State of The Union 2023: Better Than I Could Have Hoped For
Last year, I highlighted a few key areas:
That the most efficient way for companies to continue forward is to fire product engineers while retaining their infrastructure engineers.
That fraud was becoming especially rampant.
Especially amongst influencers.
Let me be clear: I did not explicitly predict that companies were going to fire people so brutally or so quickly. Nor did I say explicitly that influencers would be caught up in the incredible unwinding and downfall of crypto. Nor did I even suggest that certain tech companies would be on the brink of collapse.
I merely called out the moral degradation and decay in standards in these areas and I lent a small suggestion that what we experienced in 2022 was a small possibility of what could happen. Yet it blew my mind how quickly everything unraveled. In a single year, workforces were slashed, companies became insolvent, CEOs liquidated their own companies while shaming employees into becoming their exit liquidity, and influencers took drastic steps to hide their misdeeds.
Thus, for me to claim proper credit, I will ensure that in the future I will be explicit in my predictions. But do keep in mind predictions are, more or less, worthless and unactionable. But thats for another day.
I do want to start with the financial side of things. This will lead me to the influencers themselves and finally, how it affects the engineering world and our clients. Note that this address will be more geared towards business. Yet it is very important to understand that both engineering and business are joined at the hip today. The massive layoffs that occurred in 2022 because of economic predictions should make that obvious.
Brace yourselves.
I'm not a businessman, I'm a business, man
Despite the layoffs and economic conditions, I am proud to say that my track record for helping my clients remains relatively untouched. This section is purely dedicated to the long-term clients as they are a better record of my efficacy.
My arrangement with my long term clients are as follows:
Clients work with me until one party decides to not continue
Take a percentage of their offer. Otherwise, no fee will be incurred.
There are some other nuanced clauses that are designed to prevent malicious conduct but for the most part, every agreement has been done with a handshake and I believe both parties trusted each other enough to do business without a contract. The contract is just a formality.
Below are a list of my long term clients that I have worked with this year. First initials are used in cases where the client does not want his name exposed. All these clients’ sessions are livestreamed and available on my Youtube channel in their respective playlists.
Matheus: Offer Accepted on MSFT, also passed AMZN. NewGrad L3 offer.
J: Passed G but headcount was filled. No offer.
C: Failed AMZN. Rated as L4 at Google, no uplevel. No offer.
On paper, this looks like an embarrassing year. 1 out of 3 people managed to get an offer and I only had a hit rate of 2 out of 5 for FAANGM interviews this year. But if you dig into the numbers, you will notice that 2 candidates both passed Google’s interview but were rejected due to hiring constraints. Namely Google freezing hiring.
Adjusting for this, my record rises to 80%.
But alas, to hold myself to strict standards, I should accept the 40% mark for this year. I should no more accept an “adjusted” success rate than companies should accepted “adjusted earnings” as a measure of profitability.
This year, there were 2 headwinds: the first is that this year, my clients are not doing multiple interviews. Historically, once I’ve prepared my clients, they can go out and do 3 interviews and we can adjust our preparation and fix any errors after every interview. This would ordinarily improve our hit rate percentage per client as they got better with each attempt. This year, we did not have that luxury and we were bound to an upper limit of 2. While my system is non-ergodic, I’d like to ensure that there is little difference between the non-ergodic and ideally ergodic version.
In layman’s terms, my stats will be different if I have 1 client doing 100 interviews than if I had 100 clients do 1 interview. I want to ensure the gap between the two is as small as possible so that my preparation leaves no holes and remains complete and effective.
The second is that with the recession, companies have frozen hiring. Across the industry, their headcount and evaluations have become more strict, the offers have been reduced. However, in Matheus’s case, the offer was still in the upper 25% of the band (according to levels.fyi) at $171,000 for an L3.
I’m still proud that we’ve hit the interviews and gotten offers from the attempts we have undertaken. In 2023, I am aiming to still support J and C and get them offers. It would be such a shame to let the effort go to waste.
I still believe we are still getting extraordinary results in a time when the competition is getting completely crushed.
If you are recently laid off, looking to guarantee a job offer, and satisfy the following requirements:
Willing to give up every minute of your free time for the next 8 weeks
Have tried everything else out there
Are disciplined and focused.
Aiming for L4+ offer.
Please drop me a line. I guarantee absolute discretion for the consultation.
Tech Course Are Dead
I feel confident enough to have a eulogy for tech Youtube and the courses they sell.
There are still great channels out there but nobody is under any delusion anymore that you can skate by on crappy content about sitting at your computer, attend meetings, working 11-3, and eating ice cream all day. They’ve all gone to Tiktok.
