That’s Not Going to Be Easy to Do
Apr. 29th, 2022 08:13 amAs April winds down, and we move from calendar jokes about weed (4/20) to calendar jokes about Star Wars (May the 4th Be With You), in this, Our Age of Nerditude (I tried a lot of other options before settling on that one, but if you have a better one, I’ll entertain it if I’m entertained by it), news articles are starting to grapple with elements of reality that Just Ignore It and It Will Go Away did not already get rid of.
https://www.cnn.com/2022/04/28/tech/elon-musk-authenticate-all-real-humans/index.html
Here, Brian Fung at CNN engages with what Elon Musk might mean by Authenticate All Real Humans, and also mentions Musk’s focus on spam cryptocurrency tweets and accounts. This is highly unusual! Most coverage I’ve seen so far in regular media on the topic of The Acquisition has fallen into a deep hole of fear surrounding “What Does Free Speech Mean” and “Will Orange Guy Return to Twitter”, which as near as I can tell is the closest thing to zombie movie in real life ever, at least for people sufficiently distant from Russia (if you are not sufficiently distant from Russia, obviously that’s the closest thing ever to zombie movie in real life).
“Musk's drive to "authenticate" Twitter users stems from one of his biggest pet peeves with the platform: spam accounts, particularly those that push cryptocurrency scams. It's often not hard to find these accounts lurking in the replies to Musk's tweets; many even attempt to trade on his celebrity and lure the unsuspecting by impersonating him.”
This is _entirely accurate and relevant_! Thank you Fung! Thank you CNN!!! (<— Sincere!!!)
OK, now to the criticize it part.
“Musk's verified account was affected by a widespread Twitter hack that led to users including former President Barack Obama and Kanye West unwittingly spreading a bitcoin scam.”
NOT accurate! “Users” did not unwittingly spread a bitcoin scam. _Their accounts on twitter did that when not under their control_. This distinction matters. If Fung — who so far is just about the most reality based analysis of The Acquisition ever — can’t distinguish between an account and its user, and attributes actions taken in that account by a hacker to its legitimate user, _even years after the hack is documented and publicized_, how can we possibly expect anyone else to make that distinction, much less _before_ documentation and publicity? Fung getting this wrong severely undermines criticism of Musk that suggests Musk is overly concerned about this problem.
OK, fine, whatever, and we’ll step over the snark in the next paragraph, and move on to Fung’s speculation about what Musk’s “Authenticate All Humans” might mean. He offers three possible options:
“One possibility is an expansion of Twitter's existing verification program. Currently, to receive a blue check on their accounts, users have to supply a link to an official website that they're affiliated with, an official email address or a government-issued form of identification. Musk could stop short of requiring identification but require that users use their real names.”
Currently, in order to use the existing verification program, you have to be “notable”, and Twitter describes what that means.
https://help.twitter.com/en/managing-your-account/about-twitter-verified-accounts
As a practical matter, if you are not famous, working for a famous something, or something kinda adjacent to that, you cannot get a blue check, no matter how well behaved and/or documented you are. As far as I can tell, the blue check is the only current twitter verification program (they have had others in the past, I think). One possible thing that Musk could do is create a new category of verification that has all the requirements except the famous / famous adjacent part (that is, you would have to supply all the authenticating information, and behave well, but other than that, you would not have to be of particular Public Interest). Fung does not entertain this possibility. Fung _does_ entertain the Real Name possibility, which is a rhetorical device known as a Straw Man. No one is going to do that. Facebook already did that.
Fung entertains using CAPTCHa, or credit card to determine that someone is a Real Human, again, these are rhetorical devices known as Straw Men. These do not establish “real humanity”, and everyone is super clear on that. However, Fung instead steps from one Straw Man to another, altho, to be fair, this segues to a source, York. Maybe someone doesn’t have a credit card. Maybe they are worried about credit card fraud, etc. Bunch of nonsense. No one at scale uses credit cards to establish “real humanity”. (<— If you argue, I’m going to point to “at scale” and “real humanity”, so come prepared for that.)
