3 ms·
I think this can be generalized to, “when you lose the ability to create a certain Form, you also lose some of your ability to imagine new types of that Form, a
by keiferski 4y ago
I think this can be generalized to, “when you lose the ability to create a certain Form, you also lose some of your ability to imagine new types of that Form, and instead just rely on shuffling existing versions of it around to generate novelty.”
A bit wordy, but I have the skill of drawing/sketching in mind. There is a big difference between sketching an idea on a blank sheet of paper, and using some kind of software tool that puts you on predefined rails from the start.
That said, I think the notion that ChatGPT replacing “good writing” as an artistic activity is basically nil. It’s more like the next generation of word processing: immensely useful for the right operator, not that useful for people without something interesting to say in the first place.
- TacticalCoder 4y ago> I think this can be generalized to, “when you lose the ability to create a certain Form, you also lose some of your ability to imagine new types of that Form, and instead just rely on shuffling existing versions of it around to generate novelty.” And it has a very serious consequence: "when you lose the ability to write, you lose the ability to create new content on which future LLMs are going to be trained". One guy has been exposed for producing one computer programming book per week, for six months or a year straight, all with 5 stars reviews. How many people like him are gaming, for their own benefits, the system? It's all too obvious on GitHub and StackOverflow already: people gaming the system with automatically generated code that just, to reuse you words, "rely on shuffling existing versions". How can this LLM-generated content possibly be used to created a better LLM? Basically: how do we train future LLMs using content produced by those among the humans whom didn't lose their ability to think?
- leroy-is-here 4y agoI think the generalization is more minute than that. Take mathematics for example, we can write that down and so we have. A lot of people “hate” math because they understand math as the shuffling of symbols. They hear “math” and think “arithmetic”. Math appears a lot like some puzzle-game because when they read textbooks they see these symbols being shuffled around which produce exact answers, like 4.8473. But mathematics is about expressing ideas which can produce answers for specific problems, not specific answers in general. It’s the allegory of the cave all over again, mathematics has become a shadow of itself in many people’s minds. Instead of a tool in which to express ideas, it just became stuffy definitions and arcane symbols which produce hard yes/no answers. Why? Because when those ideas are expressed by writing down the mathematics in textbooks, they are expressed in arcane symbols and stuffy definitions which have exercises that have hard yes/no answers.
- TheBigSalad 4y agoHonestly ChatGPT is hot garbage. I can't find inputs to make anything it writes sound like it comes from a real person. It's very good at sounding like a specific, boring person. It's impressive that it has mastered human language, but is it really close to mastering humanity?
- keiferski 4y agoI think it is much more useful as a "word processor" than as a source of information. It's really good at reformatting text, changing it into questions/answers, summarizing it, etc. For example, I'm a huge fan of Anki and I've found ChatGTP to be very helpful for generating questions from a block of text and in formatting Cloze cards. But I don't recommend asking it for information directly. I wrote a bit about it here: https://neurotechnicians.com/p/generative-ai-and-anki-part-1-chatgpt https://neurotechnicians.com/p/generative-ai-and-anki-part-1...