I heard this today from a colleague today. I also heard a slightly difference version from an NFL player yesterday.
“We need to give 110%!”
Of course, it’s impossible. There is only 100%. That’s a mathematical fact.
What we’re really saying is, “I’m giving it my all!”
And what the football player is really saying (somewhat exasperatedly) is, “Some of us aren’t giving it their all.” And that likely is true.
But we can’t really give 100%, 100% of the time. Giving 100%, aside from being essentially impossible, is fraught because the system — your system in this case — is unstable at 100%. Nowhere to go except down. You don’t need to give 100% all the time, nor should you.
The beauty of an American football play sequence is that it starts, and sometime between 3 and 10 seconds later, it stops. Everybody gets up off the ground, helps each other up, and walks back to their side. Giving it their all for a few seconds, then recovering.
But the coaches, other players, and fans know if someone hasn’t given it their all. It’s right there on the tape for all to see.
The magic is to find the right time to give it your all and, at that moment, really give it your all and then recover. Don’t miss the moment.
I believe God exists, or at least I want to, so I conduct myself as if. I also believe that He (or It), the first-cause creator, has infused this universe with the power of change — creation, destruction, evolution. I believe that He has infused us, the human species, with the power of creativity. Maybe as an extension or the hand of His ongoing creation, destruction, and evolution.
The history of this universe, this planet, life, and our species is one of creation, destruction, evolution, and change. We can argue about the specifics and methods, but we’re always advancing. To what? I don’t know. Neither do you. But advancement is the path. We’re part of that advancement. It’s our job.
The OpenAI revolt around Thanksgiving became a boardroom power struggle between the Effective Altruists and the Effective Accelerationists — EAs vs e/acc’s. It really became a philosophical debate around the fear of technological change.
“The reality is that no one at the beginning of the printing press had any real idea of the changes it would bring. No one at the beginning of the fossil fuel era had much of an idea of the changes it would bring. No one is good at predicting the longer-term or even medium-term outcomes of these radical technological changes (we can do the short term, albeit imperfectly). No one. Not you, not Eliezer, not Sam Altman, and not your next door neighbor.”
The e/acc’s won at OpenAI. I think that’s good, but I don’t really know. Nor do they, or the EAs who lost. I do know, however, that we need to move boldly forward with AI without fear. Open market. Crowdsourced. Democratic. Individuals plus the collective.
Yes, things will change. Yes, not everyone will benefit. Yes, maybe even something goes bad for a time.
But we must continue to be creationists. It’s what we’re built for.
A world where AI does stuff — a lot of stuff — is coming.
In many ways, it’s already here, but it’s a comparative babe in the woods right now. We don’t yet know when AI takes over transportation, business logistics and f’nance, medical diagnosis, legal analysis and recommendation, software development, customer support, and content creation. But it’s definitely a question of when. Not if.
Currently, we’re in the throes of coming to grips with it. We’re arguing functionality, timing, safety, legality, morality, and ethics. This will continue, as it should. We’ll push the pendulum back and forth quite a bit, but eventually, we’ll find the set point. I don’t know when. Probably more than a year and less than 50 years from now.
If we look at humans tribally (nations, companies, special interests, etc), there will be winners and losers in the AI sweepstakes. Some will benefit and others will suffer. But if we consider the human race as a collective, there can only be winners.
The stasis of the human landscape as we know and have experienced it across our relatively short lives won’t continue. It never does.
Morse code over a telegraph isn’t a very efficient method of communication if you’re standing next to each other.
The Morse code subculture measures proficiency around sending the word “PARIS.” Those five letters require 50 “dits” (the technical term for the beeps). An average proficiency for PARIS is 20 words per minute. That’s a hundred letters and a thousand dits per minute.
Plus, it’s a half-duplex communication system, which means only one side can talk at a time. Therefore, when one side is talking, the other side must listen. So that’s that’s 10 words, 500 dits, per minute for each of you.
Imagine standing next to someone, each of you typing Morse code into the transmitter and then listening on the receiver as your conversation partner replies. Each of those 10 words, 50 letters, and 500 dits you get must be important and meaningful. You must think about each one.
Speech, however, is very efficient.
Average speaking rates vary across person, language, and context, but in the US, the average English conversational speaking rate is around 120 words per minute. The average English word is five letters. That’s 600 letters per minute.
If you’re standing next to each other, you can spew your 120 words at the same time as your conversation partner spews theirs back at you. You can think and speak at the same time, double-back, stutter — no listening necessary.
SBF was found guilty of seven counts of various frauds against FTX customers and Alameda lenders.
But it didn’t have to be this way. To change the outcome, he didn’t have to make any changes to the salacious tidbits that the public ridiculed him for — crypto, creative trading practices, lavish spending on high-end real estate, tawdry relationships and almost-communal living amongst the execs, playing video games during public interviews, massive celebrity endorsements, extravagant marketing campaigns, over-the-top spending, and justified bajillions in the name of effective altruism.
None of that really mattered, nor was it illegal.
All he had to do was one simple, if not easy, thing — bring in at least one financial adult and give them a voice. One person that says,
“No, Sam, it’s not OK to lose track of people’s money. You need a better accounting system.” “No Sam, it’s not OK to back door client money from FTX to Alameda. Borrowing the FTX customer funds needs to be an above-board operation.” “No Sam, it’s not OK to take the reigns off the risk management engine. If you’re losing money, find a legitimate way to handle it, which may include copping to your investors that you’re losing money. If Alameda loses money, that sucks for you and the investors, but it’s not illegal. If FTX customer funds disappear, that’s fraud.”
Yes, one naysayer adult. One guy or girl that says, “Um, you can’t do that.”
Is it age? Is it experience? Is it worldview? Is it conduct?
