Viewing profile — verdverm

with how overton is shifting, might be in the cards, watch polymarket for insider activity but seriously, it should require an executive order, unless you stop believing in experts…

pardon their French? or is the limit gallows?

Their value is in turning away people by destroying the very product that built the empire, good imo, bad for investors. I hear Google is a slop fest internally, citing people who …

I can only imagine the company culture that enables and encourages creating gambling opportunities for this and the other troubling worldly events

at this point it seems like everyone feels the need to have a story like this, the hype factory must grow

I'm running ~6 small models on an oem spark to back the various features of my personal knowledge base tool. A couple of LLMs and some non-llms too https://github.com/verdverm/gmd …

This is a good bridge to something new eventually. There is Timoni, which was a CUE based attempt to replace Helm, but imo, it stayed too true to helm concepts and missed the oppor…

re: coordinated regulatory capture with the oil giants, we call them a cartel, which I postulate is also accurate for the hyperscalers and big ai

I suspect this is part of why people are cheering for the Chinese. It's nice to see frontier competition coming from another country. And they are making the open weights a way of …

Have they aimed for 86 uptime?

This seems persistent and endemic, have they said anything about why it's been so bad and what they are going to doing to improve going forward?

oooh that smell, can't you smell that smell? I've architected our CI pipeline (newer folks are doing the CD portion), it's a monorepo tamed with CUE and Dagger, runs on Jenkins tha…

Yeah, for sure don't use it. It's complete slop. The help strings do not match the real args/flags. We went back to calling the API directly, writing scripts for the agent, then ta…

likewise! I cannot understand the field of devops willingness not only to endure this, but find double templated yaml (Argo+helm) appealing. So many of my peers prefer to do what e…

Have you seen the markdown agent instructions they provide in their new agentic `twg` cli? 70k tokens one average, there are more than one... Rovo is the worse Ai I have used, I su…

I'm using CUE (config) + Helm (reconsiliation) This post is also using helm, putting a DSL in front because yaml sucks and templated yaml is insane. The problem here is the DSL, de…

> Analysts say such weapons – which cost over $1 million each – would also be important in any conflict with China. Said without context, there are not very many places close to Ch…

https://vllm.ai/blog/2026-07-27-k3 Original full blog on their own domain

Yes, Ai can code, kind of, but it cannot make good design

[dead]

Trash bin? ~1M LOC for what? If you cannot review or test it, then why would you trust it? This is why it's better to stay human in the loop during the process. Sounds like you don…

Looks like you are using Ai for this kind of thing based on your submission history. I don't think people are looking for proxies to design slop

It's best not to go back and forth with someone like that. They did not answer my only question in their pass-by commentary, so I feel no need to respond to the quotes they pulled …

I'm not sure how much we can trust software that hasn't been security audited by Ai going forward. Humans are not that good at writing code, we are going to need help that relentle…