theneverfox

theneverfox, 1 month ago

I’ll never forget the one professor who put up a side of code… And had no idea what the class was about. We spent most of the class reading together with him to try to figure out what the lesson was supposed to be about

Apparently the guy was one of those crazy low-level guys who can do things I don’t understand but build on top of. Guy just constantly looked bewildered by reality, he belonged in the code world

theneverfox, 1 month ago

Semaphores. It was obviously C++ code with a bunch of threads, but as it was a standalone C++ program it wasn’t really clear why it was lol

theneverfox, 1 month ago

By low level, I mean like kernel work. I’m told he worked on one of the 'nixes way back when.

It was a data structures class, we did Java or Python in the into classes, php & js for Web + db basics and C++ for theory classes. Then you pick your path

Anyways, the guy taught OS, language design, and data structures. He could code fine, he was just a terrible lecturer - extremely disorganized, no lesson plans. He only wasted the one full class forgetting why we were there, but reading his code (labeled by week) then scribbling on the whiteboard was his lecture

I guess I ended up understanding data structures and I never fell asleep, so maybe he wasn’t a bad teacher. It was just mostly just assignments, he didn’t really do quizzes and the final wasn’t much of the grade

theneverfox, 1 month ago

Ah, I phrased that ambiguously - it was in C++, all of our computing theory type classes were.

I just got distracted realizing I graduated proficient in 9 languages and reasonably comfortable in another 3. 2 were from internships, but the rest were all from coursework. The last couple years, I was juggling 2-4 at all times, plus the odd scripts

I always thought I was really good at picking up and switching languages, but I just realized my program was designed that way.

That feels like a lot, do other colleges do something similar?

(I guess you could knock off 3 because we ended up switching every semester in software engineering because cross platform apps were pretty bad at the time)

theneverfox, 1 month ago

I’d call that hardware - if you’re code enough to the metal to be writing machine code (or even assembly), the physical architecture of the hardware is part of your code

Low level generally is one step up - manual access to memory, compiling to an architecture rather than a virtualization layer, etc

Strangely, the guy that taught OO theory did our hardware class, we built bit shifters and wrote programs in risc assembly… And ONE program in machine code with the promise we’d never have to do it again

I could understand someone who writes in assembly, but machine code is a nightmare…I think I got it without any mistakes, but my butthole was clenched for 4 hours, terrified I’d have to debug it

theneverfox, 1 month ago

Uhm… Have you considered that slack has cat picture plugins?

And meme plugins, and 30 other plugins that look for keywords then spam gifs for what you assume can only be an in joke before your time?

Oh, and one of the plugins actually creates tickets from chat, but jira is down and the guy who maintains it is busy writing a panda facts plug-in. So now it just vomits out an error message so everyone avoids the words “ticket”, “issue”, and “status”

theneverfox, 1 month ago

… Why? Surely not for advertisers

theneverfox, 1 month ago

Ah, but you’re one layer off. Projected/potential money/s (in the next 1-2 quarters mainly) is what is truly king.

It doesn’t have to be a good idea, it can be a terrible one - but good sounding words in the board room are what matter

“Hey, so we’ve decided to see if we can run 10 unskippable ads back to back. Simultaneously, we’ve launched a war on ad blockers. This time it will surely work because we found out you can ignore your customers - Elon Musk has shown us the way, he only lost bots with all his innovation. We expect people to get over it in 3 months and estimate we’ll lose 4 users. Between 10x more ads and half our users off ad blockers, we project 20x ad revenue next quarter!”

-Words of a future CEO, probably

theneverfox, 1 month ago

Several months ago, fresh off the high of following through on my resolution to leave Reddit forever, I made the same decision with YouTube. Once ublock stopped working, I’d try out peer tube, or maybe sail the seas

But ublock never stopped working. I watch more YouTube now than ever before, I got totally addicted as I binged in preparation to leave

At this point, I don’t know if it’d be good for me, or send me in a desperate arms race to get my fix

theneverfox, 1 month ago

I mean, I’ve got one of those “so simple it’s stupid” solutions. It’s not a pure LLM, but those are probably impossible… Can’t have an AI service without a server after all, let alone drivers

Do a string comparison on the prompt, then tell the AI to stop.

