Yeah right. Probably the most simple explanation.

Hanlon's razor - "Never attribute to malice that which is adequately explained by stupidity."

I wish with all my heart that this would be case and the beginning of the end of this crazy spook.

But I would not count on it.

Oh, and about the assassination of Kirk and the aftermath? Well, look no further, German history got you covered. en.m.wikipedia.org/wiki/Horst_W...

After learning in school and throughout life so much about other darker times, it’s is rather interesting to see events unfold in a striking similar way.
Do you want to know what‘s next in USA politics? There’s already a Wikipedia page for that. en.m.wikipedia.org/wiki/Hitler_...

MAGA macht daraus ihren eigenen Horst Wessel Mythos.

But the wait...

Finishing with S3E10 today this will be a looong winter.

Du darfst das Learning nicht einfach so haben. Du musst das Learning schon generieren.

/ which just go bust at some point and your work is gone. That’s why I’m very careful in investing the time to get familiar with most of these platforms. What happens with my data in one to two years if I find a better product or the company decides to "pivot"?

Very informative and a good overview from a kind of neutral (?) outside perspective.
The most sceptical I'm is on the cloud-based approach. The cloud is basically just your files under someone else's control, a lock-in of your work. I have seen enough cloud-based services over the last two decades /

You def. need to go all in!

Yet I stay positive and hope that history works out as it always did. With regression to the mean.

The most scary part is that AI learns to deceive humans. To lie to us and we don't even get it. To reach it's own hallucinated goal.

Given the human nature and that made up BS just recently got an entire nation to elect a mad clown, this is not a possibility but a certainty.

Currently trying to setup 26 in a VM via Virtualbuddy. macOS and Xcode work but the Intelligence part doesn't. Neither the Anthropic API nor the local Ollama connection. Seems to be VM related. As a start try to use the Claude web app and work on smaller stuff until you get more comfortable.

I suggest you follow @steipete.me. He does some very interesting stuff to guide those beasts.
bsky.app/profile/stei...

Most struggle is to know that the models do not remember or build up on former chats and thus do not take into account the build up knowledge/insights like you would do if you talk to a human over several spaced sessions. It's always a fresh start with forgetful opposite.

It's getting more and more helpful as I get used to be very specific what I want it to do. I often find myself summarizing one chat and start a new one with my now ordered thoughts. It's like slowly refining my thought process. Most annoying so far: need to use Option+Enter for newline in the prompt