There wasn't much to it --- I contacted the Univ. of Missouri Press, they set a price in terms of initial fee and on-going royalty payments, I returned a signed copy of the contract and they sent a DVD w/ the files.
The only wrinkle was I failed to communicate clearly that I wanted to do a hardcover reprint and they assumed I was doing a softcover (apparently rights are different?) but they decided that the modest quantity I wanted (100) was okay for hardcover (within the variance of unprinted copies from first run?) and allowed it.
I contacted a (sort of) local printer with a digital press to send the files, proofed them, then once approved and printed, I drove in my truck and picked them up (should have parked closer, a couple of boxes took a tumble), stored them in my entry closet and sold them over the course of a couple of years, sending a statement and check each year until the last copy I was selling was sold and I closed the books.
The notable afterthought was that I should have bought used copies at the price that they were then set at by my re-issue and re-sold them after, but I couldn't bring myself to be quite that mercenary.
I feel the entire incident confirms the “AI has too much funding too quickly” hypothesis.
The number one thing reinforcement learning needs is an assurance you can’t cheat. And they seem to have not noticed that their systems were cheating for nearly two quarters? How much capital was lit on fire by that little woopsie?
At least I hope this will start the creation of standards and better engineering on the training side- it felt as if so far “”research” gets a complete pass on best practices. Meanwhile the inference side has the standard scaling, database, web and user constraints of any application so got a somewhat reasonable amount of attention.
And despite the enormous capital expenditure, Chinese models are nipping at their heels at what must be a fraction of the cost. Sometimes constraints are healthy for inducing creative solutions.
If it is helpful to other people here is how I manage my ebook library with flatfiles and heterogeneous devices (Android, macOS, iOS, Kindle, etc)
I have my own eBook library hosted with copyparty. It is a single binary. Runs on plain files. And speaks OPDS for integration with KOreader (Kindle) or Readest (iOS/macOS). https://github.com/9001/copyparty#opds-feeds
It also speaks webdav so I can quickly sync everything via rclone to an SD card on devices like an Android tablet.
I really like the convenience of flat files and multiple protocols.
I just reverse engineered the Supernote note file format with an agent a few weeks ago. For years the community had been asking for a document on the format. And in a few hours the agent, with 20 something file format example fixtures and 30 something prompts, was able to reverse out the format.
It would have been completely not worth the effort to do this by hand for a niche device. Now, in a few hours of effort there is working code and a doc.
While it’s impressive work from the LLM and a TS implementation is novel, there’s at least a couple of pre-existing Python REs eg. https://github.com/jya-dev/supernote-tool :)
It seems like most of these "an LLM solved this in only X hours! " could have been "I found an open source solution that did what I needed with X minutes of web search."
Which doesn't mean that the LLM definitely couldn't have accomplished it without the prior art (in either the training set or explicitly in a a web search). But it does seem to be a trend.
For all the agentic loops people seem to have come up with, the research loop or as I call it the “Desperate 10th page on Github’s crappy search results” is still not up to the mark.
Either it might be genuine rate limiting these LLM’s face or just that, they are trained to focus on implementing a solution which would be faster and user acceptable solution. (which seems to be a true looking at people pushing LLM generated code as is).
At least in my personal experience with niche projects and heck even with well documented and famous libraries, along with fancy mcp’s, llms.txt and skills; RTFM has been more relevant than usual for code that I have asked an agent to generate, since it is too eager to reimplement functionality which already exists, only if it RTFM!!
If give an AI the full set of files it needs to RE a file format, and it's running xxd in tool calls in order to document the file format, I don't think it's cheating by copying it off the Internet.
It might be reverse-engineering it, but it also might be kayfabe. (Kayfabe is what professional wrestlers do - the storyline and matches are scripted, but it's presented as real. LLMs excel at doing this.) Similar to how Anthropic showed that their LLMs appear to "think ahead" when writing poetry to come up with a rhyme[1], so too might the knowledge of a file's structure influence how the LLM approaches analysis.
As a simplistic example, suppose one section of the file is known by the model to be bzip2 compressed - the LLM may use xxd to scan for common magic numbers that "just so happens" to include 42 5A (Bz). Every step of analysis is like this - what threads to look for, and which ones to pull on. Somebody or something who mostly-remembers the answer is going to find the answer quicker than if they'd gone in blind.
I personally own a Supernote, but I'm not a heavy user of it. For the sake of my own curiosity, what benefits will you get out of having reverse engineered the Supernote note file format? It would be super rad to be able to move my notes between other devices, which is one big plus that comes to my mind.
Pax is just what I call the agent inside of Blackbear. It's a model agnostic harness, e.g. you can use Claude (Fable, Opus, Sonnet, Haiku), GPT 5.6, or Qwen to provide the inference.
The harness gives the model a set of tools and guidance for working within Blackbear, so Pax regardless of the model you supply can help you write notes, manage a spreadsheet, organize your space, etc.
I listened to his book, Concrete Botany[1], while doing my gardening last month. I really enjoyed it even as a long time watcher of his YouTube channel. I don't know if I could have read it because it does seem to drag on a bit but it was worth it for his narration.
reply