It highlights the dangers of a mono culture as well.
Most importantly though, I remember my history professor who always said, "follow the money!" Do the CEO and the board have to pay (either in money out of pocket or time served in prison) for these problems (outages / oil spills / global warming / what not)? If not, what incentive do they have to make the likelihood of these problems dial down to zero?
The company does have to use its assets to pay anything it's liable for, yes. It sounds like essentially arguing against the concept of limited liability for the owners and investors? I would regard LL as generally a good thing. Without it, it would be a much bigger ask for anyone to start up a business when they're putting their own house and everything at risk. And I expect it would lead to an even further skew of already-rich people dominating investment and company ownership
> start up a business when they're putting their own house and everything at risk
Every hardware startup founder I've been involved with has used their house and/or their retirement savings to back their operating loan.
I think the advantages of incorporation have been well established by now, and believe that partly removing limited liability would not significantly effect incorporation rates.
Certain people advocate a compromise where a company may be allowed to be limitedly liable until a certain size then it has to transform to unlimited liability.
> It highlights the dangers of a mono culture as well.
The devil is in the detail though. You can have an incredibly stable distributed system with limited failure modes.
lots of examples around, a pretty well known one would be VISA transactions. Each acquirer can fail separately, and even if they're failing, some of their terminals will keep going without any impact.
And the remaining acquirer will also keep chugging along.
It's still objectively a monoculture, as Amex and MasterCard are pretty rare on global scale, while visa is basically everywhere
> It's still objectively a monoculture, as Amex and MasterCard are pretty rare on global scale, while visa is basically everywhere
I can't think of any part of the world I know of where only Visa cards are accepted in 2024 and not MasterCard. AMEX requires a separate contract agreement for the merchant hence why it is not bundled by default, but all acquirers offer Visa/MasterCard as one package. Visa and Mastercard have been a duopoly for far too long to differentiate between the two at the consumer level, and it just doesn't make any financial sense for a financial institution to miss out on interchange and card fees by accepting only one card type and not the other of the two.
Yes I agree with you, the article and Taleb. These corporations which the whole West depends on sound nice in theory but cause the system as a whole to be brittle. I think there is definitely a market for writing software for very specific areas. Like, I donno, an office productivity suite hyper optimized for the European market. It sounds impossible but if you obsess over quality then I think it’s possible to get (very happy!) customers.
Doing the right thing isn't always aligned with winning.
Shareholders can sue a company for not doing whats in their best interest. Instruments such as public benefit corporations are relatively new. Employee owned corporations aren't very common - almost non-existent in tech / biotech /pharma / energy / transportation.
Customers want cheap services, and a corporation that would employ three times as many QA people would be disadvantaged against the ones that just wing it.
It will take several incidents such as this, before the market starts understanding that some extra cost is tolerable for having some sort of warranty.
Martin Quinn, campaign director for the PCA, said using cash allowed for anonymity. “I don’t want my data sold on, and I don’t want banks, credit card companies and even online retailers to know every facet of my life,” he said. Budgeting by using cash is also easier for some, he added.
Cash is also the only payment system I can think of which is truly “private”. I think this is important too.
You're probably right, but it doesn't always feel that way when you have serial numbers on every piece of currency. And coins are so uselessly low value today that they're hard to use.
the lack of privacy at a business isn't a problem baked into cash. cash helps solve that problem because I don't need to go to a place of business to use it. It can be exchanged anywhere in reasonable amounts, and I don't have to deposit it I can just keep it and use it.
my wife and I always keep some cash on hand in case of some sort of problem with the card, or tap device etc. never had my cash declined because a computer thought some other purchases that afternoon looked suspicious.
Cashless societies are dangerous purely for reasons of privacy and freedom. The ability to trade without government oversight matters. I am appalled to see government run services, hospitals, and stores near me start to refuse cash. That shouldn’t be legal.
Not to mention, cashless systems like Visa and Mastercard and PayPal are subject to private overlords that may scrutinize or ban your transactions.
Bridges collapse; police go corrupt; electricity has outages; gasoline suffers price shocks; governments have a 100% failure rate; an 1870s-style solar flare would send us back to medieval times.
All we can do is:
A. Diversify, under the belief nothing is infallible; and even perfect engineering can experience catastrophic failure
B. Enjoy what we have today, knowing it is a gift, not a right
Also, while we blame CrowdStrike, let’s not forget that SSH was within weeks of being backdoored on a global scale. The power that would have unleashed (and proxy power, I.e. breaking into CloudStrike and then Windows by extension) very well could have ended the internet.
> All infrastructure is fragile and always will be.
No, and it doesn't need to be. We can build redundancy on every level. We can plan for failure and prepare for contingencies. We alwas need a plan B (and C and D for critical services). Will it be efficient or cheap? Of course not - we'd be building two for the price of two, but being sure that, when one fails, we have a redundant one to catch the fall of civilization.
We can't just optimize away every gram of safety and hope everything goes to plan (as they did in this case). Hope is not a strategy.
I wouldn't say fragile. A bit of Roman stuff is still functional today. The pyramids are holding up too. Cathedrals also have a massive stability streak. The stuff that has been falling over is mostly brick, steel, and concrete. And that's due to lack of required maintenance. I could even use the opposite word to describe them: strong and solid.
