h-bradio

6 ms·

1. 1.▲
   by h-bradio 1y ago

   Here's a great piece from Ian Bogost in The Atlantic on this topic (gift link): https://www.theatlantic.com/technology/archive/2015/11/progr... (Please focus on his discussion of PE licensure, not o
2. 2.▲
   Local Live Captioning with Whisper.cpp and FFmpeg(brady.fyi)

   2 points by h-bradio 1y ago|1 comments
3. 3.▲
   by h-bradio 1y ago

   This is a fun little set of commands to run live-captioning on my Macbook! Nothing serious. A tutorial and demo video are in the link. I saw a post on HN that introduced me to Whisper.cpp. I’ve been excited to see so many AI tools start to
4. 4.▲
   by h-bradio 1y ago

   Here is how my prompt ended up! https://gist.github.com/hbradio/2f504c3fdb6f7113181b2d8c6862... I just asked an LLM to make it similar to a working Qwen prompt. To make the LLMs able to use tools, I had to configure th
5. 5.▲
   by h-bradio 1y ago

   Thanks so much for this! I updated LM Studio, and it picked up the mlx-lm update required. After a small tweak to tool-calling in the prompt, it works great with Zed!
6. 6.▲
   by h-bradio 1y ago

   OP here! Going into it, I definitely agreed and thought that easily fact-checkable claims would be the minority. But as I worked, I found that many of his claims were "this paper says this". So checking the claim was as simple as
7. 7.▲
   by h-bradio 1y ago

   OP here! Thanks for calling out this important point. As I fact-checked each claim, I was surprised at how many of the checks were "does the paper he's citing say what he says it does?" You can see them here: https:/&#x
8. 8.▲
   by h-bradio 1y ago

   OP here -- thanks for your reply! You're exactly right! I included the NYT/PolitiFact graph at the top as an example of that problem. In the second half of the post, I propose what I think could work a little better (sampling comp
9. 9.▲
   Fighting Brandolini's Law with Sampling(brady.fyi)

   30 points by h-bradio 1y ago|16 comments