The post is mainly just a CTA against further internet centralization and government control of core infrastructure, which is fine. We need more of these, and we need more examples of their harms for folks to draw on. HN often gets distilled down to a singular cause - EU's Chat Control, Elon's shutdown of Starlink over Ukraine, a regional outage of a public cloud provider - but generalized topics like these aren't really discussed all too often I find, or are often flagged for a variety of reasons and shutdown.
As technologists of multiple stripes and disciplines - programmers, developers, engineers, architects, designers, product managers, etcetera - we need to collaborate more on the direction of our industry as a whole, rather than just specific niches we find appealing. From my specific perspective in IT, the increasing centralization across every vendor category (three major x86 server manufacturers, two CPU vendors, two GPU makers, three global-scale public clouds, ISP mono- and duopolies, a handful of commercial operating systems, a near-monopoly EUVL supplier - the list goes on) is a dire threat to not just the open internet, but open technology in general.
We need to be better advocates for and champions of the technological future we envision, rather than just blindly celebrate startups and tech fads all the time. Mr. Schneier is merely the latest and largest canary in the proverbial coal mine.
"In February 2022, two days after Russia's full-scale invasion, Ukraine requested that the American aerospace company SpaceX activate their Starlink satellite internet service in the country, to replace internet and communication networks degraded or destroyed during the war.[2][3][4] Starlink has since been used by Ukrainian civilians, government and military.[3][5] The satellite service has been employed for humanitarian purposes as well as defense and counterattacks on Russian positions.[6]"
"In 2022, Elon Musk denied a Ukrainian request to extend Starlink's coverage up to Russian-occupied Crimea during a counterattack on a Crimean port, from which Russia had been launching attacks against Ukrainian civilians; doing so would have violated US sanctions on Russia.[18] This event was widely reported in 2023, erroneously characterizing it as Musk "turning off" Starlink coverage in Crimea.[19][20]"
It’s hard to be a better advocate without diving into the politics of why we’re in the situation we are, which also doesn’t address the amount of political power you and I have relative to the interests that want said technological consolidation to exist.
And given that the tech community trends towards political philosophies like libertarianism, which is inherently anti-organization and anti-collectivist, I’m not sure how you begin to scratch the surface of what a real solution looks like.
Politics are a factor but economics is a bigger one. With any technology, each successive generation inevitably requires larger and larger capital investments. Ideally governments should do more to preserve competition but when it costs >$10B to develop a new microchip manufacturing process that inherently limits how many players the market can support. And if one company bets on the wrong technology or gets the timing wrong that can leave them too financially weak to survive.
> when it costs >$10B to develop a new microchip manufacturing process that inherently limits how many players the market can support.
Does it though? TSMC's market cap is over a trillion dollars. Likewise Nvidia. What's $10B compared to these numbers? Less than 1%. Maybe we couldn't have a thousand of them, but why couldn't we have ten?
Not only that, this technology isn't a single invention, so why does it have to be a single company? Couldn't some companies make the fabs and other ones operate them, causing them each to require less capital and be easier to compete with on its own? Couldn't the various pieces of equipment in the fabs each be developed by a separate company?
"It costs >$10B to do this as a vertically integrated conglomerate" is bad, so maybe don't have that.
This piece could be infinitely long trying to address every single angle that is relevant, big or small. Or it could just cut to the heart of the matter and ask us all to fill in the rest. I’m fine with the latter, personally, as the “why” is not really what they’re debating. Whatever the cause(s), the end result is currently undesirable and necessitates action. We can unpack the “why” as we try to fix it.
Do unions work against corporate mergers? I’d imagine they do as they tend to work against corporate interests in general but I’m not that well versed in this sort of history.
While it may not be practical from a technical perspective, the current US president has suggested shutting down parts of the Internet to ostensibly combat terrorist recruiting.
I thought this would be advocating "chaos monkey" style intentional shutdown to test institutions for resiliency in an outage situation. Might not be a bad idea. Maybe once every four years on leap day or something.
would it be better to start the intentional shutdown at say a couple of minutes before midnight so you know the shutdown wasn't perhaps caused by the leap day bug?
multiple satellite operators are coming on line. what are the odds all of them coordinate to shut down in one region invalidating using the other providers as fail over?
Its become clear that the axiom “The Net Interprets Censorship As Damage and Routes Around It” as no longer true. It hasnt been since before 2010 anecdotely but the data Schneier presents here is undeniable
This is still somewhat true. For example, Russia is now frequently shutting down mobile Internet. Ostensibly for protection against drone attacks, but even it had to relent a bit and allow at least some whitelisted services to work.
So immediately local VPN companies started providing the unrestricted access through proxies at these services.
Worship of the eternal steady-state. Whoever speaks against any intervention to preserve it is a heretic, and must be excommunicated.
Whether it’s ML training, pentesting, or old-fashioned engineering, we have to throw the occasional curve-ball at our systems in order to improve them. Surprise internet shutdowns are good, even if the ostensible reasons for them are dumb. Maybe people will host more information offline, and become less dependent on cloud services…
One more reason to resist the fragile lifestyle that requires constant internet access. Even if you don't live in a totalitarian country where shutting down the net would be easy and probable.
