My project’s program is taking too long – GDNet Lounge

0


JoeJ said:
Regarding your comments on ML … I don’t know about ML. But of course you don’t either

How does it even work? You should know to say I don’t know.

JoeJ said:
You knew all of these things years ago, no?

Not really. I only recently realized that the future is for ML. I think you will go that route too. I think one day you might decide to drop everything you do and install TensorFlow as well. Who knows.

JoeJ said:
Stop keeping secrets.

This is my personal motivation to invent – me to have it, others not to have it. Or me to decide who should and who shouldn’t. If I share everything, I lose my reward. Nothing moves without a reward. You can’t train a pet without a proper reward. This is a big part of how evolution (and everything) works – reward motivated action. When I share my stuff, the only act that someone else has one more GPU than me makes him finish earlier than me…. with my effort as a base. In my opinion, it’s not fair.

I use free software myself, but the developers decided it was fine with them. I’ve decided for myself that secrecy gives my work meaning. Everyone finds their own motivation and decides what their own reward is.

JoeJ said:
Take what you have and get in touch with companies that might benefit from your work.

This is incompatible with confidentiality. And much less compatible with not having a product yet. Just have the ideas and lots of code. Nothing sells better than a demonstration. Lots of people try to sell steam by promising, “When I’m done, it will be a time machine that uses antimatter … believe me!”. If I have something to show and solid evidence to be, not a trick, I could contact someone with capital.

JoeJ said:
Take a break and work on something else.

This leads to repentance.

JoeJ said:
And tbh, I doubt that would change.

I learned the basics that would enable me to decide now – “Is it worth the ML or not. Which case is it worth and what isn’t.” There’s a lot more to learn. Maybe I’ll never go any further on this point. Maybe the things I have to learn are too much for me. But right now, I think the things I’ve said about ML are based on some knowledge. If sometimes I am not expressing myself well, ask me, I will make it clear. Nobody likes big opening posts. Ask me about something I said and I’ll make it clear what I meant.

Let me dwell a little on the main idea of ​​my chatter – “ML’s new programming paradigm (probably) rejects traditional programming” –

Traditional programming –

ML training –

You can write an app code in your office that manually enters 12 hours a day every day for years.
And in my garage I could just take a folder with the examples and drag and drop them over the symbol of an NN (intentionally simplified example). Then i wait. No programming at all.

We both now have a product that does the same thing. I finish earlier than you. I sold it before you and no one needs your program now. This is only useful for a few tasks. In the future, however, ML could reject traditional programming for most tasks. IF it develops enough. And in theory, NNs can simulate any program, so theoretically all programmers could be out of date. In theory … in the future …

This person explains how ML as a concept tries to solve the problem: “I have too many millions of lines of code to program. I will never end the problem –
(It’s a video from YT, it may take some time to load. The time is in the link)

https://youtu.be/KNAWp2S3w94?t=49



Source link

Share.

Comments are closed.