Cory Doctorow actually goes more in depth on the radiologist example in a post from last year:

'If my Kaiser hospital bought some AI radiology tools and told its radiologists: “Hey folks, here’s the deal. Today, you’re processing about 100 x-rays per day. From now on, we’re going to get an instantaneous second opinion from the AI, and if the AI thinks you’ve missed a tumor, we want you to go back and have another look, even if that means you’re only processing 98 x-rays per day. That’s fine, we just care about finding all those tumors.”

If that’s what they said, I’d be delighted. But no one is investing hundreds of billions in AI companies because they think AI will make radiology more expensive, not even if that also makes radiology more accurate. The market’s bet on AI is that an AI salesman will visit the CEO of Kaiser and make this pitch: "Look, you fire 9/10s of your radiologists, saving $20m/year, you give us $10m/year, and you net $10m/year, and the remaining radiologists’ job will be to oversee the diagnoses the AI makes at superhuman speed, and somehow remain vigilant as they do so, despite the fact that the AI is usually right, except when it’s catastrophically wrong.

“And if the AI misses a tumor, this will be the human radiologist’s fault, because they are the ‘human in the loop.’ It’s their signature on the diagnosis.”

This is a reverse centaur, and it’s a specific kind of reverse-centaur: it’s what Dan Davies calls an “accountability sink.” The radiologist’s job isn’t really to oversee the AI’s work, it’s to take the blame for the AI’s mistakes.’

In short, we definitely could (and indeed should) be using tools like tumor detecting machine vision as something that helps humans build a better world for humans. But we’ve seen time and time again, across countless fields that it never works out that way.

That’s because this isn’t a problem with the technology of AI, but the fucked up sociotechnical and economic systems that govern how this tech is used, who gets to use it, who it gets used on, whose consent is required for those uses and most significant of all: who gets to profit?

!Not us, that’s for sure!<

Marmite is really good for this. Some of the other go-to “umami bomb” condiments aren’t vegetarian, but Marmite works well for vegetarians

I’ve found that I actually seem to use more em-dashes since they became understood to be associated with AI — it’s a defiance thing. I mostly type on my phone, and to type an em dash, I just need to long press on the dash.

I didn’t know that about Framework HiDPI displays

I agree, but like others have said, it bodes well that they’re open about this in the article

Thanks for sharing that talk, it’s pretty good.

I always find his talks interesting, because it often feels like it’s the same core talk each time, but changed through iterative refining of his ideas and rhetoric due to how often he gets the chance to talk about these issues.

Even though I’ve seen many of his talks before, I am surprised at how skillfully he is able to weave in new angles and themes. I mean, he has been doing this stuff for literal decades, but still, it’s impressive.

The datasets they are trained on do in fact include CSAM. These datasets are so huge that it easily slips through the cracks. It’s usually removed whenever it’s found, but I don’t know how this actually affects the AI models that have already been trained on that data — to my knowledge, it’s not possible to selectively “untrain” models, and they would need to be retrained from scratch. Plus I occasionally see it crop up in the news about how new CSAM keeps being found in the training data.

It’s one of the many, many problems with generative AI

Thanks for sharing this. This bolstered my spirit.

I liked the bit where it discusses how, regardless of the effectiveness of whistles in deterring ICE, they have proven to be helpful in regular people feeling less alone.

The idea of copyright is to protect the financial rights of creatives, thus incentivising people to make more stuff, right?

Well even before AI, it wasn’t doing its job very well on that front. The only ones with the power and money to be able to leverage copyright to protect their rights are those who are already so powerful that they don’t need those protections — big music labels and the like. Individual creatives were already being fucked over by the system long before AI.

If you haven’t read the article, I’d encourage you to give it a try. Or perhaps this one, which goes into depth on the intrinsic tensions within copyright law.

