That's a good summary of one fundamental flaw of LLMs. One. Of many more. Fuck the slop.
this post was submitted on 17 Jan 2026
151 points (98.1% liked)
Comic Strips
21169 readers
2453 users here now
Comic Strips is a community for those who love comic stories.
The rules are simple:
- The post can be a single image, an image gallery, or a link to a specific comic hosted on another site (the author's website, for instance).
- The comic must be a complete story.
- If it is an external link, it must be to a specific story, not to the root of the site.
- You may post comics from others or your own.
- If you are posting a comic of your own, a maximum of one per week is allowed (I know, your comics are great, but this rule helps avoid spam).
- The comic can be in any language, but if it's not in English, OP must include an English translation in the post's 'body' field (note: you don't need to select a specific language when posting a comic).
- Politeness.
- AI-generated comics aren't allowed.
- Adult content is not allowed. This community aims to be fun for people of all ages.
Web of links
- !linuxmemes@lemmy.world: "I use Arch btw"
- !memes@lemmy.world: memes (you don't say!)
founded 2 years ago
MODERATORS
These people are better at using AI than like 99% of everybody, since they are actually critically and carefully reviewing its output.
One of the big problems with actually using AI for work is that, even for people who carefully and critically review it, sometimes, it is just too good at making slop look like something real.
Also having the discipline to do an actual proper review of the output is hard. I bet that gets skipped s lot even for people who know better.
Some of the hardest part in software development are testing, validation, code review, is so tedious, the majority of people don't like to do it so is hard to find qualified people and so on. This AI thing just put way more stress in the already dire part of the whole development.
It's amazing how fast AI can provide an answer to almost anything.
It's like I'm extremely fast at calculating.
If you ask what is 257 time 389 I can give you an answer in a split second.
4809!! It's obviously wrong, but it's FAST!
If you ask what is 257 time 389 I can give you an answer in a split second. 4809!!
So awesome! Would you like some billion dollars venture capital?
About time someone came with an offer, thank you I think I deserve it. And I'm sure you will be very happy with the speed improvements I can provide.
I love how complex and confident sounding some of the replies are, and then you click to see the "reasoning" and it's something like:
Alright I'm diving in into the concept of numbers. First, I need to understand what a digit is. Digits are the protrusions that are often found at the edge of a human hand. Wait, that is incorrect, digits are mathematical symbols. I'm making progress, my search results suggest that digits can be both mathematical unitary symbols and human anatomy terms. The user asked for the result of 1+1, I'll invoke the Python agent and code the operation, analyze the input, and re-frame the answer. It appears the Python agent returned with a malformed output, I'll check the logs. I'm frustrated - the code is clean and the operation should have worked. I've found the error! The output "NameError" clearly indicates that I've accidentally mixed data types in Python, I've been crunching through a fix and am confident the calculations will proceed smoothly. Writing final answer, factoring in the user recently asked about the job market in 2026.
Based on the current job market and listings found on online sources like LinkedIn, you will appreciate that the answer to the expression 1+1 is 2, would you like me to create a graph showcasing this discovery so you can boost engagement on your LinkedIn Profile?
The inefficiency for each query is bizarre.
The explanation is a separate query and doesn't necessarily have anything to do with how it presented the answer initially.
You're absolutely incorrect. The explanation is not an explanation made after the fact, it's a simple technique called chain of thought where the LLM must append a log with this type of "reasoning" during the entire process, as that's been shown to reduce the rate of errors in complicated queries.
Explanations that are a separate query are only the title it gives to the conversation and the little one sentence "progress" updates it gives (in certain UIs, like Gemini, others just leave a default "Thinking...)
If you want to add 1 and 1 together, you must first invent the universe.
This is a pretty perfect explanation of why you're not going to see AI being used for anything serious in the art world for a very long time. you can't randomize your character designs every time you need more than one clip, angle, iteration, that breaks the function of comercial art entirely. Yes I know it's technically been done, but it's exactly what people call it, AI slop. You can't randomize every element of a film every time you need a new shot, a cut, a new scene with the same characters, backgrounds, or lighting.