fridder

I also wonder what this moment will mean for chip design going forward. The fact that we are so constrained on the GPU side along with the success of unified memory with apple silicon does make me wonder what chips in 3-5 years will look like.

Sony has so often footgunned itself it is comical. Occasionally great engineering hobbled by terrible business choices

or libghostty

Royalroad has been a great place for me

I did find a plugin at least

I wonder if it works with jj workspaces as well. I guess I have something new to tryout today

Take a look at zellij as well

I think it is rather disgusting to equate raising humans and training an LLM

Well, I was trying to be a little charitable

His most innovative move was acquisition

Didn’t he famously trash folks over 30 years back? He has really done nothing innovative outside of ads.

Exactly. With the broad layoffs some companies do, you learn the company doesn’t value you, so why should you value the company?

I think you are spot on. Also if you are actively listening you may learn that the fault is in how you are communicating your ideas instead of the ideas themselves being flawed

really is too bad FreeBSD couldn't be an alternative

there has been a decoupling of productivity gains and wage increases for some time. Not completely disconnected yet but still

We need a SETI@Home but for model training

I wonder if it will be better at building TUI's. It has been absolutely abysmal at interacting with them and building them

One problem: folks that are on blood thinners shouldn't supplement with vitamin k

I personally like how I barely had to configure it, how nerd fonts just worked, and how nicely it renders text

Cost, time, and for things like CT's, trying to limit your radiation exposure

I'm also a Doom user, albeit a new one. What is making you consider the switch?

Not OP but if you run this from the CLI it does make the ordering make a little more sense

Not too shabby. I like the regular Qwen but prompt prefill on my m1max is slow as hell

Is there a local harness designed around the local model use case that is claude code like? Opencode has been problematic at times, pi works for one off for me but not back and forth conversations with the LLM. Considering I only use Qwen or Gemma models I'm close to just writing my own at this point

Yeah. I usually do this by telling it to be adversarial and find gaps and holes. Not fool proof but it does seem to increase the quality. It has helped when using local models in particular.

AKA Woz vs Jobs

dirvish is nice

It truly is the SOTA for local inference on mac. Even when there are regressions the dev(s) are insanely responsive. It is the most impressive opensource project I've seen in a awhile

My pushback isn’t the credit part it is when they try to spring things on you directly instead of going through normal channels. A lack of planning on someone’s part is not automatically an emergency on my part. I see this far more often than credit stealing

The more I think about it and your comment the more I wish it was local only. It could be useful to analyze your editing habits and interactions with AI but I want that for my own benefit not random coworkers