Next up! If Twitter collects enough personal information of any sort to establish “real humanity”, then they’d be a target for all kinds of shenanigans by criminals and repressive governments. _Without engaging with who might do that_, I would point out that it should be fairly straightforward to collect the information for the purpose of verifying an account, and then _purge the information completely_. You don’t have to save it, and you probably shouldn’t (for account NOT of public interest — accounts of public interest are a different matter, but Twitter seems to have a solid solution there). But if you decided _to_ save it, there are tokenization methods to save it in a disconnected, silo’d manner, and only permit access by a very limited number of people in a highly controlled way. Companies that lose their credit card data usually have not set up these kinds of controls (for a long time, many companies were using credit card numbers — untokenized! — as customer identifiers, which is beyond insanity).
I wasn’t going to post a long thing about this, because while it’s great that Fung has engaged with the core issues here — that Musk has a stated goal (authenticate real humans) to solve a stated problem (cryptocurrency spam bots who are stealing his likeness) — the _way_ that Fung has engaged with that core issue falls short of typing this much.
But I decided to post a long thing about this because of how it wraps up.
“After years of trial and error, tech platforms have already developed important lessons about user authentication that could benefit Musk, said York.
"If he merely means things like CAPTCHAs, I think he's in for a surprise," said York. "He's talked a lot about how he'll get rid of bots, but Twitter's been trying to do that for years and I think he'll soon realize it's not an easy problem to solve."”
It’s hard to know really what to do with this. I mean, do I snark on EFF? Do I snark on York? Do I snark on Fung? Do I look at it and point out how all statements of the form “If FALSE then” are all by definition true? Because “he merely means things like CAPTCHAs” is _obviously false_. Duh. No, he does not mean that. Come on. Rockets landing on their butts. Starting a new car company in the United States. Judge signed off on the SolarCity acquisition. Harassed Al-Rumayyan in email and is still walking around quite fine and with even more money now than before. Tweeted about how he wanted to go mano a whatever with Putin as a way to resolve WTF and no one even bothered to mention that again. Did not had to pay up for calling someone a pedo who clearly was not and who went to the bother to take him to court over it.
I am pretty sure this is a person who knows what is a hard problem and also how to solve hard problems. He’s not thinking it is an easy problem.
I mean, unless he _is_ thinking it is an easy problem, in which case a sensible person with any access to wikipedia and the ability to read should sit back and wait and see how it turns out because if it _is_ an easy problem for him, then he should damn well solve it so twitter works better for everyone.
ETA:
Sometimes, I’m struggling with something, and my husband shows up and tries to take it over. This is terrible! He should not do this. He does it to the kids, too, and he _really_ should not do it to them, either. It is terrible parenting. It induces dependency. It is disruptive of process. Etc. I’ve done a lot of screaming about this (with cogent arguments — I don’t just verbally abuse people, I raise the volume and stay on topic. I’m not new to this game). Still, he’ll show up and try to take over a task.
Lately, if I don’t have specific concerns about him ruining something that cannot be corrected, I’ve been letting him. I do this now because I slowly realized a couple of things. First off, if he can do it, and he wants to do it, sure, fine, whatever. Go ahead. I have plenty of things to do, and if I run out, Lego makes more things for me to do. Second, if he thinks he can do it, and I really do know he can’t but also that he’s unlikely to break anything irretrievably, it’s far more fun to have that as another example of the list of Remember That Time When You (I have a very good memory. And a blog. And other documentation of my life) than it is to scream at him. AND — and this is golden, seriously, keep it in mind — the amount of conflict in my life is drastically reduced. Partly because I’m not screaming, which is important because I hate screaming, but mostly because it turns out that even a short list of Remember That Time When You is a profound deterrent. Honestly, I can evoke most of it with two or three words in a particular tone of voice. He’s even started _waiting_ (arguably hovering, but I’ll take it) until asked before reaching for something and just taking it over.
Musk is a little feral still. But if he’s really got a trick that’ll get the stuck lid off the pickle jar, I say let him give it his best shot. Twitter is not in a good place as a public company, and there is no clear path forward for it to get to a good place as a public company.
