mattew

6 ms·

1.
by mattew 3mo ago
I mostly use Opus for skill development. Once I have a solid skill implementation with a good eval, I move ongoing execution to a cheaper model running under Goose. With the eval you can see if the cheaper model works well enough.
2.
by mattew 5mo ago
It was odd that there was no mention of the forced adaptive reasoning in the article. My guess is they don't have enough compute to do anything else here.
3.
The Lovable Shopify Integration(lovable.dev)
2 points by mattew 11mo ago|1 comments
4.
by mattew 11mo ago
Looks like Lovable is expanding out to platform specific development with their just announced Shopify Integration
5.
by mattew 1y ago
I’ve only played with Junie and Aider so far and like the approach the Junie agent takes of reading the code to understand it vs the Aider approach of using the repo map to understand the codebase.
6.
by mattew 2y ago
I saw something mentioned of voice control in aider.chat
7.
by mattew 2y ago
Thanks!
8.
by mattew 2y ago
Thanks for sharing. Can you add a simple example that is more concrete and less abstrwct to the docs? It would be easier for people to understand if you did.
9.
by mattew 2y ago
This makes sense. Can you clarify what you mean by journal management in this context?
10.
by mattew 2y ago
I’ve found the OpenAI assistants API not really up to snuff in terms of predictable behavior yet. That said, I’m very bullish on agents overall though and expect that once they get their assistants behaving a bit more predictably we will se
11.
by mattew 3y ago
Good stuff. How does this compare to Instructor? I’ve been using this extensively https://jxnl.github.io/instructor/
12.
by mattew 3y ago
This right here is actually the coolest part about developing with LLMs. You just changed the functionality with a sentence rather than a config file, or writing code. It’s great to be able to break out functionalty into things that can be
13.
by mattew 3y ago
I also think of pudding guy every time something like this comes up.
14.
by mattew 3y ago
That’s right. It was selective availability and you used differential post correction to clean the data up and get accurate locations for the data you were capturing. Thanks for the correction!
15.
by mattew 3y ago
The interesting thing about GPS back then was that the location data you got back were randomly slightly incorrect every time you got a reading. I’m pretty sure this was so it wasn’t useful for military purposes. I think it was called diffe
16.
by mattew 3y ago
Don’t forget classification. Even the zero shot classification can be quite decent, but put a solid prompt together, or fine tune a set of examples and it’s way easier with an LLM than previous methods in my experience.
17.
by mattew 3y ago
Maybe I’ve been using PostgreSQL too long but when faced with the choice of adding vector support to PostgreSQL or using a new technology, my first choice was to start with the PostgreSQL addition. I’m not criticizing the specialized case f
18.
by mattew 3y ago
I’m fine if you have a “contact us” for a true enterprise plan, but companies that have actual pricing plans should post them. I way prefer to buy from SaaS companies that display their pricing up front rather than trying to get me on the p
19.
by mattew 3y ago
Haha - I posted this one too. Always good for a laugh reading the reviews for this one.
20.
by mattew 3y ago
The 55 gallon drum of lube on Amazon always cracks me up. The reviews are classic. https://www.amazon.com/Passion-Lubes-Natural-Water-Based-Lub...
21.
by mattew 3y ago
What I find entertaining/confounding is how difficult the abstracts to these new AI papers are to understand. It feels like academia is pushing this style, so it’s hard to blame the authors since they have to play the game. For referen
22.
by mattew 3y ago
I’ve thought recently about using LLMs as part of the data build process or ML projects. This looks like a decent start in that direction by providing access to an LLM from scikit
23.
Scikit-LLM: Sklearn Meets Large Language Models(medium.com)
6 points by mattew 3y ago|1 comments
24.
by mattew 3y ago
That blew my mind as well. I would have thought a much larger dataset would have been required.
25.
by mattew 3y ago
I’m thinking about data for an LLM from a different angle for ecommerce analytics. 1. Build up a set of instructions on how to interpret a particular type of data in the system prompt 2. Run a set of analysis and instead of outputting the d
26.
by mattew 11y ago
I'm using the server monitoring with StatsD integration pretty heavily and am pretty happy with it so far. I'm guessing the app monitoring is probably pretty well done based on how nicely the server monitoring works.
27.
by mattew 12y ago
Are there any good online or real world assessment tools out there to identify people who may be suffering from depression / more predisposed to be depressed?
28.
by mattew 13y ago
We've been using MapBox and are pretty happy with it. They are filling a much needed niche for businesses that are not big enough to be able to afford the $10k google license. We have had good success leveraging their tools for non p
29.
by mattew 13y ago
Down for me and another user.
30.
by mattew 15y ago
"Currently, Netflix uses a service called "Chaos Monkey" to simulate service failure. Basically, Chaos Monkey is a service that kills other services. We run this service because we want engineering teams to be used to a constant level of fa

More ›