If you'd like to know what brain surgery is like, read this piece. He had a 13-hour operation - done by proper doctors, not Muskys - and was conscious the whole time.
David was one of the finest people I've ever known. He died in 2016, but this op gave him some respite for a while. https://www.theguardian.com/society/2002/oct/15/health.lifeandhealth
A weird little extra spice in the online US rightwing's renewed obsession with eliminating porn is that they're having to pay the piper for old deals. And the piper is leaving town.
I mentioned 6 months ago that MAGA on Twitter was entering its swimsuit edition. Those memes of Trump as Jesus weren't completely disappearing, but they were being displaced by photos of busty white women wearing skimpy US- or Trump-themed bikinis.
Doesn't really matter that Yandex searches confirmed the "wholesome" models shown were almost all from former East Bloc countries, not the US.
What it really confirmed beyond reasonable doubt was that a very large percentage of "MAGA" is smoke and mirrors. On Twitter, particularly, its spambots and scambots - not real people.
Left to their own devices between major elections, they revert to normal business, and a lot of that is porn.
>>Note: I’ll have to stop early due to Dr appt but will add what I miss when I return.
As always, I can’t reply during live-thread of multiple reporters. Plz remember to preface q’s w “Not for Laffy” or something similar since I’m too distracted to answer. Others can for me. TY.
Fascinating example of how public reality can influence corporate culture came from a strange source this week: the Springboks.
When asked in a presser about how the Boks had dealt with the pressure of defending their world title, manager Rassie Erasmus opened up with: "Pressure is not having a job" (rather than playing a rugby match).
Erasmus is an excellent coach, and he himself has been coached here.
I'm not suggesting he's not a decent man who understands sport's role in society. I'm saying he's been trained to consider the wider implications of the role the Boks play in South African society whenever he speaks publicly.
The team enjoys almost total support in this country, despite rugby's racist history. Mandela's canny olive branch was accepted and it's still paying dividends for both sides. Everyone benefits.
It's a good lesson to teach all those who claim to represent big groups. This is what is meant by a social contract.
Good investigation into how widespread fake-account trolling really is on social media.
This is about corporate trolls, but many big accounts use exactly these methods. (I've been on the receiving end of it at times.)
@NatureMC There's been a lot of discussion on the merits (if any) of gatekeeping, whether it's in reviews or news.
What I mean here is the value of having "professionals" in some way qualified to vent an opinion.
This includes, unfortunately, twats like the "Top Gear" team, as they do know a lot about cars, even if they're not officially part of the industry. So there's a lot to be said in favour of taking that gatekeeping role away from elites.
But I think we might be approaching a problematic extreme, particularly - as you say - with AI.
It's Yeats's "things fall apart" notion, when the "centre cannot hold" as there's no consensus on what opinion is valid.
Eh, I'd better explain how bots affect a site's metrics again.
Since the original thread was motivated by observations about pornbots on Twitter, I'll focus on the one that's most talked about in terms of Musk's dumpster fire.
The famous DAUs. That's Daily Active Users and is a metric used almost exclusively by Twitter.
(There's also the MAUs - Monthly Active Users. MAUs are the commercial industry standard and are used to set advertising rates. They are audited and what sites are allowed to define as "active users" is limited.)
They both have an "m" before them - mDAU and mMAU - that stands for MONETISABLE.
To simplify for the purposes of this thread, they measure the number of accounts that are logged on to a site and therefore have the theoretical possibility of seeing an ad.
To be a bit more specific, they supposedly measure the number of devices that are logged on. (This explains the old pictures of clickfarms with thousands of phones. These days there are programs to fake the notion of it coming from individual devices.)
DAUs don't have the same constraints as MAUs, which is why they are so beloved of Twitter. You can fit a lot of bot activity into the DAUs.
It's not a Musk-created con. DAUs were widely used until advertisers realised how much sites were rigging the numbers.
Twitter persisted in publishing its DAUs because it has always struggled to generate ad revenue.
I'm not going into all of this deeply again. I'll focus on the bots.
I group bots into four categories. I'm willing to add a fifth - "good bots", for those not intentionally involved in fake amplification - as people seem so butthurt about this.
But these are broad categories that simply explain purpose.
There are not only four bot networks. There are hundreds of networks, each containing thousands of "botfarms".
They interoperate within their own categories and with other categories. Classic pornbots are the most likely to stay in their own lane.
Outsiders pay either individual botfarms or networks for amplification.
The botlords very seldom pay each other. Instead, they work in a barter economy. Something like: "I'll give your customer 50 followers if you give mine 100 likes", though it's more often trading like for like.
What drives this entire model is human beings and their insatiable desire to appear popular or relevant on social media. The attention economy. Politics plays a role, but it is not the driver.
The reason "Views" has become such a big metric on your page is that Musk is trying to bullshit everyone about engagement.
Like almost everything he does, its roots are so venal it's funny in a ghastly way.
Musk's ego demanded that his be the most-followed account on Twitter. To achieve this - short of Tom from MySpace-ing it - he needed bots. Twitter allowed Donald Trump to do the same in 2015, with similar consequences.