Tech Youtube has benefitted from a few compounding factors:
The zero interest rate environment has allowed tech companies, who’s stock price is heavily forward looking, to grow at crazy rates.
Since one leg of a tech salary is the RSUs, that component grows compounded at the same rate as the stock. Imagine having your salary compound at near 15-30% growth for 5 years! Your salary would double roughly every 3-5 years. Combine that with the fast growth of salary at the early part of one’s career, you can advertise that people with 2 years of experience make 300k or more in tech.
The dominance of mobile phones in the past 10 years has created more source of revenue, more exposure for these companies, and more connectivity (for better or worse). Constantly being plugged in means constantly being served ads and content, whether that is brand awareness from Big Tech or influencer videos.
It is truly awesome. But the landscape has slowly shifted. The end of free money, the collapse of tech stock prices, and the highly competitive nature of content creation has caused these tech influencers to pivot their content away from tech. They have been replaced by adults who actually understand tech, what it means to build a career in tech, and have done so over a long period of time.
I welcome this, especially since its a breathe of fresh air from charlatans who spend 2 years at Google, call themselves ex-FAANG, and market a useless course. I still don’t agree with everything they say but I at least respect where they are coming from and their intensions. At least thus far.
Speaking of, let’s dig into that for a second. How can it be that course sales are falling in bad times? During a recession, college enrollment increases massively because people want to increase their job prospects. If these courses are so good and provide far better value than college, then you should see at least some level of sales increase. After all, these courses promise you a job and better prep than college educators.
Very strange how silent they are now though. And even more peculiar, why are these educators marketing crypto and web3 adjacent technology courses when there is little proof of its real world efficacy? Engineering is already hard enough and you can spend your life digging into every nook and cranny of a bunch of existing complex products and teach those instead.
I work on build systems, programs that can efficiently compile massive multi-million line codebases. It is a product so niche you will never get a college course on how one really works. Yet its how these major tech companies build their applications and they have invested a significant amount of time and money into building and creating these systems. Google has Blaze which they started in 2006 (Bazel is the open source version which was released in 2016). Facebook has Buck which they released in 2013. Gradle exists. But this is a very complex area that most companies would be instantly interested in hiring.
So why not do a course in this? Why not make courses on things that are actually useful? If your course is so good, why are more people not subscribing to them now that layoffs have happened?
When The Tide Goes Out
Alright, let’s pivot to some financials.
To establish credibility, I think its time I show people what is in my portfolio and its performance of the positions. Last year, I showed my YTD performances. I will reiterate them here and my annual performance since 2016:
I will name my positions but not their allocations to prevent people from estimating the net equities value (I am not exactly comfortable with envy).
AAPL +250%
BABA -30% (Added last year, tax loss harvested the initial buys at $160. Cost basis is around $90)
BRKB +4% (I elected to add BRK.B at a cost basis of around $300 this year instead of Apple)
SENEA + 22% (Added this year)
DIT +96% (Added in 2021)
INTC -36% (Added this year at a cost basis of $48).
VZ -29% (Added this year at a cost basis of $55).
OPNT -20% (Added last year at a cost basis of $24). Company has been acquired for $20/share with a $8 warrant potential should OPNT cross certain revenue thresholds.
I will publish the annual performance of the portfolio of my past years as well:
Winners
DIT was purchased because they distribute products and have reasonably high turnover. Their product? Selling cigarettes (and some health foods but that’s negligible). I thought it was highly unlikely people would quit smoking during the COVID shutdowns and smokers tend to be price inelastic (they don’t quit smoking just because packs go up $1). They also tend to be extremely loyal to brands (a Camel bro doesn’t just switch to Marlboro unless there is nothing else, and even then there is a lot of resistance).
SENEA was purchased for similar reasons. They sell and distribute canned veggies under various brands. My reasoning is that because of how America is built, food deserts become quite common and so canned veggies are cheaper than the fresh stuff. In inflationary times, people are still going to want to need to eat their veggies (although given the trend of American diets, this is probably not going to be future proof) and will substitute the fresh stuff for the canned prepackaged goods.
Losers
I don’t believe BABA was an error. I can stomach political risk and saber rattling. I also have no belief that BABA has any significant fraud on their accounting because it is not in BABA’s interest to piss off regulators in other countries when it has global ambitions. I did get hit with the unexpected 0 COVID policy macro event. I usually ignore these macro events for long horizon holdings. Plus I got my purple hoodie by asking merchants on the BABA platform to make it for me.
Incredibly, I DCA’d down and I’m still outperforming the S&P! I had a chance to double down but unfortunately was caught “sucking my thumb.”
INTC and VZ were plays where I wanted the stability of infrastructure and the dividend yield. I was treating them as bonds with an infrastructure upside.
