ShagilityAgile Data Coach, Author, Agile Data Guides · Your Agile Data Guide

Local models beat panic, maybe #substack #shorts

6 Aug 2026 · 2 min

Watch on YouTube

1:55

In this short conversation, Shagility and Speaker A discuss the growing importance of privacy‑preserving AI tooling, especially local transcription with the open‑weight model SuperWhisper. They argue that while frontier LLMs remain the performance ceiling, local/open‑weight models can deliver a large portion of the value at a fraction of the cost, describing a Pareto‑type trade‑off (≈80% of value for 10‑20% of cost). The dialogue then shifts to organizational questions about who should own and manage the massive amounts of unstructured data (voice, video, etc.) that constitute a company’s “brain.” They speculate whether this responsibility will stay with data teams, migrate to a new CRM‑like function, or be taken on by engineering groups who are already building layered data‑engineering and analysis pipelines. The speakers note a cultural trend where engineers, motivated by the belief that “knowledge is power,” spin up their own solutions when existing teams show little interest, leading to ad‑hoc architectures that blend data engineering, analysis, and application layers.

Chapters

I think where in theory, somebody building tooling for that would be useful. Yeah, I totally agree. I think privacy, I think people underestimate privacy like to like a crazy extent. Right. I still think like I use a tool called SuperWhisper, um, and because I do local, local transcriptions most of the time, because I am still conscious about, about that. Um, I still think that local LLMs and open weight LLMs will have a chance. They're never going to be as good, I think, as Frontier, but I think the evidence is starting to show that you can get like I don't know, 80% of the value for 10%, 20% of the cost. It's kind of, you know, Pareto. And I really am starting to, we're starting to see it. I do like how everybody was like, you can't use a frontier model to train your cheap model. That should be illegal. And I was like, isn't that like you guys using all the books and videos and copying it and complaining and whatever?
Yeah. Yeah. I think the other thing that's going to be interesting though is that you know, we've, we've talked about data teams dealing with unstructured data before, you know, voice, video, all that kind of stuff. I think now it's going to become a reality. It'd be interesting to see whether that, that corporate brain, whether that sits in the data team's remit, in the data domain, or we get a new equivalent of the CRM team who, uh, who end up doing that. So where does it live?
I don't, I don't, I don't know. I mean, right now I'm like, I'm seeing it very clearly in engineering teams potentially. And then a lot of people are spinning them up themselves. And then they're like, they're trying, everybody's like, I've seen this a lot of places where like an engineering person is like, understands that knowledge is power, like most data people. And it's like, well, nobody's interested in doing this, so I'm just going to start my own thing, right? And then they just, they do that and they start to do a layered architecture and you're like, oh, you mean you're doing data engineering and data analysis and then you're creating— I was like, oh, welcome to my world.