And then, do a partial string match with at least x matching characters on the prompt, buffer it x characters, then stop the AI.

Then, put in more than an hour and match a certain amount of prompt chunks across multiple messages, and it’s now very difficult to get the intact prompt if you temp ban IPs. Even if they managed to get it, they wouldn’t get a convincing screenshot without stitching it together… You could just deny it and avoid embarrassment, because it’s annoyingly difficult to repeat

Finally, when you stop the AI, you start printing out passages from the yellow book before quickly refreshing the screen to a blank conversation

Or just flag key words and triggered stops, and have an LLM review the conversation to judge if they were trying to get the prompt, then temp ban them/change the prompt while a human reviews it

theneverfox, 1 month ago

But why have drive by wire? Like you touched on, planes have orders of magnitude more testing, redundancy, and need. Not to mention maintenance

Is there a reason cars need it? Powered steering seems to be pretty effective with a better failure mode

theneverfox, 1 month ago

I remember watching a video from a psychiatrist with eastern Monk training. He was explaining about why yogis spend decades meditating in remote caves - he said it was to control information/stimuli exposure.

Ideas are like seeds, once they take root they grow. You can weed out unwanted ones, but it takes time and mental energy. It pulls at your attention and keeps you from functioning at your best

The concept really spoke to me. It’s easier to consciously control your environment than it is to consciously control your thoughts and emotions.

theneverfox, 1 month ago (edited 1 month ago)

Floride is an element. What we use in toothpaste isn’t nearly the same as the industrial byproducts dumped in the drinking water

Edit: I was half asleep when I posted this, fluorine is the element, floride refers to salts with ionized fluorine in them. The stuff in toothpaste and the dentist’s office is sodium fluoride

What is added to drinking water is hydrogen floride mixed with God knows what else, because it’s an industrial byproduct with lax restrictions.

Most of Europe and Japan don’t use this, and despite having great data to do statistics on, there’s little evidence it’s doing anything for dental health

This wasn’t science that led to public health policy, this was a solution to a business problem and a PR campaign

theneverfox, 1 month ago

That’s a reaction to propaganda

It’s better than internalizing a lie, but rejecting a lie over and over pushes someone towards overreaction. It taints your ability to see nuance

theneverfox, 1 month ago

it is a little funny to me that they’re taking about using AI to detect AI garbage as a mechanism of preventing the sort of model/data collapse that happens when data sets start to become poisoned with AI content. because it seems reasonable to me that if you start feeding your spam-or-real classification data back into the spam-detection model, you’d wind up with exactly the same degredations of classification and your model might start calling every article that has a sentence starting with “Certainly,” a machine-generated one. maybe they’re careful to only use human-curated sets of real and spam content, maybe not

Ultimately, LLMs don’t use words, they use tokens. Tokens aren’t just words - they’re nodes in a high-dimensional graph… Their location and connections in information space is data invisible to humans.

LLM responses are basically paths through the token space, they may or may not overuse certain words, but they’ll have a bias towards using certain words together

So I don’t think this is impossible… Humans struggle to grasp these kinds of hidden relationships (consciously at least), but neural networks are good at that kind of thing

I too think it’s funny/sad how AI is being used… It’s good at generation, that’s why we call it generative AI. It’s incredibly useful to generate all sorts of content when paired with a skilled human, it’s insane to expect common sense out of something easier to gaslight than a toddler. It can handle the tedious details while a skilled human drives it and validates the output

The biggest, if rarely used, use case is education - they’re an infinitely patient tutor that can explain things in many ways and give you endless examples. Everyone has different learning styles - you could so easily take an existing lesson and create more concrete or abstract versions, versions for people who need long explanations and ones for people who learn through application

theneverfox, 1 month ago

They’re famously terrible at math, you can relatively easily offload that to a conventional program

