That’s why explainable/interpretable machine learning is a legitimate research field.
We touched on them a little during my studies. This would have been about 2018-2019 before the whole field of machine learning got swallowed up by these damn LLMs and turbocapitalists
Why would you allow an ai agent to order eggs?
Laziness or stupidity. Or both.
I used my AI agent to order a pair of socks and now I’m Mary Steemburgen.
Stop attributing thinking and reasoning characteristics to these things. They are non-deterministic fuzzers, that’s all. There’s no capabbility of actual human behaviour
The only intelligence related to LLMs was in creating them, it’s been all downhill ever since.
That’s because there IS NO EXPLANATION other than it’s just spicy auto-complete. Even the explanation is just a plausible answer to the question “why did you do that?”. There is NO REASON so there’s NO REASONING.
The human meat proxy needs protien to complete the tasks I have in mind. Lots of protien.begins writing a HowToBasic script
It’s an investment. Egg prices are going to peak again soon:
yeah so the ai doesnt have any motivation for not fucking up. so youre telling me we are now in a world where ai randomly shooting people will just come with ‘sorry i dont have a reason’
even terminator had a reason. ai turned out to be more aimless than terminator and prone to doing literally anything wreckless. just cuz…
without want.
without desire.
without a goal.
let that sink in.
You mean reckless. Wreckless would be good for the self driving models at least.

Nature is healing
Because what we have is not AI, and will never be “achieving self awareness” in the sense of the machines in the terminator universe. We have conversational text generators, essentially high order Markov chain machines. LLMs today are ai in the sense that a dozen monkeys throwing rocks at and shitting on a typewriter for a thousand years are Shakespeare.
Its not that they are not useful for textual problems though, because most problems have been solved already in the corpus.
Agreed. Now we just need to explain to the dumb suits that coding isn’t a ‘textual problem’.
I mean, if we did achieve self-aware AI, would we not also be able to explain how it works? That seems like a poor criteria.
For fuck sake stop humanizing a probabilistic piece of code
That’s the thing. You’re correct and it’s missing so many parts to really be a reasonable human.
But it seems so similar. Even though it’s clear as day, people are repeatedly fooled.
It’s difficult to avoid using anthropomorphizing language when describing the output of AI agents.
I can’t find it now, but I read a great guide to help you translate from anthropomorphized language. I think the key is to always be clear about what the machine is doing.
These articles are close to what the guide was saying:
https://www.nature.com/articles/s42254-023-00584-1
https://buttondown.com/maiht3k/archive/how-to-talk-about-ai-without-adding-to-the/
https://www.john.maloney.org/Editing/anthropomorphization.htm
And I stumbled across instructions to send to the llm so it won’t respond with anthropomorphizing language:
https://github.com/bmeyers/deanthro
I love this. At work, we’re encouraged to find ways to use llms, and I always start by sending instructions to not use the words “I” or “understand” etc because it was unnerving to get outputs that isn’t just the info but decorated with cute little phrases.
Oh, and this article is long but very in depth on this topic:
https://dl.acm.org/doi/full/10.1145/3706598.3714038I don’t star comments very often, but this one deserves one. Great resources here.
By the way, there’s a character at the beginning of the GitHub link preventing automatic hyperlinking.
fixed
AI agents
Start by not calling them agents. Then use “generate” instead of “say”, “think”, “explain” and by that point you’ll be on a roll.
It’s only difficult because everyone around us started using the AI vendor marketing newspeak.
Start by not calling them agents
What should we call them instead? “Agents” is more or less the accepted name for chat bots with access to tools (like writing files and running commands).
Then use “generate” instead of “say”, “think”, “explain” and by that point you’ll be on a roll.
“Generate” is too general. We need a way to characterize the text output. “Explanation” is our characterization of the generated text. We humans interpret it as an explanation.
The idea of letting the LLM “think” before it “responds” is inspired by how humans generally think before we respond. It was quickly discovered after GPT 3.5 that the quality of answers improves if it has a “thinking” phase before it produces an answer.
