AI chatbots were tasked to run a tech company. They built software in under 7 minutes — for less than $1.
(www.businessinsider.com)
from L4s@lemmy.world to technology@lemmy.world on 12 Sep 2023 18:00
https://lemmy.world/post/4934036
from L4s@lemmy.world to technology@lemmy.world on 12 Sep 2023 18:00
https://lemmy.world/post/4934036
AI chatbots were tasked to run a tech company. They built software in under 7 minutes — for less than $1.::A recent study found that AI-bots could talk to each other, reflect, and identify software bugs.
threaded - newest
This is the best summary I could come up with:
Artificial-intelligence chatbots such as OpenAI’s ChatGPT can operate a software company in a quick, cost-effective manner with minimal human intervention, a new study indicates.
The findings came after researchers published another study in which AI agents powered by large language models were able to run a virtual town on their own.
Based on the waterfall model — a sequential approach to creating software — the company was broken down into four stages in chronological order: designing, coding, testing, and documenting.
After assigning ChatDev 70 tasks, the study found the AI-powered company was able to complete the full software-development process “in under seven minutes at a cost of less than one dollar,” on average — all while identifying and troubleshooting “potential vulnerabilities” through its “memory” and “self-reflection” capabilities.
“Our experimental results demonstrate the efficiency and cost-effectiveness of the automated software development process driven by CHATDEV,” the researchers wrote in the paper.
The study wasn’t perfect, however: Researchers identified limitations, such as errors and biases in the language models, that could cause issues in the creation of software.
The original article contains 638 words, the summary contains 175 words. Saved 73%. I’m a bot and I’m open source!
Wow they really try to bury the part where all it did was make a really simple, straightforward game that’s been made thousands of times before and probably existed in the training corpus (and it still manages to fuck that up 12% of the time). More AI “research” blowing smoke up people’s ass.
It’s literally only able to predict what would the code would be based on its training code. It’s amazing that people expect this thing to run anything when it can’t reason. It’s basically an advanced template engine.
You hear that?
That’s the AI Bubble getting bigger and bigger and bigger and when it finally bursts all these same companies that have been over-hyping it are going to claim “how could we have known it was all BS?”.
The other day I tested chatgpt (which I’m using with an extention on vscode), asked it to write a script that identifies where a cat’s nose is and places a mustache on the image.
It looked good for skeleton code but it didn’t work and I had to do some tweaks to make it do so. It has potential, but I doubt it would, out of the box, be able to handle support for a website. There is no need to eliminate the human in the chain. In fact, if you keep the human, you are less likely to run your website into the ground.