Think of it as Bezos buying the WaPo. When El Jefe bought the paper, there was a lot of wringing of hands and fear. But now, it’s a really nice news source, very well run, even the comments are often worth a glance. And honestly, the odds of Musk being worse for our country than the Murdochs have been seem kinda low.
https://www.cnn.com/2022/04/28/tech/elon-musk-authenticate-all-real-humans/index.html
Here, Brian Fung at CNN engages with what Elon Musk might mean by Authenticate All Real Humans, and also mentions Musk’s focus on spam cryptocurrency tweets and accounts. This is highly unusual! Most coverage I’ve seen so far in regular media on the topic of The Acquisition has fallen into a deep hole of fear surrounding “What Does Free Speech Mean” and “Will Orange Guy Return to Twitter”, which as near as I can tell is the closest thing to zombie movie in real life ever, at least for people sufficiently distant from Russia (if you are not sufficiently distant from Russia, obviously that’s the closest thing ever to zombie movie in real life).
“Musk's drive to "authenticate" Twitter users stems from one of his biggest pet peeves with the platform: spam accounts, particularly those that push cryptocurrency scams. It's often not hard to find these accounts lurking in the replies to Musk's tweets; many even attempt to trade on his celebrity and lure the unsuspecting by impersonating him.”
This is _entirely accurate and relevant_! Thank you Fung! Thank you CNN!!! (<— Sincere!!!)
OK, now to the criticize it part.
“Musk's verified account was affected by a widespread Twitter hack that led to users including former President Barack Obama and Kanye West unwittingly spreading a bitcoin scam.”
NOT accurate! “Users” did not unwittingly spread a bitcoin scam. _Their accounts on twitter did that when not under their control_. This distinction matters. If Fung — who so far is just about the most reality based analysis of The Acquisition ever — can’t distinguish between an account and its user, and attributes actions taken in that account by a hacker to its legitimate user, _even years after the hack is documented and publicized_, how can we possibly expect anyone else to make that distinction, much less _before_ documentation and publicity? Fung getting this wrong severely undermines criticism of Musk that suggests Musk is overly concerned about this problem.
OK, fine, whatever, and we’ll step over the snark in the next paragraph, and move on to Fung’s speculation about what Musk’s “Authenticate All Humans” might mean. He offers three possible options:
“One possibility is an expansion of Twitter's existing verification program. Currently, to receive a blue check on their accounts, users have to supply a link to an official website that they're affiliated with, an official email address or a government-issued form of identification. Musk could stop short of requiring identification but require that users use their real names.”
Currently, in order to use the existing verification program, you have to be “notable”, and Twitter describes what that means.
https://help.twitter.com/en/managing-your-account/about-twitter-verified-accounts
As a practical matter, if you are not famous, working for a famous something, or something kinda adjacent to that, you cannot get a blue check, no matter how well behaved and/or documented you are. As far as I can tell, the blue check is the only current twitter verification program (they have had others in the past, I think). One possible thing that Musk could do is create a new category of verification that has all the requirements except the famous / famous adjacent part (that is, you would have to supply all the authenticating information, and behave well, but other than that, you would not have to be of particular Public Interest). Fung does not entertain this possibility. Fung _does_ entertain the Real Name possibility, which is a rhetorical device known as a Straw Man. No one is going to do that. Facebook already did that.
Fung entertains using CAPTCHa, or credit card to determine that someone is a Real Human, again, these are rhetorical devices known as Straw Men. These do not establish “real humanity”, and everyone is super clear on that. However, Fung instead steps from one Straw Man to another, altho, to be fair, this segues to a source, York. Maybe someone doesn’t have a credit card. Maybe they are worried about credit card fraud, etc. Bunch of nonsense. No one at scale uses credit cards to establish “real humanity”. (<— If you argue, I’m going to point to “at scale” and “real humanity”, so come prepared for that.)