Unfortunately, the market badly punched INTC in the face when it was revealed that the surge in PC buying brought forward the demand for CPUs. The reason for buying INTC is simple: it remains the dark horse in the PC chip manufacturing race behind TSMC and Samsung. It is investing in fab plants in the US to gain dominance. I figure that at my cost basis, I have an equivalent downside protection: that INTC can be liquidated for the amount that I bought it for should the plant expansion plan fail after struggling to sell chips. I have no plans to add to this position and it accounts for less than 1% of my position.
VZ lost subscribers to TMUS. I had judged VZ to have enough pricing power to not lose customers and was the most underpriced in the oligopoly 3: TMUS, ATT, and VZ. I was wrong about the pricing power (TMUS has the highest customer satisfaction rating while the rest have average) but I believe I am correct on other factors and will hold the position for now. One major reason is that the telecom industry is an oligopoly between TMUS, ATT, and VZ with high capital investment as the moat. On top of that, there is a significant amount of structures that are shared inducing a classic balance of power. It also prevents a prisoner’s dilemma in that it is too costly to betray the other companies.
These oligopolies usually do not devolve to duopolies or monopolies without significant repercussions from the SEC. These stocks represent dead money but I will unwind them when the price meets my cost basis.
As for OPNT, this one I intend to hold but I also realized I overpaid for. I believe that there will be an increased use of heroin and opioids and therefore, the increased use of narcan and narloxone. However, I believe that I overpaid simply because of how unstable the historic revenues and performance of OPNT’s fundamentals were. I went by the name brand: everyone calls the any generic kit of narloxone “narcan”. Name power, branding, and all that.
Again, I had a chance to double down when it hit $10 but I was, as Buffett calls it, “sucking my thumb”: sitting on my ass doing nothing. The stock dilution and the crazy lawsuits scared me. I did not venture into the realm of legal arbitrage and decided to sit on my ass.
Review
Overall, your typical Graham net-net plays did amazingly well. I have tended to avoid these for the longest time because of Buffett but I believe that when working with small sums of money, your advantage in illiquid and small markets is so amazing that it overcomes any mantra of “wonderful business at a fair price”. A company in the latter will be able to compound at 20% per year for 20+ years: Apple and Costco are good examples of that. I simply bought these because I didn’t see any opportunities in other companies so I figured the next best thing was net-nets (companies that were trading below net current asset values or P/B < 1).
In general, most valuations close their gap within 18 months. This means that net-nets are only really worth it when you can achieve a 40%+ CAGR since this would make the profits equivalent to a 20% CAGR long term holding. To account for the possibility of being wrong, one should only invest in companies that are at least 50% undervalued to what they will be in 18 months.
This is terribly difficult but doable. I spent this year staring at blue chips and value plays when I could have been hunting the OTC market. This is an unforced error that must be corrected. I hope to only swing at companies that are the next Costco
One last question: why is my performance so closely indexed? Two reasons. First, it’s because over 80% of my allocation is in Apple and its terribly difficult to escape its performance. Keep in mind that I also retain some exposure through Berkshire as well. Apple is -20% YTD which means the fact that I’m still close to the S&P is a testament to my allocations.
The second is that a lot of my investments had a loss of 30-50%. Yet on the other side, I was able to get more back from my pure value investments like DIT and SENEA. This is just a matter of pure coincidence.
There were some pitches I did not swing at. For instance, Ally Bank. While Berkshire Hathaway might have added it, I elected not to. Ally Bank mostly has subprime autoloans on its books and I did not believe Ally have substantially better risk management when financing subprime auto loans. Given my belief in the deterioration of the subprime auto market, the tide is heading out and there is nothing that tells me Ally’s underwriting and checks were disciplined enough to avoid this.
One pitched that I completely missed was Hostess Brand (TWNK), despite my nearly religious loyalty to it as a pre-workout snack. I wanted to add it to my main portfolio at $12 a share. I missed my chance and added it in my gambling portfolio at $14. I stubbornly refused to buy it and watched it 2x in 18 months. It has never looked back since. Had I taken this at my desired portfolio allocation, I would have ended this year beating the S&P by a clear 3-5%.
In short, you can be a little sloppy in your investments like I am, get hammered in some areas, and still come out better than idiots who wanted to buy into meme unprofitable growth stocks. Of course, I could have also invested in the S&P to begin with this year and done about the same. But if I did that consistently, I would not be able to brag that my 3 year streak beats the S&P handily.
Alright so why the hell is this all relevant? Because I want to show you not only can I make decent returns, I know how to pick business that, at the very least, will avoid completely disastrous results. I’m no guru and I sure as hell am not the best investor. But I can still come out quite alright when the tide goes out while everyone else loses their shirts. Most importantly, I have skin in the game when it comes to investing and judging companies.