In that case, the finished product would be of pretty mediocre quality, in my case. That’s because most of my projects are things that I actively want to learn, and I tend to be overambitious when setting aims. Despite this though, even when I completely botch it up, I struggle to think of a time where I have regret my endeavour. I like the learning

No. The main reason I do projects is for the learning. Any useful items I produce as part of the process are just happy side effects.

Something that I’m super chuffed with is that a few years back, one of my most cheapskate friends asked me for advice on buying a new laptop. When I presented their options to them, they were reluctant to cheap out and get a mediocre laptop that wouldn’t last them very long, but they also balked at the price of even the midrange laptops (they weren’t keen on spending more than £250 on a laptop, which wasn’t enough to get anything that they’d consider to be decent and worth the effort/cost).

As a long shot offer, I told them that I could always try installing Linux on their laptop if they wanted to wring another couple of years out of their existing laptop. I was a tad surprised when they opted for this, and even more surprised at how well they took to it; I jokingly call them one of my “normie” friends, because they’re one of the people whose perspective I ask for when I’m trying to calibrate for what non-techie people know/think. I only had limited experience with Linux myself at that point, having only played around with things on live USBs before. I had heard that Linux could give new life to slow computers, but I was surprised at just how effectively it did this.

(A small amusing aspect to this anecdote is that when I was installing it, I said that one of the side benefits of running Linux is that it could boost nerd cred amongst folk like me. They laughed and said that they didn’t expect that this would be a thing that would ever end up being relevant. Later that year, they got a girlfriend who saw that my friend was running Linux, and expressed approval, which is quite funny to me)

That’s sad. Regardless of whether it’s one of the reasons for Microsoft’s nosedive, it does make me feel some unexpected sympathy for Satya Nadella. I also feel pity, because most high up CEOs do not seem happy with their lives — Many of them spend an absurd amount of time at work, even if they never seem to actually do much work, and I can’t imagine how hard it must be to weather grief under such conditions. No amount of money can buy you more time with a lost loved one.

It really seems like a hollow existence.

Being anti-capitalist is entirely fair, but I think being opposed to all businesses is a less justifiable position to hold.

Yeah, this really was an excellent answer. I’m impressed by how he used this example to illustrate the wider strategy without it feeling like a weird kind of “we will cut programs like this and then everything will be fine”. It gives the sense that this is being used as an example because it’s simple enough to be at the top of a large pile of potential savings, which will take time and work to dig through.

To put it a different way, his answer abstracts away the right stuff. He doesn’t pretend that everything will be as simple and easy as cutting this one program was.

Often the promised cost savings never materialise, because when the chatbot fucks up, it’s humans who need to clean up the mess.

This goes way before the Spotify thing. This is just a court ruling that took a long time to come through.

At a certain point, if your entire raison d’être is to do illegal things, you might as well fully commit

That sounds super cool! Do you have any pics?

So many features like this have gotten so much worse over the years. Google assistant is the big stand out one for me. I first switched to Android in 2014ish, and I got heavily into tinkering and automating stuff. I could say “Okay Google, make a coffee”, or “pop a coffee on please”, and Google assistant would hear this, parse it and understand that this wasn’t a command it knew. This would lead to that input being passed over to Tasker, the app I used for automating stuff, and that would then do the behind the scenes magic of turning on the coffee brewer as I was on my way home (It was very funny, because I didn’t have a fancy smart coffee pot or anything — I just used a ball bearing on a track to hit the on button)

Nowadays, I say something simple like “Okay Google, make a note” and it will say “I’m sorry, I don’t understand that” more often than not. The speech recognition used to be so good, especially after training it on your voice for a while. Now it’s just shit.

It makes me disproportionately sad. Like, enshittification is everywhere, but this is something distinct, even if it is linked to enshittification. If they were gating better voice recognition behind paywalls, I’d be annoyed, but much less sad, because at least that functionality still exists. Modern software, especially that produced by the tech giants, has gotten so complex that I wonder whether even the most proficient engineers in Google understand their software nowadays.

Windows update will frequently reenable OneDrive. Microsoft systematically undermines people ability to understand and control their own device.