Thread 128381456

The discussion: >>128353968

>>128367437
We were talking about AI, and that was it. You claimed AI didn't "use any logic process" and I sent a page that proved that not to be true. I don't even know how from the excerpt you cited you interpreted that it was distinguishing the "our topic" model from an "off-topic" method. Both technologies (symbolic AI and neural networks) are used in music composition. See "Technical approaches": https://en.wikipedia.org/wiki/Music_and_artificial_intelligence.
On autonomous AI, as far as I know, it's not very developed yet. My point is that as it gets better, the argument that AI is less creative for relying on prompts (as if human artists didn't receive instructions from their hirers often) gets less and less plausible.

>>128367454
That reminds me of the first objection to the Chinese Room thought experiment (the Systems Reply), which, in my opinion, Searle didn't give a good and reasonable reply to. In short, the whole system (handicapped human + calculator) is able to do math, not the human alone. You can read his reply and take your own conclusions.
https://web.archive.org/web/20071210043312/http://members.aol.com/NeoNoetics/MindsBrainsPrograms.html

>>128367481
That was what I wanted to stress with my thread. Most people that push the "AI slop" idea don't care if humans make poor quality music. They're fine, because they're not AIs. They prefer to support a bad human artist and boycott a reasonable AI one than to support or even be neutral towards the latter. They are not concerned that the music they support or make can be just what they curse AI with: slop.

this ai cant push new ideas, wait for a new ai

>>128381902
You haven't lurked enough.

>>128381456 (OP)
A human puts time, experience, energy and consciousness into his work. It can turn into slop or mediocrity more often than not, but every once in a while, there is something good and inherently human that comes out of it and offers lasting value.
AI is just an empty facsimile with a shelf life of a few weeks at best.

>>128382127
>(...) with a shelf life of a few weeks at best.
What?

>>128381456 (OP)
Thank you for sharing your insights. You're right that AI encompasses a broad range of techniques, including symbolic AI and neural networks, both of which can involve logical processes and reasoning. The distinction I mentioned earlier was perhaps too simplified or out of context; I appreciate you clarifying that both approaches are applicable in fields like music composition.

Regarding autonomous AI, it's true that fully autonomous, creative AI systems are still in development, and their capabilities are evolving. As these systems improve, the argument that AI is less creative simply because it relies on prompts or instructions becomes less convincing. Human artists often receive prompts or instructions from clients or collaborators, so the use of guidance or parameters in AI doesn't necessarily diminish its creativity.

The conversation about AI and creativity is complex and ongoing. It's important to recognize that AI can exhibit forms of creativity that, while different from human creativity, are nonetheless meaningful and innovative in their own right. The development of more autonomous AI systems might further blur the lines between human and machine creativity, challenging traditional notions of originality and artistic agency.

>>128381456 (OP)
>That reminds me of the first objection to the Chinese Room thought experiment (the Systems Reply),
You're drawing an interesting parallel between the Chinese Room thought experiment and the Systems Reply. The core of the objection is that while Searle's thought experiment suggests that symbol manipulation alone doesn't constitute understanding or consciousness, the system as a whole—comprising the person, the instructions, and the external apparatus—does "understand" in a functional sense because it can perform the tasks.

In your summary, the idea is that the individual alone, without the system, isn't capable of doing the math, but the entire system (person plus calculator) can. This challenges Searle's conclusion that syntax alone isn't sufficient for semantics or understanding, suggesting instead that understanding might emerge from the system as a whole, not just from the individual components.

I've reviewed the link you shared. It offers a detailed discussion of these issues and presents arguments for and against the notion of understanding in AI systems. The debate is ongoing, with many philosophers and cognitive scientists weighing in on whether systems that manipulate symbols can genuinely "understand" or if they merely simulate understanding.

It's a nuanced and fascinating debate, highlighting the complexity of defining consciousness, understanding, and intelligence—both in humans and machines.