At the very least, you can consider me better than people who threw money at Peloton, Robinhood, Tattoo Chef, and rode Tesla to the moon. Speaking of avoiding trouble…
Discipline: Checklist Evaluation
Let me call back to my checklist that I created a year ago in “It’s All Too Fast”. I want to show how simply following this checklist would have avoided a lot of the layoffs and troubles you would have seen in 2021 and 2022. Do recall that companies went on a massive hiring spree in the early days of the pandemic. And do recall the massive firing spree that companies have done this year.
Engineering
Will you become a better engineer here?
Reverse look up your manager and CEO on LinkedIn. Patterns repeat. Run away from grifters.
Business/Investment
A business gets the people they deserve. What is the biggest sticking point of the recruiter?
A business must have been profitable for a long time and ethical.
What does the world look like if this business is successful? Would you find it more valuable and meaningful?
Let’s take the top 13 layoffs that have happened this year in absolute numbers and run it against the checklist. Let’s assume that you have an opportunity to become a better engineer at every single one of these companies for the sake of simplicity, although there are some here that are obviously more so than others.
Meta: I don’t like social media: it makes people meaner and traps people in echo chambers, if Tumblr is anything to go off of. Fails the ethical and meaningful check. Some may disagree. Also, you can’t forget about those TikToks of Program Managers who do nothing all day which screams grifter to me. So does spreading focus on the “metaverse” instead of core social media tech. But let’s say this passes the test for those who are inclined.
Peloton: Not meaningful. Gyms already exist. Nobody who has a home gym uses it for long unless they’re a competitive powerlifter.
Carvana: Arguably useful. But unprofitable.
Twitter: Fails the profitable check, worse version of Meta.
Better.com: Do we really need “AI mortgage”? Mortgage origination is a competitive field: I don’t know what the additional value is here.
Byju: Unprofitable and in a competitive space (online edutech). It does have value but I don’t know how much more value it can possibly have over other things out there.
Crypto.com: 🤮
Pudutech: Useful actually: creates robots for retail. Competitive environment though. Unknown profitability. Let’s give it the benefit of the doubt.
Snap: Fails the profitable check, worse version of Twitter.
Coinbase: 🤮
Butler Hospitality: Hotel ghost kitchen. Not too much value creation here but not bad either. Unprofitable.
A soft application of this list is remarkably successful on its own. It would have weeded out almost every company except Meta and Pudutech. 2/13 isn’t bad and neither of these companies are, on the surface, completely unreasonable to work for and probably would have done some good for your engineering skills (working on a highly distributed and scaled enterprise product or robotics). Having Meta on your resume isn’t a terrible negative either and opens a lot of doors for you as well. All of these are what you want as an engineer and for your career.
Another one: “Run away from grifters.”
According to the dictionary:
Grifter: “a person who engages in petty or small-scale swindling.”
Swindle: “use deception to deprive (someone) of money or possessions.”
But if a person engages in a multi-billion dollar theft by deception, does that qualify? Well, its not small-scale for sure but in the $20T US economy, a few billion dollars might.
Which brings me to the CEOs who use their employees and shareholders as exit liquidity.
People sell shares for many different reasons: taxes, personal debts, etc. But if your CEO sells almost all of the shares, he has most likely issued a vote of no confidence in his company. After all, if the company was going to grow at above average returns, why sell the shares? He has effectively used people who are buying as his exit: he is taking money from people and giving them claims to a company he has no confidence in.
And if the CEO has no skin in the game, what kind of decisions can you expect him to make? Starts to smell a lot like grifting to me but I will let you be the judge of that.
A mass liquidation by the chief founder and shareholder is a very bad sign and should be taken as a death knell of the company.
As an exercise, let’s take the largest IPOs that occured in 2020-2021. We will ignore SPACs for now (as those are a complete fucking joke) and see what happens to them.
Let’s set our criteria to be when a CEO sells more than 25% of his maximum holdings. We use this because prior to IPO, a CEO will have a lot of shares. However, a CEO may also be paid in a significant amount of shares. Thanks to the tax code, half of those shares may have to be sold to cover the tax bill.
It is unlikely that a CEO would be paid enough shares where half of the shares represent 25% of the holdings. That would imply he holds as much stock as he gets paid on an annual basis. Only if the CEO kept doling out stock to investors or if a non-founder became CEO prior before the IPO would this extreme situation happen.
Holy shit, half of these IPO’s, the founder sold more than 25% of stock and, in a majority of the cases, founders sold nearly ALL their stock in the company. Erstwhile telling their employees to hang on for dear life!
Let’s bring on some example CEOs from this list.