Next up! If Twitter collects enough personal information of any sort to establish “real humanity”, then they’d be a target for all kinds of shenanigans by criminals and repressive governments. _Without engaging with who might do that_, I would point out that it should be fairly straightforward to collect the information for the purpose of verifying an account, and then _purge the information completely_. You don’t have to save it, and you probably shouldn’t (for account NOT of public interest — accounts of public interest are a different matter, but Twitter seems to have a solid solution there). But if you decided _to_ save it, there are tokenization methods to save it in a disconnected, silo’d manner, and only permit access by a very limited number of people in a highly controlled way. Companies that lose their credit card data usually have not set up these kinds of controls (for a long time, many companies were using credit card numbers — untokenized! — as customer identifiers, which is beyond insanity).
I wasn’t going to post a long thing about this, because while it’s great that Fung has engaged with the core issues here — that Musk has a stated goal (authenticate real humans) to solve a stated problem (cryptocurrency spam bots who are stealing his likeness) — the _way_ that Fung has engaged with that core issue falls short of typing this much.
But I decided to post a long thing about this because of how it wraps up.
“After years of trial and error, tech platforms have already developed important lessons about user authentication that could benefit Musk, said York.
"If he merely means things like CAPTCHAs, I think he's in for a surprise," said York. "He's talked a lot about how he'll get rid of bots, but Twitter's been trying to do that for years and I think he'll soon realize it's not an easy problem to solve."”
It’s hard to know really what to do with this. I mean, do I snark on EFF? Do I snark on York? Do I snark on Fung? Do I look at it and point out how all statements of the form “If FALSE then” are all by definition true? Because “he merely means things like CAPTCHAs” is _obviously false_. Duh. No, he does not mean that. Come on. Rockets landing on their butts. Starting a new car company in the United States. Judge signed off on the SolarCity acquisition. Harassed Al-Rumayyan in email and is still walking around quite fine and with even more money now than before. Tweeted about how he wanted to go mano a whatever with Putin as a way to resolve WTF and no one even bothered to mention that again. Did not had to pay up for calling someone a pedo who clearly was not and who went to the bother to take him to court over it.
I am pretty sure this is a person who knows what is a hard problem and also how to solve hard problems. He’s not thinking it is an easy problem.
I mean, unless he _is_ thinking it is an easy problem, in which case a sensible person with any access to wikipedia and the ability to read should sit back and wait and see how it turns out because if it _is_ an easy problem for him, then he should damn well solve it so twitter works better for everyone.
ETA:
Sometimes, I’m struggling with something, and my husband shows up and tries to take it over. This is terrible! He should not do this. He does it to the kids, too, and he _really_ should not do it to them, either. It is terrible parenting. It induces dependency. It is disruptive of process. Etc. I’ve done a lot of screaming about this (with cogent arguments — I don’t just verbally abuse people, I raise the volume and stay on topic. I’m not new to this game). Still, he’ll show up and try to take over a task.
Lately, if I don’t have specific concerns about him ruining something that cannot be corrected, I’ve been letting him. I do this now because I slowly realized a couple of things. First off, if he can do it, and he wants to do it, sure, fine, whatever. Go ahead. I have plenty of things to do, and if I run out, Lego makes more things for me to do. Second, if he thinks he can do it, and I really do know he can’t but also that he’s unlikely to break anything irretrievably, it’s far more fun to have that as another example of the list of Remember That Time When You (I have a very good memory. And a blog. And other documentation of my life) than it is to scream at him. AND — and this is golden, seriously, keep it in mind — the amount of conflict in my life is drastically reduced. Partly because I’m not screaming, which is important because I hate screaming, but mostly because it turns out that even a short list of Remember That Time When You is a profound deterrent. Honestly, I can evoke most of it with two or three words in a particular tone of voice. He’s even started _waiting_ (arguably hovering, but I’ll take it) until asked before reaching for something and just taking it over.
Musk is a little feral still. But if he’s really got a trick that’ll get the stuck lid off the pickle jar, I say let him give it his best shot. Twitter is not in a good place as a public company, and there is no clear path forward for it to get to a good place as a public company.
Think of it as Bezos buying the WaPo. When El Jefe bought the paper, there was a lot of wringing of hands and fear. But now, it’s a really nice news source, very well run, even the comments are often worth a glance. And honestly, the odds of Musk being worse for our country than the Murdochs have been seem kinda low.