rchaves

7 ms·

1.
by rchaves 1y ago
Hello HN! tl;dr: We built Scenario, an open-source testing library for AI agents. It simulates real conversations with your agent, its code-driven, and lets you assert anything mid-dialogue. Repo: https://github.com/langwatc
2.
by rchaves 1y ago
well I think hype is not bad per se, I'd do it even if not trying to make a buck, it's okay (up to a point) to hype up something so that eventually it finds a problem where it fits well, but yeah, I'm still waiting on this on
3.
by rchaves 1y ago
same here, but I would even avoid "strong arguments" because that's what we all have been doing so far what I want is real use cases, show me real-world production examples from established companies where multi-agent collabo
4.
by rchaves 1y ago
is this multi-agent collaboration though, or is it just a workflow? All examples you listed seem to have pretty deterministic control flows (write then validade, context exceeded, after each response, etc) when I think of multi-agent collab
5.
by rchaves 2y ago
Nah it's just a marketing problem, "GPT" and "ChatGPT" names is the biggest asset OpenAI has, people have expectations so high for GPT-5 that they cannot burn this name unless it's something truly majestic, bor
6.
Discuss HN: Agents are the new object-oriented programming
3 points by rchaves 2y ago|0 comments
7.
LangWatch: A UI for DSPy(langwatch.ai)
5 points by rchaves 2y ago|0 comments
8.
LangWatch: A UI for DSPy(github.com)
5 points by rchaves 2y ago|0 comments
9.
by rchaves 2y ago
Erm, he wrote the article with “you” to invoke the feeling of the reader thinking about their own use case, which I did Different because I ran without good practices before, got more and more messed up over time, grinded to a halt, and it
10.
by rchaves 2y ago
Yes exactly, from the experience he had as Facebook massively scaling up, while all good practices were thrown down the window (except for foundation and critical parts) and extreme go horse php being written as fast as possible, until it w
11.
by rchaves 2y ago
I’ve seen people spending 10 minutes to test things by hand, would have taken them less to write and run a test, specially with AI now When writing test actually makes it faster to code, THEN it’s worth it. You can even throw the tests away
12.
by rchaves 2y ago
Not true. Kent Back’s 3X is a much better take, test and good practices for what is high risk and hard to change, move fast for most of it on the rest to try to find that black swan as soon as possible. Yes, I do feel this time is different
13.
by rchaves 2y ago
yeah I guess base models without built-it CoT are not going away, exactly because you might want to tune it yourself. If DSPy (or similar) evolves to allow the same or similar than OpenAI did with o1, that will be quite powerful, but we sti
14.
by rchaves 3y ago
Nope, the trains are not often late, this is just in Germany
15.
by rchaves 3y ago
If you look closely it actually does give multiple instructions per screenshot! However it cannot get too far, because the screen changes under it. For example when it starts typing a tweet, the tweet box expands and the send button moves,
16.
by rchaves 3y ago
indeed! Ideally I want it to have very real time human-machine feedback, so you can interrupt it in the middle, point at things, then ask new things, and so on, kinda like if there is someone else pairing with you, and you are telling them
17.
by rchaves 3y ago
thanks! I took my inspiration from Vim browser plugin ( https://chromewebstore.google.com/detail/vimium/dbepggeogbai... ), they have a shortcut F that allows you to choose any element on the website to navigate from
18.
by rchaves 3y ago
yes actually, but I only saw it after I've implemented it, I had actually searched for something like that before but I guess Google is worse and worse those days however, I tried self-operating-computer, and it could not find the righ
19.
by rchaves 3y ago
Hey, I was working on something to allow GPT-V to actually do stuff on the screen, click around and type, I tested on my Mac and it’s working pretty well, do you think it would be cool to integrate? https://github.com/rogeri
20.
by rchaves 3y ago
great naming innit
21.
by rchaves 3y ago
Hey there everyone, now that AI can "see" very well with GPT-V, I was wondering if it can interact with a computer like we do, just by looking at it. Well, one of the shortcommings of GPT-V is that it cannot really pinpoint the x,
22.
Show HN: GPT-V and OCR for Screen Control(github.com)
22 points by rchaves 3y ago|10 comments
23.
Ask HN: BlitzJS, RedwoodJS, RefineJS or something else for fullstack JavaScript?
2 points by rchaves 3y ago|4 comments
24.
Convert Python Functions into OpenAI Functions(github.com)
2 points by rchaves 3y ago|0 comments
25.
by rchaves 3y ago
Yeah I never know where memory goes exactly in langchain, it's not exactly clear all the time. But sure, the main insight I remember is this, take a look at their MULTI_PROMPT_ROUTER_TEMPLATE: https://github.com/hwchase
26.
by rchaves 3y ago
hmmm, just had a chat with GPT-4, it didn't seem convinced that ETLs would do well the same things that LiteChain is trying to achieve: https://chat.openai.com/share/88961bd1-8250-45f0-b814-0680ba... I'd be h
27.
by rchaves 3y ago
this inspired me on writing a new section in my project "Prompts on the outside" ( https://github.com/rogeriochaves/litechain#prompts-on-the-ou... )
28.
by rchaves 3y ago
100% this! What is worse is that LangChain hides their prompts away, I had to read the source code and mess with private variables of nested classes just to change a single prompt from something like RetrievalQA, and not only that, the defa
29.
Closure, from why the lucky stiff (2013)(github.com)
153 points by rchaves 3y ago|137 comments
30.
by rchaves 4y ago
Right now Arc browser

More ›