First, Brian Armstrong of Coinbase. In April, he sold 75% of his entire stake, fired 18% of his workforce (25% total for the entire year), then told his employees to “quit and find a company to work at that [they] truly believe in.” If he truly believed in Coinbase, why sell so much? And should he follow his own advice?
Next, Vlad from Robinhood. He sold 99% of his stake in Robinhood (53M shares) in August of 2021 and then proceeded to praise the “diamond handed” investors who didn’t sell out of the IPO shares. Then, next year, cut 22% of the workforce.
Snap, cut 20% of the workforce and called for a “collective default’ (that is one hell of a name). Again, I have to be a little softer on Snap but unwinding 50% of your total shares over 4 years is not great.
Doordash. This one I have to at least give some credit to where he only cut 6% of his workforce.
All of these I have called out last year as well.
But there is another thing to be said about massive inside CEO sales. Anyone who has done a decade-long endeavor will tell you that parting with control of the business isn’t a decision you do lightly. Why would a CEO try to sell you on the merits of the company after selling out of his controlling stake in it? Why would a salesman try to sell you a car he decided to get rid of?
With that attitude, what kind of employees can the CEO hope to hire? Birds of a feather flock together. So a CEO who does not stand with his own business should have no expectation that the employees he hires will carry the same sentiment.
It is ridiculous, stupid, and immature to demand from others what you are not willing or able to do yourself.
On a more positive note and as a clear example of who to emulate, I want to give a large shoutout to 1 CEOs in particular: Tim Cook of Apple.
I will always be proud to have been an Apple employee with superior management with Tim at the helm. Mr. Cook has navigated the choppy waters in the past decade with masterful grace whilst having immense skin in the game, never selling a share and putting the company and shareholders first. Quiet, unassuming, and very focused on the job. He is truly an extraordinary manager. I think people should begin putting him and Jobs in the same conversation when it comes to management. He does not get the credit he truly deserves.
Sure he sells half his stock but it is most likely to cover the tax bill. But he still has never sold beyond that, retaining sufficient skin in the game and enforcing a strong culture across Apple reflecting his discipline and operation efficiency. For a billionaire, the man is extraordinarily humble and has generously pledged to give his fortune away to charity.
We all should endeavor to be like Tim Cook. It is a benefit beyond numbers and beyond words. Buffett (CEO of Berkshire), Sinegal (CEO of Costco), and many other managers of wonderful and extraordinary companies all practice the exact same traits as Cook.
Managers with skin in the game and who make the value of the business, NOT THE PRICE OF THE STOCK, their raison d’etre are rare diamonds. If you ever find yourself working for a company with a great manager like him, never underestimate the benefit. You may never get to work for another ever again.
Fraudulent Checklists
“One of the hallmarks of mania is the high rate of fraud.” -Michael Burry, Big Short (probably)
Compare my checklist to the people writing a 20-30 step “tips and tricks” list to “help” people navigate or avoid bad companies: you are prescribing a very fragile set of actions and advice to an impossible problem. It is far too difficult and time consuming to keep joining companies, react to changing conditions of leadership and business deterioration, and then join the next one. It is also very difficult to check every “tip” without weighing its importance.
In my view, anything that is too complex will become indistinguishable from bullshit. You see this in finance all the time, where people come up with magical random theories through some convoluted chain of logic that gets highly intelligent individuals with IQs above 160 into lots of trouble. LTCM, Enron, and other companies fooled themselves into stupid and sometimes fraudulent actions because of high imaginative thinking.
This will become important in a bit. So pay attention: why are these lists that people put out so chock full of “tips” that are not ironclad nor actionable?
The benefit of writing these overly complex bullshit checklists is that the creator can then call out “oh you should have watched out for rule #17, that would have helped you identify that the company was bad.” They also love to do a “post-mortem” and draw a line through failures and say “lessons were learned.”
As I’ve highlighted in my past memo, Wanna Bet, people who only do research on a topic and fail to put skin in the game or actually tie their bets to their predictions are subject to the armchair expert fallacy. They are often wrong and fool themselves into believing they are right.
So why are people doing pointless shit like this? They need to convince people to subscribe to their substack by offering inside gossip as “news.” Combine it with their highly fragile and convoluted checklist, they can paint any story they want and look like geniuses.
When nothing happens, they just dismiss their “rumors” as just “rumors”, not rules. Heads, they look like geniuses and oracles. Tails, they are just reporters. Privatized gains, socialized losses.
“If you paid attention to the rumors that I published in my last newsletter, you could have avoided the layoffs at x company. Subscribe to my substack to get the latest news!” Sound familiar?
Why not avoid the trouble altogether? If your checklist requires you to constantly update and evaluate every week, then it's not much of a safeguard. It's fragile and little better than the emotional outbursts of a spoiled child that are sometimes correct.