I didn’t mean for children (aside from generating learning materials). They can be wrong - it’s crippling to teach the fundamentals wrong, and children probably lack the nuance to keep from asking leading questions

I meant more for high school, college, and beyond. I’ve been using it for programming this way - the docs for what I’m using suck and are very dry, getting chat gpt to write an explanation and examples is far more digestible. If you ask correctly, it’ll explain very technical topics in a relatable way

Even with math, you could probably get a better calculus education than I got… It’ll be able to explain concepts and their application - I had zero interest in calculus because I little explanation on why I should learn it or what it’s good for, I only really started to learn it when it came up in kerbal space program and I had a reason

But you should never trust its math answers lol

theneverfox, 1 month ago

Try reading something like Djikstra’s algorithm on Wikipedia, then ask one to explain it to you. You can ask for a theme, ask it to explain like you’re 5, or provide an example to check if you understood and have it correct any mistakes

It’s fantastic for technical or dry topics, if you know how to phrase things you can get quick lessons tailored to be entertaining and understandable for you personally. And of course, you can ask follow up questions

theneverfox, 1 month ago

I mean… Yeah? Most explanations aren’t great compared to a comprehensive understanding in your head, you already understand it - it would have to be extremely insightful to impress me at that point

The results vary greatly based on the prompt too - not only that, it changes based on the back and forth you’ve already had in the session

It’s not a god, it’s not a human expert, but it’s always available, and it’s interactive.

It doesn’t give you amazing writeups, but (at least for me) it makes things click in minutes that I might need an hour or two to understand through reading up on it. I can get a short summary with key terms, ask about key terms I don’t know, ask for an example in a given context, challenge the example for an explanations of how the example can be generalized, and every once in a while along the way I learn about a blind spot I never realized I had

It’s like talking to a librarian - it gives you the broad strokes of a topic well, which prepares you well enough that you’re ready for deeper reading to fill in the details.

It doesn’t replace a teacher, a tutor, further reading, or anything else - but it’s still a fantastic education tool that can make learning easier and faster

theneverfox, 1 month ago

No, software developer isn’t a fallback term for software engineer, they have slightly different implications. They’re all very loosely defined so they’re almost interchangeable

theneverfox, 1 month ago

It’s usually one or the other. It just doesn’t matter which one

At my first job I was on a contract as a software engineer I with the job title junior developer, because that’s just how the titles mapped

theneverfox, 1 month ago

In general, I think missiles are bad. I think shooting down missiles is good.

There’s the rare exception to this, where the thing the missile is aimed at is about to do something worse than the missile, and the missile has a chance at preventing great harm

This is not one of those exceptions. Missiles hitting in this case would not save anyone, they’d just increase the risk of war

All that being said, you don’t try to negotiate as missiles are literally en route to a country. That’d be extremely messed up, that’s not how you treat an ally, no matter your relationship. You’d want to shoot them down, playing up your contribution if possible. Make them not want to think about how it would’ve gone without your help. Then leverage that later

theneverfox, 1 month ago

They use military ranks, military equipment, and go through training to dehumanize and terrify them to encourage them to kill at the slightest hint of danger

Maybe we should stop letting them larp around like an occupying force…

theneverfox, 1 month ago

I’d rather they just not have any of that, but a stricter judicial system for those carrying the monopoly on violence sounds like a good idea

theneverfox, 1 month ago

I bet they’d supercharge enforcement of the laws they’ve been testing - such as intercepting women leaving the state for suspected abortions, or parents suspected of taking children out of the state for gender affirming care

The laws are set up that you could basically set up roadblocks and force a fight through the system to leave the state… Keeping people from leaving is important if you want a fascist state, because they suck and only “true believers” wouldn’t consider moving

That’s why those laws are so terrifying… They don’t have to convict anyone, they can just be used to suppress movement

theneverfox, 1 month ago

I’m so confused… You hold the long, floppy cloth perpendicular to gravity when you fold it?