I’m actually surprised at how well LLMs handle this. I too expected them to be literal, unable to understand a prompt where I tell it to “make it cooler” The LLMs handle it just fine. They make assumptions and try to give you something. You tell it “put the button Read more
3 years ago I played around with dalle and tried to make a nonfiction book about aloe Vera. The image quality was horrible but I managed to whip up a bunch of usable pictures and write a short book. Instead of trying to get exactly what I wanted, I went Read more
To elaborate on this. The only way a central authority can fairly distribute resources and work is by being near omniscient. I predict a truly sci-fi dystopia where everyone is monitored, everything is fed to a benevolent AI with instructions to optimize everything. And people will get told for example. Read more
AIs need to get their own sensory organs. They’re brains in a jar. They hallucinate because that’s what brains in a jar do without external sensory input (literally what happens to human brains in a sensory deprivation tank). That will become a problem as they become more and more complex Read more
One of the things that really annoys me with LLMs is that they don’t tell you when they can’t do something and they simply bullshit their way around it. My father wanted to do some work and asked me to grab the medical advice from a YouTube video. Chatgpt didn’t Read more
AI video generation is quite slow and that’s why I don’t really like playing with it. It’s a slot machine with no fine control. I prefer static images. We used to wait half a minute or more for image generation, now it takes a couple of seconds. Allows for experimentation Read more
Sudowrite now has a mobile app so you can fiddle with your book while lying in bed like I am. Also they added scrivener import support, so you can easily go and work on existing old projects. Finish that book!
Written. Published. Moving on to the next book. Why would I ever stop using AI?
Made this cover art a while ago in Midjourney. it’s epic but it’s a mess on the details. Experimented with grok and chatgpt, finally gemini managed to do the fixes I wanted for the hands and faces. Does it use nano banana automatically now? There wasn’t a selection anywhere. Anyway. Read more
I finally got in the mood to revisit my most successful experiment, the Level Up Harem series. I had a draft of the book already written, sat down with Claude opus 4.6 I think inside cursor and spent over 50 dollars in credits to complete it. Ran a lot of Read more
