vektormemory

I'm liking all of the remixes.

We built a combined books/video version as well in our app library.

https://medium.com/@vektormemory/vektor-2-0-4-hacker-news-bo...

Great idea, I do a lot of work testing small LLM's in this space and agree if you add some guardrails and train a small embedder, you won't need these large models, as your info is static, you can ship it all as one.

Or try these with your trained embedder: https://ollama.com/library/minicpm-v4.6 https://ollama.com/library/bonsai-2

If only there was a company that didn't use ads, protected people's privacy, didn't sell their data, and had local data sovereignty?

Full article with desk routing benchmarks

We have to test many different api's and models as we have over 600+ in our harness now, Sonnet 5.5 via cloud code provides the highest quality results.

Over the last few years we have added mcp tools, memory graph, md files and skills, automated everything to the point that on desktop the system runs itself, besides proceed/approved/explain.

We have 100's of PRs/jobs/projects and Claude builds down the list automated and only asks a few questions point form if needed.

Very impressive now, compared to 6 months ago.

You can view it if interested: www.vekmem.com

Full unpaywalled article

A full unpaywalled article is provided.

Those close x's are exactly how the govt would make them, peak design...

80% Sonnet 5.5, Opus 5.5 to finish the last 20% cleaning up the fine details.

Gemini 3.1 pro as a nutty professor, researcher, and design verifier.

My custom orchestrator for all local LLM work: https://vektormemory.com/vektor

Can someone remove the extra LLM and just have an embedder do the classifier work?

It's turning into pimp my llm...

Massive updates ahoy...

After reading through the Anthropic threat intel report, you should read it too.

One of the best free reports I have read in a long time, actually, was it written by Claude Fable… I don’t know?

Countering misuse of AI: September 2026 / Anthropic Case studies from threat actors disrupted between December 2025 and August 2026 across seven areas of harm, from cyber… www.anthropic.com

It didn’t scare the crap out of me; it confirmed my already jaded suspicions.

The nature of who can carry out sophisticated cyber attacks has fundamentally changed, but not the attacks themselves.

The report’s core finding is that sophistication has stopped being a reliable signal of who is behind an operation, as AI has collapsed the gap between well-resourced state actors and lone wolf individuals.

A hacktivist with stolen API keys, a financially motivated dark crew, and a state-nexus espionage operator all used the same playbook: agentic AI bots running multi-victim campaigns that previously required entire teams.

I get a real sense that we are in a “deep situation” caused by a very small percentile of the population using the tools made by teams that have way too much VC fun money and are not really aware of who is using their tools; the age of agentic bots is much further along than anyone can conceive.

The oroborous of bot slop continues. It’s going to be a great ride for cybersecurity, though, 1000x the problems to solve.

And the govts are once again asleep at the wheel, too worried about GDP numbers and the other 100 calamities they caused this week instead of focusing on secure spaces, quality infrastructure, education, and medicine.

Why buy the cow when the milk is free? They should back an open-source projects?

All countries should back open source projects with their combined funds and data centres.

This information to train the LLM's is already in their library resources, they stole it already from lib gen.

I like to synthesize my unconscious thoughts whilst my headphones are on but unplugged whilst on the playa man.

I love synthesizing the unconscious mind while shit in silence.

I never push the play button on my unplugged headphones whilst on the playa. Lol