff.

So, AI’s first challenge is GIGO. Garbage in, garbage out. Over the last twenty years, search has become next to worthless as multinationals gave used their money to game it. 

Companies are starting out behind the 8-ball.

At least, at first. Keep in mind, companies large and small are already at wits end trying to streamline “efficiency” programs they bought over the last ten years.

Everything from Client management systems to inventory programs. The problems have come from the following areas. (To name a few.)

Inventory management.

So, you think it’s just ‘add a bar code’ and it manages itself? I had a friend that managed a big box home and garden store. In one year, they “lost” 45 refrigerators! They had 8 tankless hot water heaters, sold one and had none left. These were the biggies. They lost hundreds of hand tools and hundreds of square yards of carpet. 

Much, if not most of this was stolen. But controlling shrink was easy when money was free and profit margins huge. You simply wrote it off. But this is no longer the case. Inflation is biting hard, and costs are up. The problem is the solution. More physical labor. 

Is adding AI going to make this easier?

Client management systems.

A client of ours is a management consultant. His first rule in consulting new companies is, “Get rid of your CRM.” Because they don’t work? No, they do work. 

The problem is, they’re inefficient. If you have salesmen, I can assure you, your productivity will drop because they will spend more time spanking the monkey with their phones than making sales calls. Were that not enough, huge reams of CRM data are useless drivel. 

We do not use a CRM at Landon Fillmore Systems for a number of reasons. First and foremost, they are absurdly hard to implement. They require continual updating and input. How much time are you going to spend inputting data on prospects that will never be clients? 

Finally, we simply don’t trust the data security any cloud system has. If your data gets hacked, it won’t be through us. And keep in mind, it’s not just some Russian working in an abandoned factory in Krasnoyarsk. It’s more likely a disgruntled former employee!!

HR and payroll. 

10% of your employees will work like dogs. Another 10% will be somewhat efficient. The last 80% will spend their days trying to game the system. This is probably a company’s biggest problem. 

An old client of ours once said, “If my employees spent as much time thinking about work as they do inventing excuses for why they’re never on time, I’d be rich.” 

This is not an exaggeration. Having not been an employee in 30 years, it stuns me how few people are honest and work hard. Online working has made this demonstrably worse. 

My point is, adding AI to an already stressed, half-managed cluster fuck will not be easy. Considering the oil industry in 1990 and tech in 2000, I’d expect AI to be another nail in the coffin. And believe me, companies WILL add AI. They are keenly aware of what I mentioned above. 

They will FLY into AI. This could be a great opportunity to short companies. 

To summarize: 

Corporate America is under heavy stress to become more productive with systems they already have. Tossing a new system on top of that will make the situation demonstrably worse.

If you recall, in the early days Microsoft would not use ITS OWN remote office software, which was why it built its huge campus in Redmond, Washington. AI will be first implemented by B-school wazoos that don’t understand the software but want to look smart at their next TED talk.

Diversity will not die quickly and removing it from the system in many cases will be impossible. Corporate Governance will hide many of these issues but eventually somebody will have to pick up the tab. In our opinion, perhaps half of the S&P 500 will not survive the transition. 

Elon Musk, a misunderstood guy.

To understand Musk, you should ask yourself this question. What did he invent? The electric car? Hardly. Online payments? Nope. The rocket? Haha. Neuralink? We can give him that one. 

Musk’s beauty is that he isn’t just an engineer, he’s a systems thinker. A good article on systems thinking can be found here. (2) 

Musk’s greatest attribute (Aside from his huge balls) is the fact that he can change on a dime. He is keenly aware of the fact that businesses can expand exponentially if they operate successful forward loops and can avoid being stifled by rigidity. 

In time, all natural systems, be they business or environmental, become rigid. And some rigidity is good. Human DNA is extremely rigid. If any strand has too many errors (mutations) it either dies or the organism dies. 

If you think of an assembly line, imagine that you sub out a process to another company. If that company fails, you fail or at least, get gummed up until you fix the break in the system. 

In fact, cells as well as DNA itself has highly intricate repair systems. 

Nobody moves faster than Musk. 

Musk is not a B-school graduate. He literally designs cars on the shop floor. The future CEO will be like Musk and, in fact, CEOs in the past were like Musk. 

But most companies cannot change that fast. I suspect that in the next 20 years, half of the S&P 500 will disappear. Given that the average S&P company is 18 years old, and their average debt maturity is 21 years, it hardly takes an aluminum foil hat to believe this. 

Twitter is a far better company than it was pre-Musk, and he did this after firing off 80% of the workforce. I can’t believe the B-school wazoo’s running todays companies can pull that off. 

A woman who runs a companywide meeting from a family picnic with an iPhone is not a likely candidate for innovator of the year.

(1) Complex Systems Won’t Survive the Competence Crisis.  Harold Robertson.  Palladium magazine. 

(2) Systems Thinking: The legacy of Steve Belichick. Landon Fillmore Systems