Tsarp

I've really liked lancedb for similar use cases. Not just that it is OSS. But Lance treats ANN as a secondary index similar to what turbopuffer v3 does. Rows sit in fragments, and the vector index never moves them.

Lot of these demos tend to squeeze in as many animations and transitions. Its usually hard to take away anything from the video at the end.

I wish more people optimized for learning than "how fancy can i make it look"

Waiting on simonw "Generate an SVG of a pelican riding a bicycle " benchmark to judge this model

Apart from fictional use cases, what is the real use case here? The pricing on some open models are absurdly low for generic tasks. For the privacy conscious it makes sense to run something like a 8-27B on local network and get the work done.

Are there perhaps some industrial or agri use cases?

Kinda weird to generalize "LLM". Every lab, every model is different. Has its own biases, reward functions etc.

One great use of agentic coding is being able to add and very granular tracing instrumentation to help with these sort of optimizations.

Its just far easier to ask an agent to do the first past ( with references to actual code ) and then start to dig in.

A good implementation of something like this is https://deepwiki.com/ for OSS projects.

Lot of this is outdated. Including stuff that made a lot of sense in the past like diataxis and the google guidance.

I've started to build docs structured as a skill. Since every model/harness is trained to handle skills well today. The entire docs of a project or modeled as a skill.

I also have additional frontmatter on the md files. Specifically 2 keys -> when should you read, when you should not read this doc. This with a simple cli to help parse this has made the docs experience much quicker and nicer.

You can also generate svgs, mermaids etc on the fly now when you are really stuck on something.

Ah yes, the usual drop down where we need to confirm we are above 18.

Might have to do with a lot of thought leaders and influencers pushing this.

In my own life/career I have seen that over a long enough horizon that lower shame seems to generally work out for people on the long run if you can back it up with some amount of confidence and competence.

I wish folks spend more time on agentic music genertion instead of a single black box prompt to music.

Similar to coding, where you are have a bunch of tools that can do the drudge work for you (the mundane things in mixing mastering for example), but still retain full creative freedom and expression.

I am not articulating this well, but basically transitioning from IDE->Coding harnesses what is the DAW-> ??

With the Devin subscription even at the 20$ plan, they offered unlimited SWE 1.7 usage. Wondering if they do the same for SWE 2.

"SWE-2 is post-trained from Kimi K3"

sqlite for everything

NVMe drives + Litestream + object storage(S3/R2..). sqlite simplifies things for the entire long tail of apps/services that aren't the Ubers and AirBNBs of the world.

whisper.cpp

Any chance you can benchmark against whisper large and large v3 turbo? These run comfortably on older Macbooks and are still far more accurate in real life dictation compared to even the parakeet models( despite ASR leaderboards) with an RTF < 1.

What harness are you guys running?

+1 Latent Space (and AI Engineering ). I hope it stays like it is. Minimal fluff, great guest list and very relevant.

This really is driving a muscle/super car, or drinking expensive wine. At the end none of specs or tests matter. It is a form of art. If it makes the listener feel better (even if its just psychological) then its probably worth it.

That makes sense when things are mostly stable and it makes little sense for most teams to work outside the norm.

Currently though we are in a world where things change every week, model capabilities, harnesses, pricing etc. Forcing a norm wont work, because there is no such norm.

Awesome to see this. Like a few others here, I hand-rolled (well, Codex-rolled) something similar that works great for me. I keep going back and forth on open-sourcing it, but my hunch is people won't really adopt these kinds of things anyway.

Everyone ends up with a workflow shaped really tightly around how they work, and it's gotten so cheap to just build and evolve your own as the models and harnesses change that picking up someone else's stops making much sense.

I dont have an answer and you are mostly correct. I received some advice based on this that made sense which was to pick the roles in your career that naturally made it easier. Sales, PM, Dev etc and not support, Devops, escalation management, CSM etc.

Ian Rush said it best: "It's best being a striker. Miss five, score the winner, you're a hero. The goalkeeper plays a blinder, lets one in, and he's a villain."

Every place I've worked rewards the firefighter over the person who made sure nothing ever caught fire. And the worst part is the math is obvious to everyone except the people who set the incentives.

Guess its "Opted-in" by default

Nice work. Wish they had picked another name given how popular lance/lancedb is.

Hmm please share more. I have had the max CC sub since it came out. Religiously follow all of Boris/Cats advice but still struggle with it. Meanwhile a really badly written AGENTS.md will still get the work done.

Wondering if enterprises have a modified version of CC that doesnt have to optimize to stop bleeding on fixed cost subscription plans.

The article really does not align with the current sentiment. Everyone with a choice has mostly moved on to codex (ofc in this world all it takes is a model update/harness update to turn things around).

CC is great at a lot of things, but repeatedly misses out reading on crucial parts of the code base, hallucinates on the work that was done and a bunch of other issues.

Would be nice to see if this number dipped from before. International students typically end up paying out of station tuition and is a huge source of income for the univs.

Docker can be small too. In this example I was able to compile a full server (rust binary) and package it in a docker (scratch image base) and the total was < 5MB.

https://github.com/srv1n/kurpod/pkgs/container/kurpod-server

wasm version compiles to under a MB though!