Not too bad for a family of four over the course of 7 or 8 years, some devices even older. I do wish there was a better way to recycle this stuff though.

I'm on Spain's IRS (Agencia Tributaria) web site. I don't know if this is just a really bad translation or someone being facetious but apparently they have a "deprive you area". This is not a joke or fake. This is real!

Giraffes are still a thing in modern AI. In the attached transcript, I was trying to get ChatGPT to remind me of the word gobsmacked. I gave up after giraffe. @janelleshane.com

I tell everyone I know: transitions are always the hardest part (as evidenced by this video - watch to the end)

This is just getting better and better… I had it create this image based on the 4th conversation starter…

First time off leash!!!

What is RAG, LLM & “context window”?

As a linguist and a seasoned developer, I try to use accessible language. Technical language is often exclusionary, creates barriers, makes it hard to relate and connect. Sometimes, however, it is impossible to escape neologisms and still be perspicuous…

And finally, here is a link to a slide presentation I created to explain the lessons I learned about trying to fine tune on construction machinery: https://buff.ly/4fi64hJ

This was a great project for interns because it covered three common programming tasks: 1) uploading a file, 2) interacting with a custom API, and 3) geolocating the user. The interns learned a LOT more along the way and seemed to get a lot out of the experience[3].

https://buff.ly/3CvmHrF

Computer vision was my first exposure to ML. As part of a FastAI course, I “fine tuned” ResNet32 to recognize plant machinery (forklifts, 180º excavators, dump trucks, etc.). Fine tuning is easier than training a model that results in equally useful outcomes.

#lightweightAI_SecretSource

The underestimated power of custom GPTs

IMHO, there is one AI tool that is underrated and overlooked. For the rest of the posts in this series, I’ll be describing the different ways we’re helping people solve problems using it.

#lightweightAI_SecretSource

5 years ago I saw an early version of GPT. Although the output was essentially useless and very obviously produced by a machine, it was clear how much progress had been made and even clearer what the future would hold. It reminded me very much of the very early years of the web (1993-95).