I asked ChatGPT earlier this morning if it could write an Android app for me as if I knew nothing. It said "of course", but offered me alternatives:
- to do all the work itself and all I'd have to do is download a file and install the app
- to teach me how to do it
Hey, I like the mindset it is trained on, on this matter! Applying well the proverb: "give a man a fish you feed him for a day. Teach a man to fish and you feed him for a lifetime.".
Guess what I chose? Well... the title may have give it away, but I actually wanted to go through the process and see how difficult it would be for someone as rusty as I am to do this.
So, on the way to build my Android app, the AI told me exactly what I need to install and troubleshoot one small error at some point, I created a virtual android device (I could choose the OS version I wanted/needed) on my laptop to test my apps, build and run a small demo app (which at first I screw up because I built the wrong kind of app, but it immediately figured out what I did from the error I received, and then start building the actual app).
I just noticed the background on the Android emulator is pink-ish, lol:
I'm not finished with this small app, it doesn't even matter what it does, but I can slowly build it, even if I have no knowledge of Flutter and won't pretend that now that I'm "building" it, I do. I'm mostly copy-pasting and reporting back the errors to ChatGPT.
The process seems to be working and so far I haven't run into any blockage.
It is of course an easy app, and I already know they are capable of creating from scratch to fully-working code things more complex than that.
Not all code is perfect, but if do a ping-pong with the AI it will very soon fix it. At least it did where I got errors or even warnings, also giving me the context to understand why the errors happened.
Where I believe such an endeavor might have its limitations is on medium/long term focusing on a task. I used ChatGPT without saving the conversation (theoretically, I don't believe it for a second they don't use these conversations), so it would be pretty difficult to pick up where we left off the next day or a week from now. But I know this is an issue even if you allow the AI to keep records. It "forgets" context after a while, which humans don't do.
I don't have more time today to play with it since we have to go to a celebration, but I'll do my best to see if I can complete the app and see what kind of context I'll need to provide to the app tomorrow or whenever I'll get back to it.
Posted Using INLEO
LLMs have been a great learning tool for me. I have limited time to play around. Trying to reason around the code and making few changes to see what happens has been fun. I played around withe Haskell and Python.
Enjoy your celebrations!
It's definitely a great tool! Both for non-coders and for coders who want to improve their productivity.
I'm hearing about people using the bots to write software, but I've not got into that. Maybe I am missing out. There is this whole thing of 'vibe coding' where you do not even look at what code is generated. I like to know how things work. I do wonder if my job could become less in demand, but then I could retire in a few years anyway.
!BEER
When I wrote code, I also loved to understand what I used (including when I took snippets of codes from others, often from online sources). Now I don't care as much, since I doubt I'll go back to coding and yes, I do believe this job is in high danger from the AI, especially if one isn't important enough to be on the irreplaceable side, when the first waves of cutting come. The social side may also matter when decisions are made, so a great programmer without social skills may not be as desired in a group anymore - just a guess.
It's just amazing how in few years Ai have turned... Right now as you say they are not really 'intelligent', just a pretty complete too but without much awareness
Their biggest problem nowadays compared to humans (and one I hear is in the focus of the AI companies), is that they cannot follow a plan for days, weeks, months, years, while we can. They compensate by being able to be trained much quicker and in many cases better than humans can (up to months instead of years), and once one is trained, any other AI instance of the same or similar model can copy that training almost instantly - that will make a huge difference for robots, when you can mass produce them and install the same "software" in all of them and upgrade it to all. That may also be a problem if those upgrades or the initial software contains glitches, bugs, limits that can be breached, etc. because you have an army of faulty robots potentially going "nuts", as we have seen in one public case.
Intelligence is relative. One could say they are not because they only rehash what they have learned in training and since or what they can find online or in the sources you provide or the prompts, so at least LLMs are language-intelligent and may mimic other types of intelligence based on the training. But they mimic it very well in many cases. Not to mention kids learn by mimicking their parents and people around them too.
It sounds like ChatGPT is helping quite a bit. It's not perfect, but it gets you to a starting point, and from there, its trial and error.
I probably could have done it without ChatGPT, but it would have taken days to research and install everything and get the sample app running. And a lot of frustration. About the code for the new app, I don't think I could have done it. I don't have the motivation to learn a new programming language, and it practically gave me the code, step by step, even if it had a few minor errors here and there, easily fixable (by ChatGPT).
#hive
Thanks for sharing!
Well done 👍. Your post is very informative 😁.
Thanks for sharing Valuable content 🙂
Congratulations @gadrian! You have completed the following achievement on the Hive blockchain And have been rewarded with New badge(s)
You can view your badges on your board and compare yourself to others in the Ranking
If you no longer want to receive notifications, reply to this comment with the word
STOP