So they didn't show up as fake followers, the Twitters of both Dorsey and Musk changed how they identified "real users".
(The way "real/fake followers" is calculated underpins the DAUs.)
But this was a deal for himself and didn't by itself raise other types of engagement.
So Musk established $8kun, which created a huge secondary market for fake engagement - which can be costly.
"Views" is a sop to their egos. It's aided by later announcements of "profit incentives" for engagement with other blue ticks, which means there's a small fantasy world (there are at most 800k $8kun at present) of navel-gazing constantly in Musk's mentions. He's high on his own supply.
It's extremely difficult - as it was under the previous regime - to get accurate figures on the bots. (They outnumber human users by at least 3:1.)
AI chatbots (spambot and scambots) are increasingly popular and I don't know how Twitter is "counting" them.
The difficulty is exacerbated by how well the company is hiding the loss of human users.
Yaccarino has publicly admitted a fall from around 260M to 230M in the DAUs. What she - nor any Twitter official - is not saying is how many real users have left the platform or abandoned their accounts.
My pesonal guess, based on years of research - some of which involved running fake amplification accounts - and a few discussions with botlords is that it's around 100M.
Pornbots are the category least involved in fake amplification. Their business model is almost purely getting real people to go to their sites.
Tumblr cleared out porn a few years ago. It's allowed it all back because it lost advertisers - who understand that real people look for porn.
The pornbots are quiet quitting Twitter. Makes me believe the platform is in way deeper shit than we officially know. No matter how they massage the DAUs.
People often talk about bots on social media as though they were a single huge group. They're not.
For my research, I've divided them into four categories. This is my own method, and others might set the boundaries differently. I think it's helpful in understanding what goes on, so I'm explaining my system again.
All the groups overlap, but like circles on a Venn diagram they represent separate entities.
Please note: this is about bots - which are fully automated accounts.
This is the biggest - and most problematic - group by several orders of magnitude.
They're almost always what people actually mean when they talk about "the bots".
They are not pornbots or political bots. They are purely commercial networks that will amplify any and all political messaging and will foist porn as well as advertising and "funny tweets" into your feed. They're beloved of influencers.
You just have to pay them (in cash or barter deals).
They also deal in selling fake followers, and they sell any type of engagement. They're the reason for the outlandish number of "views" on Twitter these days.
They work every metric.
This is the newest bot group, mainly because although they're a long-time fake account category, they used to be mainly sockpuppets.
They're involved in all the scams, from crypto to romance ones.
And they're very present in political discourse - even if their scam payoff isn't immediately apparent there. (They're looking for targets.)
Developments in AI mean these accounts are now fully automated. You're not talking to a catfish, you're talking to a computer.
For the purposes of my own system, they've become a separate bot category.
I re-upped this after some discussions on something I posted elsewhere: The pornbots are abandoning Twitter.
There are several causes, but it's the implications that are important.
Porn is the weeds of the internet. Eliminate them for your perfectly manicured lawn and your pretty flowers and you lose the bees and rest of your pollinators. Soon your whole ecosystem is on life support, requiring a lot of intervention to survive.
As with weeds, the pornbot networks are not disappearing. They're moving. Threads is the current major target. Being a whole lot smarter than Musk, I imagine Meta will deal with it the same way as on Insta and old Twitter. Hide it.
(There is a ton of porn on IG - similar levels to Twitter. Those who don't want to see it, don't.)
The picture is dire for Twitter, though. If the industry that normally fights for a foothold on platforms is walking away, your platform is in bigger trouble than anyone thought.
There's an endearing belief among those who don't understand finance that this is somehow all going according to plan for Musk, who will swoop in and buy up the debt for pennies on the dollar.
What they're ignoring is that Musk doesn't have the readies to do that even if he wanted to (which I've never believed he does).
Look at how his fortune has fared recently. Tesla results and price tumult are killing the gains from Starlink and others.
He's not spending another cent on Twitter.
@the_etrain It's hella confusing, but this $13BN or so is not Musk's debt. It's TWITTER'S debt.
(The money went to pay former shareholders, it wasn't an investment in the company. It's a common feature of leveraged buyouts.)
Musk owns enough shares to own Twitter. So he owns the shares (in itself) that Twitter owns too.
He could buy up the debt cheaply to reduce the financial pressures on Twitter and clear the claims on the company.
But unless he's seriously trying to turn it around, there's zero incentive for him to do that.
@the_etrain It's very rare. Why on Earth would they? They already own the company. Usually only done to prevent others buying the debt and foreclosing.
The suggestion seems to be wishcasting by people who still believe Musk's bullshit.
@the_etrain It reminds me of the cartoon meme of the person prodding something dead with a stick and encouraging it to act. There's a lot of people who subconsciously won't accept how awful Musk is.
Financial types (and the business media) are also horribly concerned because the banks hold a LOT of Musk's personal debt - which involves a lot of Tesla stock as collateral for loans.