Some time ago someone posted in Twitter a letter of Theodore Kaczynski giving life advice, one point being not to use internet for more than one hour a day. Too bad I couldn't find it anymore.
why is this flagged? (maybe Theo? I don’t know this person).
Its absolutely a good argument against fragile IoT devices that have no local/offline mode and the ever increasing lurch of internet requirements for our daily life.
I’m not sure my phone does much of anything without an internet connection. Yet it is my primary banking and authentication method (via BankID).
EDIT: Theodore Kaczynski is the unabomber… well, thats an odd name to drop and maybe not an ideal candidate for life advice.
It's getting downvoted because (1) this person is suggesting the answer to governments taking away our ability to freely communicate is to stop freely communicating (2) he's giving life advice from a terrorist mass murderer.
Yes, you're not at risk from being cut off from the world if you're not connected to it in the first place. That's not a state most of us want to exist in. Ted Kaczynski lived in a small cabin in the woods away from humanity.
As technologists of multiple stripes and disciplines - programmers, developers, engineers, architects, designers, product managers, etcetera - we need to collaborate more on the direction of our industry as a whole, rather than just specific niches we find appealing. From my specific perspective in IT, the increasing centralization across every vendor category (three major x86 server manufacturers, two CPU vendors, two GPU makers, three global-scale public clouds, ISP mono- and duopolies, a handful of commercial operating systems, a near-monopoly EUVL supplier - the list goes on) is a dire threat to not just the open internet, but open technology in general.
We need to be better advocates for and champions of the technological future we envision, rather than just blindly celebrate startups and tech fads all the time. Mr. Schneier is merely the latest and largest canary in the proverbial coal mine.
"In February 2022, two days after Russia's full-scale invasion, Ukraine requested that the American aerospace company SpaceX activate their Starlink satellite internet service in the country, to replace internet and communication networks degraded or destroyed during the war.[2][3][4] Starlink has since been used by Ukrainian civilians, government and military.[3][5] The satellite service has been employed for humanitarian purposes as well as defense and counterattacks on Russian positions.[6]"
"In 2022, Elon Musk denied a Ukrainian request to extend Starlink's coverage up to Russian-occupied Crimea during a counterattack on a Crimean port, from which Russia had been launching attacks against Ukrainian civilians; doing so would have violated US sanctions on Russia.[18] This event was widely reported in 2023, erroneously characterizing it as Musk "turning off" Starlink coverage in Crimea.[19][20]"
https://en.wikipedia.org/wiki/Starlink_in_the_Russian-Ukrain...
And given that the tech community trends towards political philosophies like libertarianism, which is inherently anti-organization and anti-collectivist, I’m not sure how you begin to scratch the surface of what a real solution looks like.
Does it though? TSMC's market cap is over a trillion dollars. Likewise Nvidia. What's $10B compared to these numbers? Less than 1%. Maybe we couldn't have a thousand of them, but why couldn't we have ten?
Not only that, this technology isn't a single invention, so why does it have to be a single company? Couldn't some companies make the fabs and other ones operate them, causing them each to require less capital and be easier to compete with on its own? Couldn't the various pieces of equipment in the fabs each be developed by a separate company?
"It costs >$10B to do this as a vertically integrated conglomerate" is bad, so maybe don't have that.
...or Ukraine's censorship of russian internet for its citizens starting 2017
https://time.com/4150891/republican-debate-donald-trump-inte...
Advantage: You no longer need to fix that leap day bug on your website.
Covering Adelaide, South Australia. Such communities should exist in most cities.
Of course, once jamming enters the picture, even that lifeline disappears.
Note that one of the higher-profile deliberate internet shutdowns was Starlink itself shutting down internet connectivity in Ukraine.
Neither option is risk-free; the failure modes simply differ.
A government can shut you off for political reasons, a corporation can shut you off for contractual or geopolitical ones.
As long as the system assumes centralized stewardship for safety or reliability, someone will inevitably hold the switch — the only variable is who.
multiple satellite operators are coming on line. what are the odds all of them coordinate to shut down in one region invalidating using the other providers as fail over?
So immediately local VPN companies started providing the unrestricted access through proxies at these services.
Whether it’s ML training, pentesting, or old-fashioned engineering, we have to throw the occasional curve-ball at our systems in order to improve them. Surprise internet shutdowns are good, even if the ostensible reasons for them are dumb. Maybe people will host more information offline, and become less dependent on cloud services…
I'll correct that to: Surprise internet outages are good
For the same outcomes though. More and varied methods of contingency.
Some time ago someone posted in Twitter a letter of Theodore Kaczynski giving life advice, one point being not to use internet for more than one hour a day. Too bad I couldn't find it anymore.
Its absolutely a good argument against fragile IoT devices that have no local/offline mode and the ever increasing lurch of internet requirements for our daily life.
I’m not sure my phone does much of anything without an internet connection. Yet it is my primary banking and authentication method (via BankID).
EDIT: Theodore Kaczynski is the unabomber… well, thats an odd name to drop and maybe not an ideal candidate for life advice.
Yes, you're not at risk from being cut off from the world if you're not connected to it in the first place. That's not a state most of us want to exist in. Ted Kaczynski lived in a small cabin in the woods away from humanity.
(there are actual web hosting companies in Kabul, and it seems its not illegal to send money there)