ChatGPT as muse, not oracle

What’s ChatGPT really bad at? Being counterintuitive and surprising us. Large language models (LLMs) are trained on the mass output of humanity to optimize for what is expected. As a result, ChatGPT routinely let’s us down when it comes to getting something we didn’t expect. If you start paying attention to the music, books, movies, speeches, essay... See more

I feel like Eno is exploring this question of what creativity is and what that process looks like for different people. The way OpenAI talks about what a large language model generates is, by definition, incurious about creativity. It’s like, what if we just spit stuff out and you have no idea where it came from?

ChatGPT didn’t say much that was useful at all. But—as the students reflected in their conversation after reading ChatGPT’s analysis—neither did they. One student noted that they were just saying “formulaic, buzzwordy stuff” rather than tackling the issue with fresh thinking. They were unnerved by how closely the empty shine of ChatGPT’s answer mir... See more

ChatGPT is not scraping the web and returning a response. Its AI is “thinking” and creating content it believes answers your question. Whether this is the job of a search engine is another matter, but it is undeniably useful for circumstances in which original content like a code snippet, blog post, email draft, or social media post is valuable.&nb

... See more

When I hear artists and AI evangelists argue about whether people who generate text using ChatGPT or images using Midjourney are truly creative, I wonder about the question behind the question.

In design, AI inspires me no more than elevator music or a business presentation. However, it often shows me what I do not want. When I ask Chat-GPT how I could better phrase something, I almost always get the most uninteresting, boring, often meaningless answer. As an author that writes to say something meaningful, get upset about this, and in res... See more