bbischof

This is great; it's honestly so f'in annoying working with documents with multiple users AND LLMs

Data visualizations and live dashboards of people tweeting that '_ is dead'

Label studio is fine if it covers your need, but in many cases the core opportunity in an eval interface is fitting in with the SME’s workflow or current tech stack.

If label studio looks like what they can use, it’s fine. If not, a day of vibecoding is worth the effort to make your partners with special knowledge comfortable.

10x speed up; what more do you want

Hello, it’s Bryan, an author on this piece.

I’d you’re interested in using one of the LLM-applications I have in prod, check out https://hex.tech/product/magic-ai/ It has a free limit every month to give it a try and see how you like it. If you have feedback after using it, we’re always very interested to hear from users.

As far as fine-tuning in particular, our consensus is that there are easier options first. I personally have fine-tuned gpt models since 2022; here’s a silly post I wrote about it on gpt 2: https://wandb.ai/wandb/fc-bot/reports/Accelerating-ML-Conten...

Hello, it’s Bryan, an author on this piece.

I’d you’re interested in using one of the LLM-applications I have in prod, check out https://hex.tech/product/magic-ai/ It has a free limit every month to give it a try and see how you like it. If you have feedback after using it, we’re always very interested to hear from users.

We allude to 2 when talking about using explanations first, but I totally agree. One minor comment is explanations after can sometimes be useful for understanding how the model came to a particular generation during post-hoc evals.

Point 1 is also a good callout. I added something on this for llm judge but it’s relevant more broadly.

Bryan here, one of the authors.

Sure. Ultimately, you want to use KG to increase your ability to do great retrieval.

Why do graphs help with retrieval? Well, don’t overlook the classic pageant example: graphs provide signal about the interconnectivity of the docs.

Also, sometimes the graph itself are a kind of object you want to retrieve over.

Super interesting. I haven’t seen much chain support tech for ts folk which is a hard requirement for more production applications.

System 1 & 2 as it applies to AGI

`AI generated content` probably

SEO is dead, or it's so back, I have no idea.

(not OC)

Probably this is the article you saw https://link.medium.com/bf81IpGM8yb

Hehe, I had the same question: https://news.ycombinator.com/item?id=33812666

Interesting! I do contribute to OSM due to my exploration but I didn’t know about this. Thanks for the reference.

I’m really into this gps game: https://wandrer.earth (I’m not the creator, just a fan)

It encourages you to walk/run/bike all the roads in a town/city/county and it tracks the percentage.

It’s a fun and creative way to go see parts of your area you wouldn’t otherwise. I’ve used it to walk 97% of the miles in Berkeley and seen so many funky little areas. I also really enjoy how much it has increased my walking over the last few years.

My favorite line in this wonderful paper is:

The database is just filled with garbage that is unusable for this project: Fonts that are completely illegible, fonts that are missing most of their characters, fonts with millions of control points, Comic Sans MS, fonts where every glyph is a drawing of a train, fonts where everything is fine except that just the lowercase r has a width of MAX INT, and so on.

The themed time-intervals suggestion is not unreasonable, but this article is completely devoid of any evidence to suggest its success besides... checks notes Elon Musk running several companies.

I bet they’re not laughing now.

This is sorta true, but not quite. Leland McInnes and John Healy (the creators of UMAP), do in-fact have an amazing paper on HDBSCAN, but it's not inventing it. In their paper, https://arxiv.org/pdf/1705.07321.pdf, they introduce AHDBSCAN which is a great extension of HDBSCAN to dramatically improve it's performance.

Their work is great but just wanted to save people a google in case they were interested.

Like others this returned total nonsense. Not even one remotely close. (I’m a white dude with publicly visible photos online ICYAI).

It’s stupid and irresponsible that they have made this publicly available. Hopefully the fact that it’s terrible will tank it.

It seems this person is not doing due diligence and some of this is factually incorrect. I’ve been seeing corrections but I don’t recommend trusting this.

My future programming language implements this and GO for better debugging, but the downside is the debugger is named “cotton eyed joe”.

Anyone happen to be a backer and can share updates?

ty

Link dead?

See also https://old.reddit.com/r/MachineLearning/comments/bzldq4/r_1...

and https://twitter.com/hardmaru/status/1138678311884738560

I love anki, and I love this use of it, but how do you quickly make cards? Especially on mobile?

Almost anywhere on the internet people will respond to this with basically "no, it's extremely inefficient to learn how to program this way", and they're right.

But.

If you want to learn _other_ things, these books are incredible. Maybe you want to learn deeply the combinatorial ideas behind data structures, or general techniques for squeezing small optimizations out of low level code. Or even maybe you want to see simple data structures used every-which-way to efficiently solve a huge range of mathematical problems. None of this is normal programming per se. I treat these as math books, and love them for it.

Like much of mathematics, it will make you a better programmer, but only accidentally.

Blue Bottle | Senior Full Stack, Senior Rails, Mobile, Senior Data | Oakland, California | Remote OK | Full time

Blue Bottle Coffee is hiring for several Senior developer roles to support our E-commerce platform, Internal Tools, Data Infrastructure, and Mobile projects.

Blue Bottle is undertaking a variety of greenfield projects in the next year and excited to nearly double the size of our team—huge amount of opportunity for ownership and leadership. (this is why we're currently only hiring at the senior level)

Also, c'mon, the coffee is great...

We are currently using Ruby/Rails, Go, and Python as our primary languages for dev. Moving towards React. Everything is on AWS, mostly using Docker, we also have CI/CD, and use the ol' github.

Some job descriptions: https://jobs.lever.co/bluebottlecoffee?department=Digital&te...

Email me at bryanbischof@bluebottlecoffee.com if ya wanna.