3 ms·
I feel like the people who think ChatGPT is pointless, unusable, boring etc, don't really get it. If you need to know a hard fact, just Google it. GPT will tell
by alar44 4y ago
I feel like the people who think ChatGPT is pointless, unusable, boring etc, don't really get it. If you need to know a hard fact, just Google it. GPT will tell you what to look for.
For example, I needed to learn about Slack chatbots. I asked it what frameworks I should look into, then how to use a specific framework or API. "Write me some boilerplate code doing X in Y framework". It's gets you up and running in minutes.
I'm exploring and learning other programming languages at a crazy rate because I can have it translate something I wrote. Or, "Hey, give me a demo C++ program that uses arrays, for loops, logical operators, and a simple class." "Great, now rewrite this in Node.js."
Sure, it might make a mistake, but copy and paste your error code and 90% of the time it compiles.
I highly encourage any skeptics to just sit for a couple days and try to get it to write code. Once you get the hang of what it can and can't do, it's immensely helpful. This is vacuum tubes to transistors paradigm changing shit.
- ZephyrBlu 4y agoYou're saying, "if you need to know a hard fact, just Google it", but at the same time saying "I'm exploring and learning other programming languages at a crazy rate". Those are somewhat contradictory to me. Information about programming languages are hard facts. If it translates your code into another language how do you know it's using best practices? How do you know there aren't subtle bugs in its translation? Is it using modern syntax? I don't really see much value unless in "exploring" another language like this. Basic data structures, control flow and logical operators are very similar in all languages. If it could show me modern syntax following best practices that is correct then that would be insanely useful, but there's no way to know whether it's showing you junk or not.
- jononor 4y agoA difference between ChatGPT and Google in this scenario is that Google cannot really synthesise or compose facts/knowledge together. These giga language models are decent at this, if anything they do it a bit too well/much, putting vaguely/somewhat but not quite right things together alot of the times. Which probably has some uses...
- insanitybit 4y ago> but there's no way to know whether it's showing you junk or not. My experience so far is that it's light on junk, and the junk is really obvious.