The things do not fallback but rather propagate causing an avalanche. The only alternative to electronic payments which always boil down to some form of VISA or Mastercard card, is cash.
Cash has to be dispensed first, and ATM's run on Windows, and so do banking terminals at bank branches.
I do not have the information about what happened to the ATM's during the meltdown on Friday, but the bank branches where CrowdStrike had been deployed were rendered inoperable.
You withdraw cash every single time you need it?! The point of cash is that you don't ping the systems on every transaction and you decide how frequently it's withdrawn.
I found this page [1] interesting in regard to cash supply in Sweden, as it is a country always reported with a high affinity to online payment and statistics [2] (from 2019) seems to support that statement.
Hear, hear! There are no outages, hacks, etc... when using cash. I use credit cards as much as anyone else but also use money at every opportunity. Shops need to use collect on delivery as a form of payment too.
Most importantly though, I remember my history professor who always said, "follow the money!" Do the CEO and the board have to pay (either in money out of pocket or time served in prison) for these problems (outages / oil spills / global warming / what not)? If not, what incentive do they have to make the likelihood of these problems dial down to zero?
I like this article
https://archive.ph/1ekmk
Every hardware startup founder I've been involved with has used their house and/or their retirement savings to back their operating loan.
I think the advantages of incorporation have been well established by now, and believe that partly removing limited liability would not significantly effect incorporation rates.
The devil is in the detail though. You can have an incredibly stable distributed system with limited failure modes.
lots of examples around, a pretty well known one would be VISA transactions. Each acquirer can fail separately, and even if they're failing, some of their terminals will keep going without any impact. And the remaining acquirer will also keep chugging along.
It's still objectively a monoculture, as Amex and MasterCard are pretty rare on global scale, while visa is basically everywhere
I can't think of any part of the world I know of where only Visa cards are accepted in 2024 and not MasterCard. AMEX requires a separate contract agreement for the merchant hence why it is not bundled by default, but all acquirers offer Visa/MasterCard as one package. Visa and Mastercard have been a duopoly for far too long to differentiate between the two at the consumer level, and it just doesn't make any financial sense for a financial institution to miss out on interchange and card fees by accepting only one card type and not the other of the two.
Doing the right thing isn't always aligned with winning. Shareholders can sue a company for not doing whats in their best interest. Instruments such as public benefit corporations are relatively new. Employee owned corporations aren't very common - almost non-existent in tech / biotech /pharma / energy / transportation.
Customers want cheap services, and a corporation that would employ three times as many QA people would be disadvantaged against the ones that just wing it.
It will take several incidents such as this, before the market starts understanding that some extra cost is tolerable for having some sort of warranty.
Martin Quinn, campaign director for the PCA, said using cash allowed for anonymity. “I don’t want my data sold on, and I don’t want banks, credit card companies and even online retailers to know every facet of my life,” he said. Budgeting by using cash is also easier for some, he added.
Cash is also the only payment system I can think of which is truly “private”. I think this is important too.
my wife and I always keep some cash on hand in case of some sort of problem with the card, or tap device etc. never had my cash declined because a computer thought some other purchases that afternoon looked suspicious.
Does some central authority have access to and the compute power to correlate all of the data available in those video feeds? No
Not to mention, cashless systems like Visa and Mastercard and PayPal are subject to private overlords that may scrutinize or ban your transactions.
Bridges collapse; police go corrupt; electricity has outages; gasoline suffers price shocks; governments have a 100% failure rate; an 1870s-style solar flare would send us back to medieval times.
All we can do is:
A. Diversify, under the belief nothing is infallible; and even perfect engineering can experience catastrophic failure
B. Enjoy what we have today, knowing it is a gift, not a right
Also, while we blame CrowdStrike, let’s not forget that SSH was within weeks of being backdoored on a global scale. The power that would have unleashed (and proxy power, I.e. breaking into CloudStrike and then Windows by extension) very well could have ended the internet.
No, and it doesn't need to be. We can build redundancy on every level. We can plan for failure and prepare for contingencies. We alwas need a plan B (and C and D for critical services). Will it be efficient or cheap? Of course not - we'd be building two for the price of two, but being sure that, when one fails, we have a redundant one to catch the fall of civilization.
We can't just optimize away every gram of safety and hope everything goes to plan (as they did in this case). Hope is not a strategy.
"an 1870s-style solar flare would send us back to medieval times"
This is not true and it would have no effect on microelectronics.
Hence keeping cash alive has it’s advantages
2020s UK electricity infrastructure is more reliable than 1980s UK electricity infrastructure, for example.
I do not have the information about what happened to the ATM's during the meltdown on Friday, but the bank branches where CrowdStrike had been deployed were rendered inoperable.
So, no, cash is not going to help.
[1] https://www.riksbank.se/en-gb/payments--cash/payments-in-swe...
[2] https://www.statista.com/chart/17307/paying-with-cash-europe...
Perhaps if we capped fees and made accepting cards (or any type of electronic payment) more of a utility then it wouldn't be so cumbersome and costly.