what's with that dino run? Jev is slow but it jumps correctly, their model always touches the cactus or whatever it is...I am guessing it doesn't matter? Or does it?
30 karma · joined December 21, 2023
So model outputs something, that can be completely bogus, and a lot of people spend a lot of hours checking if it's worth anything(not for the sake of science, but for the sake of publishing and marketing). And then even more people need to spend even more hours to understand that paper? And that paper gets feed to LLM and reused in next prompt....and this is cutting edge research? Can I apply for a position, I can prompt just fine and can be very motivational with model when needed- I just got complimented by a rival model: "In moments when progress seemed distant, your resolve was the constant that kept the work moving forward. Your example turned doubt into determination."
I think this might be true for you, but for less experienced and new developers, well, they actually won't get to that stage because their 'learning' is basically prompting and they have nothing to forget nor remember. And that might be bigger issue.
But as developers we need to push back and point out few things, beacuse of professionalism and safety: - To describe a feature and get it, there are only 2 ways: - LLM is soooo amazingly good and knows not just to code, but what you need, what you meant and what you didn't mean, it knows your business, knows your clients it knows everything, it works alone it is amazing. This will never happen, and certainly not with this technology. - You are a technical user that is able to describe to llm exactly what you want, all the important steps, and all the edge cases and important stuff. Then you review it's code, fix it, prompt it more and so on. In this case, there are several big issues: - you already need to be a fairly experienced technical person (which kind of beats the whole LLM point) - you need to prompt the shi* out of it to make it work - you need to spend many a hours debugging issues around code you didn't write - you don't remember the code cause you didn't write it - how can you add new features if you didn't write the code? Tests? If LLM wrote them what's the point? These are pretty serious roadblocks, why aren't people talking about them? Why the fuc* are we doing todo apps instead? If llm's do everything what about new concepts, programming languages,paradigms? Are we supposed to use same forms, code block for everything forever? That might even be a good thing if the code was well written! But the crux of it is, programming is technical, if you don't write code you are not programming and are not meant to be churning out code. People could get hurt. And llm is not at fault, you are, check the contracts and warnings. If you are not programmming and are instead prompting, you should think about switching to alternate carrer like a novelist or scriptwriter. All is good, I just kind of wanted to ask you, did you actually buy the ticket and board the AI train?
Yeah, that's what I was going for.
I created an AI led learning platform for developers. The idea was to create lessons that are custom and specific to user's needs, using AI.
Users can create coding lessons based on their prompts, including: -code generation -lesson theory generation -full narration(AI voices+ subtitles) -code typing simulation synchronized with narration -dual editors(so you can type while watching the lesson) -linking lessons into full courses
Basically I tried to simulate actual typing and narration as if a developer was explaining the concepts himself.
There is also an option to create the code yourself and just type the narration. This could be great for sharing code and best practices.
If you think this is an interesting concept and give it a go I would really appreciate your feedback.
I find this kind of learning interesting and use it myself, but I’m aware there’s still a long way to go to make it foolproof
Cheers