I did it full-time for 3 months and it was exhausting. I found it useful for learning a new framework (Ruby on Rails at that time) but in general I did not find it compatible with how I think and work.
It's an interesting article. To me it reads like someone who is not confident in their writing, but is very interested in getting their idea across, iterated with an LLM to produce it.
It set off all my LLM writing triggers and I noped out after about 3 sentences, because I'm designing/coding with AI all day and I'm 100% sick of its writing. I don't really care if it was written by an LLM or a person who writes in LLM style.
HOWEVER, the surprising fix for me was to ask a number of LLMs to summarize it. I omitted the "is this AI slop?" parts when I pasted it in. Like I said, I don't care about who wrote it, but I hated the writing style.
The AI summaries were quite readable.
They converged at a high level on the summary, but with a good bit of variance in nuance. I had a good follow up discussion with one of the LLMs, so I'm glad I tried it.
The way they worded that question is bad and, as you say, the outbox pattern does not transactionally update the queue itself. The outbox pattern is nevertheless very useful.
Re the spikejapan blog: The author’s About page includes this line which describes my experience of the article very accurately, having bailed after a couple of minutes:
“It's a species of anti-blog, as there is no way that you'll get through a post if you suffer from any kind of attention-deficit disorder; even then, you may need a strong cup of coffee and an hour to kill.”
reply