It’s some combination of all these, yet hard to define objectively. The paradox of startups and organizations that make change in the world is that they are quite often (but not always) driven by young, inexperienced, naive, and maybe not professional people. Non-adults. Because adults know better.
FTX and Alameda had this in spades. SBF purposely kept the employee base very young and put many “smart,” inexperienced, and unqualified (on paper) people in all positions. Some of these people rose to the occasion. Maybe they all did. But there is no doubt, in hindsight, that having zero adults in the room was the third head of Cerberus that led to the downfall.
Nobody ever said to SBF, “No, you can’t do that.” Nor did anybody ever say, “You’re super smart, but you’re not infallible.”
Or if they did, they never said it in a way that he believed or took seriously. They basically operated as a fleet of yes-men to SBF.
Interestingly, a group of adults did actually say “WTF?” to SBF in the very early days of Alameda and prior to the launch of FTX. These were the hard-core effective altruists with whom he first got started. They, like the adults they were, looked at SBF’s operations and said, “Uh, no way. Thanks, but we’re out.” Prophetic.
Combine the lack of adult supervision with SBF’s arrogance, and it led to his inability to see the real-world problems that were coming. Once they did and buried FTX and Alameda up to the neck, they had no ability to stop the avalanche.
Sometimes, you need an adult in the room. Even if they’re the naysayer, the finger-pointer, or the “that’s how they get ya” guy over in the corner.
Crypto as a concept isn’t the reason that FTX failed and SBF went to jail, but crypto as a volatile asset class is definitely one head of Cerberus.
The first head was that FTX and Alameda operated as a run-of-the-mill Ponzi scheme. That was illegal, and ultimately why SBF is in jail, but it would have all worked out if not for the volatility of the crypto market.
Starting in about January of 2022, the crypto market took a steep dive. By the fall of 2022, Bitcoin was worth less than half what it was at the beginning of the year. The entire crypto market followed roughly the same line: ~$2.75T => ~$1T.
Alameda was a hedge fund that traded in crypto. FTX was an exchange that allowed consumers to buy and sell crypto. Both very exposed to the Crypto Winter. The traders inside Alameda feverishly tried to minimize losses and find gains on the short side, eventually needing more funds than existed without “borrowing” the FTX customer money. Yes, they just took the money out of the FTX customer accounts to use for trading in the hedge fund. Consumers with accounts in FTX overreacted to negative news and swings and exacerbated the already tenuous situation, eventually trying to cash out their accounts. That created a gap in what FTX should have had in its customer funds versus what it actually had.
Crypto as a concept, although I’ve stated not the reason for the collapse, did play a minor role. One difficult, confusing, and sometimes shady aspect of the crypto world is the abundance of bullshit coins and tokens. Coins and tokens are not exactly the same thing, but for our purposes here, we’ll just use tokens to refer to both. We all know Bitcoin, you might have heard of Ethereum, and if you’re crypto-curious, you probably know Solana. However, thousands of (mostly bullshit) tokens exist to be traded, much like penny stocks. Anybody with a computer and a dream can create their own blockchain or create an application on an existing blockchain to make any new token for any purpose. It’s basically that simple.
FTX used this approach and created its own tokens, named FTT. Think of an FTT token just like a share of stock in FTX. The token’s value was tied directly to the value of the company. Therefore, a large portion of FTX’s valuation was tied to the value of its own FTT tokens.
Cue the asset class drop, the FTX – Alameda Ponzi with frazzled traders, the frightened consumers trying to pull their money out of the exchange, and the drop in value of the FTT token. Roll that all together, and you get an exponentially dropping valuation for the company. Eventually, in November, the balance sheet flips over. FTX is insolvent. They file for bankruptcy.
Next, we’ll tackle how the people themselves became the third head of Cerberus.
Although two separate companies, they didn’t keep a robust enough financial separation between the two. In fact, they didn’t separate them, which ultimately led to the collapse.
If you opened an account on FTX and dropped some money into it, your dollars were physically held inside of Alameda rather than FTX. The old shell game. They did this because when FTX first opened, no bank would supply it with a bank account. Therefore, they found a pragmatic solution in using Alameda’s bank account. That, in and of itself, is fishy, but maybe not illegal (depends on who you ask) because if you do the accounting correctly and keep those funds separate, it shouldn’t be a big deal.
But they didn’t do that.
When Alameda needed extra funds, they just used the money (“borrowed”) in the FTX user accounts. That’s illegal. That’s a Ponzi scheme.
They could’ve gotten away with it without the other two problems: the volatile asset class and no adults in the room.
Crypto’s a scam! Crypto’s full of crooks! Crypto’s not real!
You might think that Sam Bankman-Fried’s jailing and the FTX exchange’s collapse were inevitable because, you know, crypto. Crypto’s a scam. It’s full of crooks. It’s not real. Of course, it collapsed.
But that wouldn’t be exactly correct. Crypto was a minor role player — more like the drug pusher standing outside the backstage entrance greasing the party wheels. A problem-enabler, but not the problem.
The collapse itself was due to plain ole, run-of-the-mill f’nancial reindeer games. Namely, FTX-Alameda Research was a Ponzi scheme wrapped around a highly volatile asset class run by a bunch of people (or at least one person) who overestimated their abilities and underestimated the need for adults in the room.
First, let’s tackle why FTX and Alameda Research were a Ponzi scheme.
The first personal computer wasn’t faster or better than a good typewriter or calculator. But it had potential. The first horseless carriage wasn’t faster or better than a good horse. But it had potential. The first airplane wasn’t faster or better than a good train. But it had potential.
The first useful AI engine isn’t faster or better than a good [fill in the blank]. I’m not entirely sure what that blank is, but it has potential.