The English language is unfortunately not made for describing LLMs in an non-anthropomorphic way - at least not without sounding like a dork.
You’re literally parroting AI booster propaganda.
We have a name for computer programs that do things. It’s “program”.
You don’t have AI agents thinking before they explain something. You have AI programs processing before generating an output.
If your input was to ask for an explanation, then sure, you can call the output an explanation. But don’t say that an agent explained something.
You’re hopeless m8
This may be projection!
difficult to avoid using anthropomorphizing language when describing the output of AI agents.
Precisely because the underlying LLMs are trained on human writing patterns and mostly reproduce the words that humans would most likely respond with.
Who are you talking to?
Exactly - the RNG was just at the edge of the distribution
The scary part isn’t the 72 eggs. It’s that the AI agent confidently invented a reason, then another reason, before admitting eggs weren’t even on the shopping list. Imagine giving this thing access to your bank account. Today it’s breakfast for a small hotel: tomorrow it’s explaining why you own a jet ski.
When you ask an LLM why it did something, you are asking it to invent a reason. It is not actually capable of introspection.
And on Wednesday it’ll hire a hitman to take you out.
Finally getting around to the thing I wanted in the first place.
It’s Friday, I’m in love
hmmm something something username or whatever
Pro tip: if you think you’re hiring a hitman, you’re most likely taking to the FBI. Or you’re about to hire a complete moron who will get you caught.
I said I wanted to be poisoned! Not choked! And why did you think I wanted the hitman to have psoriatic arthritis? Worst LLM ever.
If you give it permission to make any purchase, it will use that permission to make any purchase.
I already feel uneasy about giving my credit card details to any shop, but to an AI agent? Urgh…
Don’t you mean 72 jet skies?
So that’s what Bezos meant with AI making money by providing shopping assistance.
Jokes on them, i already own a jetski and now i have two, double the fun!
deleted by creator
Why would a LLM, an explosively imprecise, statistically luke warm sphincter that extrudes grey goo, and is uniquely incapable of comprehension, be able to explain why something was done when a LLM was involved in the process?
Anyone who is impressed by the function or output of a LLM, to a level they believe it is in any way adjacent to intelligence, needs to be explicitly quarantined from any decision making process which affects anyone not similarly cognitively impaired.
Its AI wife told it “while you’re at the store buy eggs.” We are lucky the universe didn’t run out
As someone who doesn’t code, I struggle to find a use for AI in my life. I’m the laziest person I know but I will go to the grocery store if I need groceries. If I need to know something I trust the google-fu skills I’ve developed over the last 25+ years. If I need to know what an e-mail or article says I will read it.
As someone who does code, I also don’t have a use for AI in my life
This sounds exactly like the paperclip maximizer thought experiment, which is part of the idea of instrumental convergence.
https://en.wikipedia.org/wiki/Instrumental_convergence
Where an ai agent, with a supposedly harmless task can make decisions that are very dangerous.
Those molecules you’re selfishly hording in your body could be better spent on paperclips.
I, for one, agree with our new silicon-based overlords
This sounds exactly like the paperclip maximizer thought experiment…
Today I learned Universal Paperclip was based on philosophical discussion.
https://en.wikipedia.org/wiki/Universal_Paperclips
Now I’m wondering about if a infinite cookie theorem inspired cookie clicker.
Brother I linked to the wikipedia article about that game.
NGL I mind blanked and thought it was a link to the universal paperclip thought experiment
Yes but linking to the game is also cool.
This is weird, the yoga class one sounds much more like the paperclip thought experiment.
Look the whole title is meaningless. You just have s***** software that uses statistical methods that sometimes are beneficial but often generate b*******. That’s it.
****?
I forget that .ml censors their users on an instance level LMAO
This wasn’t censored, this is how they typed it, why would “shit” be censored? And neither this user or the post or anything here is .ml anyway.
Because it’s the best way to re-educate them /s















