I'm a software developer. In my experience, agentic programming struggle with two things: 1) Insufficient training data for a programming language or framework. 2) High application complexity. For example, I tried to optimizing CUDA kernels using an agent. It was very difficult for the agent...
In my opinion, we'll see two camps: Camp 1: Simple software that will be cheap to produce. Camp 2: Complex software that will be more expensive than before.
Why, you might ask? Because the new generation of programmers doesn't understand that debugging skills, architectural knowledge, or even...
This person hasn't shared the prompts they used. So I can upload an image and create the prompt "Turn the lady into an elephant." This is AI slope. Reddit is full of these fake AI mistakes.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.