2068
What a time to be alive
(i.ibb.co)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 08 Jun 2024
2068 points (100.0% liked)
Programmer Humor
19187 readers
1048 users here now
Welcome to Programmer Humor!
This is a place where you can post jokes, memes, humor, etc. related to programming!
For sharing awful code theres also Programming Horror.
Rules
- Keep content in english
- No advertisements
- Posts must be related to programming or programmer topics
founded 1 year ago
MODERATORS
They put new AI controls on our traffic lights. Cost the city a fuck ton more money than fixing our dilapidated public pool. Now no one tries to turn left at a light. They don't activate. We threw out a perfectly good timer no one was complaining about.
But no one from silicone valley is lobbing cities to buy pool equipment, I guess.
This is so dumb that I totally beleive it
You need to really specify what is meant by "AI" here. Chances are it's probably some form of smart traffic lights to improve traffic flow. Which is not all that special. It has nothing to do with LLMs
Honestly I'm not sure, we had circular sensors for a long time, about the size of a tall drinking glass, now there's rectangular sensors they just put up about twice the size of a cell phone and they have a bend, arc, to them, I know they weren't being used as cameras at all before, no one was getting tickets with pictures from them, it's a small town. What exactly the new system is I'm not sure, our local news all went out of business, so its all word of mouth, or going to town hall meetings.
I'm guessing it's some sort of image recognition and maybe some sort of switch under the pavement telling the light when a car has rolled up.
Nah, that need dat water to cool the AI for the light.
Linus was ahead of the game on this one. Nvidia should start building data centers next to public pools. Cool the systems and warm the pools.
I've seen a video of at least one spa that does that. They mine bitcoin on rigs immersed in mineral oil, with a heat exchanger to the spa's water system. I'm struggling to imagine that's enough heat, especially piped a distance through the building, to run several hot tubs, and I'm kind of dubious about that particular load, but hey.
A large data centre can use over 100 MW at the high end. Certainly enough to power a swimming pool or three. In fact swimming pools are normally measured in kW not MW.
"A large data centre" this wasn't. I saw a couple washing machine-sized vats of oil-soaked computers.
If all it's running is a hot tub that sounds reasonable. This bitcoin miner uses over 3kW: https://www.aozhiminer.com/showroom/high-profit-110th-95th-bitmain-antminer-s19-pro-s19-bitcoin-miner-with-power-supply.html
Whilst it's a shame this implementation sucks, I wish we would get intelligent traffic light controls that worked. Sitting at a light for 90 seconds in the dead of night without a car in sight is frustrating.
That was a solved problem 20 years ago lol. We made working systems for this in our lab at Uni, it was one of our course group projects. It used combinations of sensors and microcontrollers.
It's not really the kind of problem that requires AI. You can do it with AI and image recognition or live traffic data but that's more fitting for complex tasks like adjusting the entire grid live based on traffic conditions. It's massively overkill for dead time switches.
Even for grid optimization you shouldn't jump into AI head first. It's much better long term to analyze the underlying causes of grid congestion and come up with holistic solutions that address those problems, which often translate into low-tech or zero-tech solutions. I've seen intersections massively improved by a couple of signs, some markings and a handful of plastic poles.
Throwing AI at problems is sort of a "spray and pray" approach that often goes about as badly as you can expect.
That was a problem solved seventy years ago. If there's no one around, just go. No one cares.
(I know I'm two months late)
To back up what you're saying, I work with ML, and the guy next to me does ML for traffic signal controllers. He basically established the benchmark for traffic signal simulators for reinforcement learning.
Nothing works. All of the cutting edge reinforment algorithms, all the existing publications, some of which train for months, all perform worse than "fixed policy" controllers. The issue isn't the brains of the system, its the fact that stoplights are fricken blind to what is happing.
A lot of people in Silicon Valley don't like this AI stuff either :)
We are a small software company. We're trying to find a useful use case. Currently we can't. However, we're watching closely. It has to come at the rate of improving.
Using CV for automatic lights is not a new thing.