This is very useful! I frequently copy the response of one model and ask another to review it and I have seen really good results with that approach.
Can you also include Cursor CLI for the brainstorming? This would allow someone to unlock brainstorming with just one CLI since it allows to use multiple models.
Will do. I was thinking of also making the LLMs configurable across the agents. I saw a post from the founder of openrouter that you can use DeepSeek with Claude code and was thinking of making it possible to use more LLMs across agents
And honestly it seems ironic that a lot of people on HN want Firefox to be used by everyone but don't want Mozilla to add features that the "normies" want.
A lot of people on HN do not want people to use, or want, a lot of features that “normies” want. In part because they think that “normies” desire for those features is based in substantial part on misunderstanding of the benefits and costs of the features.
Calling Nehru "installed" by Mountbatten misses the crucial context of the time. Nehru was the undisputed leader of the Indian National Congress, which had been the primary force behind the independence movement for decades and had overwhelming popular support. Mountbatten's appointment was more of a constitutional formality in the transfer of power, not an act of kingmaking. It's like saying the Chief Justice "installs" a newly elected president.
The same goes for the Aung San Suu Kyi connection. Labeling her a "compliant native" seems to ignore the 15 years she spent under house arrest actively fighting against a military junta. That's a pretty high price to pay for being a supposed puppet.
Not if that might bring the anger of the Empire down on them in full. There's a reason even totalitarian regimes don't apply the same amount of force to all of their political enemies.
It is also a lot of work (and intuitive) to publish all the content on a website and then intentionally make it unpleasant to read just to encourage book sales.
When the presentation of the website has the appearance of being deliberately terrible, then I (for one) am lead to presume that the corresponding book will be just as disdainful.
The place I work at, we replaced our old NLP pipelines with LLMs because they are easier to maintain and reach the same level of accuracy with much less work.
We are not running a call centre ourselves but we are a SaaS offering the services for call centre data analysis.
Thanks for sharing this! I have been trying on and off to run RooCode on a VPS to use it on the go. I tried Code Server but it does not share "sessions".
KASM seems interesting for this. Do share if you write a blog post on setting it up
It’s pretty straightforward through the Linuxserver docker image deployment. I have some notes here re: configuration and package persistence strategy via brew:
Ghostty is gpu accelerated. It’s super fast, and tmux in it is a joy to use. That combined with NeoVim gives me an increadibly smooth dev experience in the terminal, something I had never have with iTerm2 and emacs.
Can you also include Cursor CLI for the brainstorming? This would allow someone to unlock brainstorming with just one CLI since it allows to use multiple models.
reply