Raw LLM Responses

There are a few things here: 1) it doesn’t work exactly as you think it does, it doesn’t just generate the average code, it’s a predictive model and it’s based on utilizing context to refine the prediction, that’s the power of attention mechanism 2) they added real-time search feature to AI, so in this case it doesn’t just gets the most likely outcome from the pre-trained data but actually adjust the prior knowledge to real-time data, so in a way it works like Bayesian model where there is prior knowledge and new knowledge which adjust the knowledge 3) if you guide the model through proper prompting and context then the AI would utilize it’s understanding of language to find the best solution from that context in other words, providing context and proper prompting secures that you have specific code and not just most common code solution 4) how did you do coding prior to AI? Don’t tell me you wrote everything from scratch, I can tell you how you did it - you used google to find an existing solution which is closest to what you want to do and then you adjust it. By solution I don’t mean entire program you obviously breakdown the complex task to sub tasks but there is nobody who would try to reinvent the wheel. So how is that any different than using AI? The only difference is AI can adjust the code to your needs for you if you know how to use it and provide proper instructions 5) you obviously don’t know how to use AI which is fine but don’t think anyone would rate you higher as programmer just because you don’t use AI since the next guy would know how to use AI properly and he would be 10 times more productive than you. So how about adapt or perish?