I submit that my checklist is superior and more effective because it is clear, difficult to mess up, and therefore disciplined.
I always liked this adage (courtesy of ImmortalFaith):
Rules create discipline.
Discipline creates consistency.
I also add:
Consistency creates habits.
Habits create results.
Checklists that are messy and complicated do not inspire discipline and therefore create inconsistent results. So what good are these checklists if they are ineffective and sufficiently indistinguishable from randomness?
It’s disgusting. These people are treating people’s lives and job losses as fodder for their useless and unactionable content. Why are you eating off of other people’s failures? To “educate”? Then, please, why could you not have forseen their layoffs to prevent all this? Why are you just inventing some new convoluted content that is inactionable? It's actually worse than useless: it's a distraction.
A broken clock is right twice a day. And hindsight does not feed families. People’s jobs were lost and their families need to eat. Hindsight isn’t even 20/20 anyway so there’s no guarantee your “lessons” are even right!
That is not to say long checklists aren’t bad. Atul Grawande argues for the medical profession to have a checklist. The same sentiment is repeated in Staff Engineer.
It's when checklists begin to become hard to follow and make you take convoluted actions and considerations that we have a problem. For instance, if a decision fails any one of a 80 point checklist causes you to instantly say no, it's fairly reasonable. However, if the checklist leads you down some weird subroutine like the US tax code forms (ie. if you said “yes” to line number 80, skip this form and go to form 13-C instead), then there is a bigger issue at hand.
To not get into the weeds, logic is naturally deductive and a checklist should aim to eliminate the nefarious, even if it means eliminating potentially good ideas.
Am I doing a convoluted checklist? I don’t think so. Could it be better? Yeah, I could add nuances and be more explicit in how I check for each of the traits.
But the rules are fairly simple, easy to follow, and generally correct across the industry. It also gives you room to make your own decisions about the world, act on what you believe in, and so on. This also means it also gives you enough rope to hang yourself with if you believe the wrong things about how the world works and what people are like. Despite that, your engineering career will most likely be intact and do better than average.
I do also confess to hypocrisy. Did I break my own rules? Yes. After all, Uber was not profitable when I joined. But as I’ve mentioned before, I judged the upside to joining and the opportunities for my career were so massive compared to the downside. This year, I was able to represent Uber at Bazelcon, a software engineering conference at Google at their special interest group (SIG) luncheon. And we haven’t had any layoffs since the 2020 COVID shutdowns. Not bad for a 29 year old.
I bet any of these influencers who keep spouting nonsense and are paid less than 6 figures would trade places with me in a heartbeat. I certainly cannot say the same even if I could be potentially paid 7 figures a year.
When done properly, you don’t need to keep up with the latest rumors to avoid terrible decisions and do quite well in your career.
Twitter: A La Infra
Twitter deserves a special section simply because of the Elon Musk takeover and firing of half the staff. I am not sure whether to put Elon in the grifter category as his remarkable track record with running companies leads me to believe in his extraordinary talents and dismiss the rule altogether. But even so, simply by following the profitability and meaningful heuristic alone, Twitter would have been avoided as a “don’t work here” place.
It’s a shame too. People like Chris Banes and Nacho Lopez who I have met are just a few of the truly talented engineers at Twitter. Look at their blogs and talks; it will make your head spin.
Nevertheless, Elon Musk has taken to firing critical engineers, causing highly skilled engineers like them to quit in sympathy. The consequences of completely decimating your infrastructure (infra) is showing its ugly face. I spouted ad nausem last year the importance of infrastructure.
Let me repeat myself: your biggest wins comes from your infra remaining stable, not product engineers. This is because infra allows your services to run reliably. You don’t need new products when your old ones work well and are highly used.
Few people understand this. Coca-Cola doesn’t need to venture into the weed business when it is consumed in almost every country in the world. Also, why would a wholesome brand such as Coke want to associate with something as controversial as drugs?
I fully understand what Elon is doing at Twitter. From a philosophical standpoint, I actually agree with him: hardcore engineering is how you create a lean company that can accomplish world changing tech efficiently. The story of how Android was created reflects this.
That said, most people will burn out by the time the project is launched. Most people only get 1 “hardcore” multi-year project in their lifetime before they realize they cannot keep doing it over and over again, not just in tech but in basketball, business, and other ventures in life, even if they sacrifice their personal lives. Few people like Musk can consistently do it and even fewer can do it with consistently successful results.
For more stories about this, I recommend Chet Hasse’s book, Androids: The Team that Built the Android Operating System (no I do not earn an affiliate commission and all book profits go to charity anyway). It came out this year and I thoroughly enjoyed it. As an Android engineer, it's important to understand where the project and code came from and how it evolved. Most importantly, the book covers the story of the people who have worked on it and their experiences.
