My biggest problem with the concept of LLMs, even if they weren’t a giant plagiarism laundering machine and disaster for the environment, is that they introduce so much unpredictability into computing. I became a professional computer toucher because they do exactly what you tell them to. Not always what you wanted, but exactly what you asked for.

LLMs turn that upside down. They turn a very autistic do-what-you-say, say-what-you-mean commmunication style with the machine into a neurotypical conversation talking around the issue, but never directly addressing the substance of problem.

In any conversation I have with a person, I’m modeling their understanding of the topic at hand, trying to tailor my communication style to their needs. The same applies to programming languages and frameworks. If you work with a language the way its author intended it goes a lot easier.

But LLMs don’t have an understanding of the conversation. There is no intent. It’s just a mostly-likely-next-word generator on steroids. You’re trying to give directions to a lossily compressed copy of the entire works of human writing. There is no mind to model, and no predictability to the output.

If I wanted to spend my time communicating in a superficial, neurotypical style my autistic ass certainly wouldn’t have gone into computering. LLMs are the final act of the finance bros and capitalists wrestling modern technology away from the technically literate proletariat who built it.

what we thought we knew about autism and... whatnot

'For decades, researchers had been measuring the wrong thing. Conflating communication style differences with empathy deficits produced dramatically inflated effect sizes and an illusion of empathy impairment'

#autism #actuallyAutistic #science #psychology

https://www.psychologytoday.com/us/blog/positively-different/202601/what-the-world-got-wrong-about-autistic-people

@davidgerard , not sure if you saw this great post about ML models in radiology?

The boosters love to say ‘look how good AI is in medical diagnosis, therefore LLMs are good’. Only, it turns out (from the article):

while models beat humans on benchmarks, the standardized tests designed to measure AI performance, they struggle to replicate this performance in hospital conditions. Most tools can only diagnose abnormalities that are common in training data, and models often don’t work as well outside of their test conditions

It also highlights a problem that’s actually quite general in medicine: we have far more data about unhealthy people than healthy ones. I was talking to a cardiologist almost ten years ago who was very excited about the data things like the Apple Watch could collect. Apparently they know that a lot of people who have heart attack have arrhythmia, but they have no idea if this is a meaningful correlation. Healthy people tend to have their heart monitored for a minute or less on a visit to a doctor every few years. People with known heart problems wear heart monitors that can record a load of things, so you have very good data on their heart rhythms but no baseline to compare it against.

This is also true for radiology. You really want to do anomaly detection: take a few million scans of healthy people, wait a few years to see if any of them have undiagnosed conditions, and then use that dataset to train a model of what a healthy lung (or whatever) looks like. Then feed new scans to the model, have it flag anomalies, and loop in an expert to figure out what kind of anomaly it is and whether it’s important.

But what you have is a load of very examples of things that are wrong, in very specific ways. And these also have artefacts that are specific to individual devices, so it’s easy for a model to learn that people who are scanned with this class of machine have this condition.

And that’s just the start of the issues they discuss.

Telling people that they need to personally refuse to obey fascist laws is... not ideal. Not everyone can equally afford the risk of legal exposure like that.

It's great when folks can refuse to obey, and we desperately need more people refusing to obey, but please be cognizant of the risks you're telling people that they *personally* must take on.

going to spend some time in a mostly advisory role helping alpine rebuild the TSC and security team.

as always, everything on my account is just my *personal opinion*. my *personal opinions* on this matter are as follows:

in general, i am at a point in my life where i just want stability and reliability from alpine, but otherwise i want to focus on new stuff. plus, the new kids are alright, you know?

my goal is to get both of these functions working to a point where i don't have to worry about them ever again.

Also hot take but the new #Android sideloading unlock requirements are kinda reasonable. We'd literally need to implement something like that in mobile Linux once it reaches the mass market. (Like actually mass not just a bigger niche.) Scams coercing people into installing malware is a VERY REAL PROBLEM.

Actually psychology research is much needed to properly test how software can try to break coercion like that.. I bet some dedicated scammers would try to guide their victims through the entire 24h delay. But it's probably good against random links and such