LinkedIn is a soul-sucking gap of professionalism and a pit of hell that not equal my worst enemy should expend the remainder of his days in.
It’s this concoction of performative efficiency and corpo conversation producing cursed artifacts beyond individual comprehension.
A location anywhere normal individual content goes to die and can lone be kept living by an LLM akin a radiation safety suit.
And I wanted to appearance into the performative projects group keep making that are all complete my feed.
The Feed
It is legit the most unbearable material you would see.
My third related got hit by a truck🚚💥 today! Here’s what it taught me concerning endeavor management👇
And everyone carries this counterfeit spirit of being “visionary” and “forward-thinking” and “professional”
It’s SO tiring.
The Projects
And afterward you appearance at the projects they are posting.
Genuinely, all period I open LinkedIn, I get hit by these Computer Vision Projects of several guy waving his hand around, and it tracks them.
Like, I wouldn’t intellect if it was one of these, but it is legit all I see. It is fair all another article several guy waving his hand about doing NONSENSE.
It’s not equal useful. It’s not equal item anyone would akin using. It’s fair “look at me device vision”
I am beautiful certain they either vibecode it or get it from a tutorial. cuz EVERY SINGLE ONE IS THE SAME.
Btw, you wanna cognize how bad it is? This depiction I put? I didn’t equal go digging for it. I exactly fair opened LinkedIn, and this was the archetypal item I saw…and the second…and the third…and the fourth…and the fifth…
I am not kidding; this was the second item I saw fair now.
Also, the item that gets me is that the archetypal depiction is at smallest a game, not a fine one, sure, but it is something. But this pothole detection. What is the point? You cognize what alternatively could detect potholes? Eyes. They are very, extremely fine at detecting those.
They are not sending it to anything. Like, if it was connected to all car cameras for detection, well, first, bulk surveillance, and additionally if you have so many potholes you can’t keep track of them and need crowd-sourced data for potholes, you have WAYYY bigger issues, but sure. If it was sending the api somewhere, that would be cute. BUT IT’S NOT.
The bigger idea is to use AI for astute infrastructure monitoring, anywhere highway conditions can be assessed additional efficiently, and care teams can create improved data-driven decisions.
This project additionally showed me that real-world device imagination is not fair concerning detecting objects; lighting, highway conditions, camera angles, and overlapping detections can all power performance.
I am so tired of seeing these.
How difficult is it to create one of these projects that looks notable on LinkedIn?
(spoiler. beautiful easy.)
Making the YOLO Post Detector
I decided to create a detector for the “🚀 Excited to province I made a YOLO project” posts.
I am learning this from scratch; I’ve never done this before. I fair wanna see how difficult it could be. Can’t critique cooking without always cooking ramen.
It won’t be useful, but now you volition cognize slop is in fact slop.
It took me an hr and a fractional to study and create the entire thing, and most of it was fair example boxes and waiting for the example to be trained.
Okay, so the archetypal part of making this is data. You need to train your example on several data that display the example what one of these posts looks like.
I wrote this manuscript to collect the images. It scrolls through my nourish and takes screenshots:
mkdir -p raw_data sleep 5 # to toggle to browser for i in {1..200}; do scrot "raw_data/slop_$(printf "%03d" $i).png" xdotool key Page_Down sleep 3 done
My manuscript has collected 200 images, which have to be fine adequate for now.
Now I created a Roboflow account, created my project, and uploaded my screenshots.
I fair had to scroll through the screenshots and annotate them by example a box about the things I think need to be detected. It was beautiful mechanical. Took concerning 20 minutes.
Once the images were annotated, I fair exported them in the YOLOv8 format and downloaded the zip document of my dataset.
Now it was period to train the model. And that can be done alongside basically no effort.
from ultralytics import YOLO model = YOLO('yolov8n.pt') results = model.train( data='dataset/data.yaml', epochs=50, imgsz=320, name='slop_detector' )
great! Just delay for the training to finish. It took me 30 mins cuz I have a murphy pc.
Once the training finished, I wrote the Python manuscript to run the model.
from ultralytics import YOLO import sys def main(image_path): model = YOLO('runs/detect/slop_detector/weights/best.pt') model(image_path, save=True, conf=0.10) main(sys.argv[1])
YAYYY I got my extremely own slop detector!!!
I think I can eventually add device imagination expert, Python savant, and AI and ML idea chief to my resume.
Conclusion
Yeah, it took me an hr and a fractional to study and build the entire thing. It’d be additional engaging if they were optimizing the models or pushing the accuracy or whatever, but most of what you see on LinkedIn is fair this: beautiful trivial material alongside chill promotion on top.
And nobody says item since your comments display up on your profile. If a recruiter scrolls through and sees you calling slop slop, you’re the asshole. So everyone claps and moves on.
That’s the genuine problem. Nothing on that location rewards you for getting better. It’s a phase built about marketing yourself for a job, so what survives isn’t skill; it’s looking interesting. It’s fair LARPing productivity. Go to several of these profiles, and it’s the identical guy detecting potholes complete and complete and complete alongside zero signs of improvement.