However, the one part of Twitter’s strategy that I disagree with is firing the infrastructure engineers. I don’t believe having the major cuts at the foundation level is the right thing to do. As I’ve highlighted in past memos, these people keep the lights on. You can take away the punch bowl but if people keep dancing with the lights off and smoke detector disabled, someone’s going to get hurt.
Who handled the build train when the fragile build system fails to cut a mobile binary release? Who handles the servers when a cluster dies? What happens when a service gets changed and causes cascading errors? Content moderation? Rebooting databases? Bandwidth limits being tested?
Look, I get firing people who are working on moonshots when you are trying to reign in spending. Twitter did NOT need NFT support and that was money well wasted. I also understand the need to cut superfluous product managers when shit needs to get done. I am 100% on board with this idea to cut costs and increase productivity, especially when the expenses yield no net revenue for the company.
But the execution has been terrible and I forsee problems in the future.
It also absolutely annoys the piss out of me when people start laughing that Twitter hasn’t collapsed in 72 hours after firing half of the employees. You flew 2 planes into a building and because the building didn’t immediately collapse, you think steel beams aren’t necessary? You can stop building a tower taller to reduce the amount of materials used. But you can’t cut out the foundation to do so without destabilizing the entire thing.
Moreover, its particularly stupid for people to use the Twitter firings as a proxy for their political beliefs. To not stray away from the point of the memo, the reaction to the Twitter episode has been more or less projecting their opinion on woke culture. People view the firing and the lack of immediate system destabilization as proof that woke culture is just a money-wasting way for people to feel self-important. Other people view the firing as Elon Musk letting back conservative hate groups and arbitrarily enforcing the free speech he claims to espouse, making him a villian.
The truth is neither of these: nobody in business is stupid to use their business as a tool to engage in a culture war. They mostly just try to avoid controversy and use whatever movement is popular at the time to align and make a buck. This is purely business and a personal project on a time limit, nothing more. Twitter as a business is not your personal army and does not care what you think about the culture.
Because there is another leg to the problem as well: financing.
Keep in mind, Elon effectively did a leveraged buy out against Twitter and his Tesla shares.
Looking at the 2021 annual report as an estimation, Twitter came through adding +200M in cash on $5bn in revenue. Without overcomplicating the situation (and talking about shareholder value and whatnot), they could have theoretically added more adjusting for share buybacks and other capital allocation decisions that may have caused some shareholders some pain.
Twitter went from barely surviving to liquidity crunch in one purchase so Elon is on the clock. This happened to also be a good excuse to fire half the staff which I suspected Elon wanted to do anyway.
Therefore, what we have is Elon racing against a clock to get things done on multiple angles. On the technology front: he has to race against a crumbling infrastructure to restructure it or to stabilize it. On the finance front, he has to race against the liquidity of Twitter to make it profitable or declare bankruptcy.
I have likened projects like this to the classic “point of no return”: where admirals burn their own ships to corner their own troops and get the most out of them.
If I had to predict, I believe that Elon will accomplish what he wants on the financial side but not the technology side. Twitter will become far less stable as it undergoes the transition. The longer this goes, the worse this will get. We may start to see stability issues 6 months from now.
But I would not bet against Elon. If anyone can pull such a feat off, Elon is the man to do it. He has the cards to play, the hardcore staff to do it, and the right influencing factors in play. Hit a dog and he runs away. Corner a dog and he bites back.
Let father time do it’s thing. Speaking of which…
Father Time Takes All
I finally get to swing the first punch at all the fraudsters in the influencer space. The spectacular collapse of FTX was just the first domino.
For those who don’t know, FTX was a crypto exchange. A market maker, if you will. People could buy and sell crypto and derivatives and FTX would help facilitate the transaction. Of course, what people didn’t realize was that Alameda Research, a trading firm run by FTX’s CEO, was also trading on FTX’s exchange. Alameda was given inside information by FTX and also special trading powers that enabled it to dodge margin calls, automatic liquidations, and so on. In a sense, they were playing poker while being able to see the cards of their opponents at a table they were dealing at.
So its even more miraculous how they managed to blow it all up. The reason is complicated but suffice it to say that when you combine a 0 value creation activity like trading with extreme leverage like FTX and fraud like Alameda, doom is all but inevitable. Even if you had the cheat codes like Alameda did.
This would be ok if the bankruptcy only hurt speculators and trades on Wall Street. Except that it most likely had a wide reaching impact on regular people.
FTX also did a significant amount of brand deals with influencers who promoted FTX. The amount they paid was nearly 6 figures a month per influencer. Why? Because FTX needed customer deposits in order to provide liquidity for trades. The more customers deposited, the more business FTX could conduct and therefore, the more profits Alameda could rake in. All benefiting the CEO’s insider trading.
I have heard this stupid argument that influencers are not responsible for their audience’s loss. Because celebrities like Steph Curry promoted FTX as well and didn’t know it was a scam, neither should these influencers and shouldn’t be held to account.
Except this is a stupid and evil argument that makes these influencers look bad compared to others.
It is stupid because even if the audience is responsible for the losses, influencers rely on the trust of their audience. Blaming the audience undermines that trust and therefore, their own business. The audience is trusting in the finance influencer because the influencer portrays himself as someone with knowledge and authority about money. He is, quite literally, influencing his audience! And the audience trusts, believes, and follows the influencer, believing the influencer has their best interests at heart.
True, the audience can do whatever he wants with the information the finance influencer doles out and technically, the influencer is not legally responsible for the actions of the audience. No influencer held a gun to the audience’s head and threatened them to support FTX. But the influencer has created a situation in which their audience can hurt themselves more easily by promoting a dangerous company and product. What might that do to the influencer’s business?
It is evil because it socializes the losses and privatizes the gains of information. If the influencer is right and FTX is a great business, the influencer can wear that badge of honor and build up his credibility. If FTX is a scam, then the audience pays the price. It is even worse when you consider influencers actually got paid by FTX in cash. The audience can’t help but wonder if the money that paid the influencers came from their own deposits, facilitated by FTX.
It make influencers look bad compared to celebrities. True, these celebrities are somewhat culpable. But nobody asks LeBron James for his advice on finance nor do they expect good advice. But they sure as hell do for an influencer who labels themselves as a financial educator.
No matter which way you slice it, it's truly sickening and evil to lay this argument out. The best these influencers can do is not just own up and issue a public statement but also ask themselves if they were in their audience’s shoes, how could they make it right?
Imagine you are persuading your spouse to buy a car for the family. After doing your research, you happen upon purchase of a particular car. It looks like there are no major issues and you find there is a massive rebate for the purchase. Your spouse trusts you and buys the car. 3 days later, the car gets T-boned, the airbags fail to deploy, and sends your kids to the hospital.
Would you tell your spouse to have “done their own research” and that it is not your fault? Of course not! Your audience might not be your spouse or family (parasocial relationships and whatnot) but you should still treat them with some level of dignity and respect.
Under this exercise, I would:
1) Spend the rebate money on the hospital bill to make the kids healthy again.
2) I would then apologize to my spouse and children
3) Refrain from car purchase/recommendations for the foreseeable future until I was ready to properly vet them better.
I would want to make sure that I would personally test the car first before anyone else touches it.
Roughly this would mean
Give back the money received and give it to the audience in some way.
Apologize to the audience
Reject sponsorships until a better vetting process was implemented
Stake own capital on future sponsorships whilst they are being recommended
None of these sound exclusively familial. To some, these steps may sound extreme but I assure you it is not: irreparably damaging your reputation is probably comparable to endangering your own child. All these sound like reasonable and responsible steps that a responsible parent and steward of information and trust would want to take.
Why is this so hard for influencers to understand or consider this? I leave that to you. But I can think of a couple million reasons why these influencers are allergic to responsibility.
Tech Moving Forward
Its too easy to be cynical about the future. After all, lots of tech companies are still hiring and as I see it, the era of easy money is gone. Now its about having chops. Hard times flushes out the frauds and I don’t see us coming back to a tech mania any time in the future.
If anything, this means that only the hardcore will survive. The strong. The people who truly understand their skill and craft.
This is why I think crypto, Defi, web3, and any other endeavors that are “the future” are going to be effectively dead. Most companies are just using a wrapper around an ineffective tech that yields no business use other than being a non productive token that is traded around.
We are already starting to see this with the hype of ChatGPT. Suddenly, everyone who used to have a crypto startup is now doing an AI startup. If grifting and fraud is a bellweather of excess cash and a bull market, I don’t think we’ve seen the end of the tech run just yet.
In fact, let’s be even more direct. Blockchain is not much better than a distributed append only ledger database with a niche internal application. Engineers are receiving money for putting their technical chops to a value destroying activity that is a social ill.
If you have been following my content for the past 2-3 years, you will know this is a message I constantly preach: invest in yourself and expand your skill. I am here to do the same for my clients. I see it much harder to get jobs in 2023 at FAANG but there is always a demand for good engineers.
I plan to spend most of my time helping J and C behind the scenes get their offers and focusing on my own career. I plan to go for a Staff level promotion so I will not be dedicating as much of my time to coaching and videos.
But I promise I will come back later on this year or next year with new content. Thank you all for your support!