nico

11 ms·

91.
by nico 8mo ago
Would love to be able to have customizable local model that only knows the stack I’m using For example choosing a model that knows rails 8 and react development running on a mac and using docker Ideally that would make the model small enoug
92.
by nico 9mo ago
Does this need depth data capture as well? The “casual captures” makes it seem like it only needs images, but apparently they are using depth data as well Also, can it run on Apple silicon?
93.
by nico 9mo ago
Nice. Would be great to have this on Slack or Whatsapp
94.
by nico 9mo ago
Very insightful, thank you How do you implement the permission scoping to the step? Do you have any shareable code or examples?
95.
by nico 9mo ago
Are you doing that as your own personal tooling? Are you open sourcing it? Would be happy to take a look and maybe contribute as well
96.
by nico 9mo ago
> Don't Do you mean "Don't give it more autonomy", or "Don't use it to access servers/dbs" ? I definitely want to be cautious, but I don't think I can go back to doing everything manually eith
97.
by nico 9mo ago
You are right, and that's great for queries How do you provide db access? For example, to access an RDS db, you have to connect from within the AWS/EC2 environment, which means either providing the agent ssh access to a server, fr
98.
by nico 9mo ago
This is probably the safest thing to do, also the most time consuming It would be nice to just be able to solve it through instructions to the agent, instead of having to apply all the other things for each application/server/data
99.
by nico 9mo ago
Yup definitely tricky. Unfortunately Claude sucks at answering questions about itself, I've usually had better luck with ChatGPT. Will see how it goes
100.
by nico 9mo ago
Do you use Claude Code? Do you say "Yes, and don't ask again" for all the commands, since you don't mind breaking things inside the container?
101.
by nico 9mo ago
Great pointers, thank you How would you go about allowing something like `ssh user@server "ls somefolder/"` but disallowing `ssh user@server "rm"`? Similarly, allow `ssh user@server "mysql \"SELECT...\&quo
102.
Ask HN: How do you safely give LLMs SSH/DB access?
85 points by nico 9mo ago|105 comments
103.
by nico 9mo ago
Would love something like this for Fusion 360. Being able to just prompt the UI to create or edit objects. It would be cool if (like with coding agents in which you can add context using @filepath), you could use the mouse to click/sel
104.
by nico 9mo ago
Some friends built a whole company around this problem, it’s actually pretty difficult to resolve, with lots of edge cases, especially if you are handling multiple banks and lots of customers with slightly different needs This tools looks p
105.
by nico 10mo ago
Super interesting. Some time ago I wrote some code that breaks down a jpeg image into smaller frames of itself, then creates an h.264 video with the frames, outputting a smaller file than the original image You can then extract the frames f
106.
by nico 10mo ago
> coax me into using my local currency instead of GBP and hid a £20 spread in the exchange rate BoFA does this for international wires as well. And I suspect a lot of companies do this to their international customers too. Unfortunately,
107.
by nico 10mo ago
Building the self-service/billing system for a CX saas ( https://autopilotreviews.co ) While trying to figure out a good ICP and reach PMF
108.
by nico 10mo ago
Could you build an inverse kinematics solver with this? (I recently watched a youtube video of someone iteratively working out the solutions for a robotic arm, by alternating modifying the inputs and the results)
109.
by nico 10mo ago
> Lots of people working in different fields end up reinventing things that have been known to math for centuries I remember reading, about a year or two ago, about a medical doctor that published a paper rediscovering calculus (I just l
110.
by nico 10mo ago
Amazing! One time I did something similar. Went to see a movie that was dubbed in Spanish, so I downloaded the movie in English and extracted the audio ahead of time, then I played it on my phone wearing headphones while watching the movie
111.
by nico 10mo ago
Has anyone done something like this but with apple silicon instead of a graphics card? Training a small LLM on an M2-M5?
112.
by nico 10mo ago
> It's not easy to find a psych who does adult Autism screenings, but they're out there And it’s very expensive to get the diagnosis, it can be up to $20k in California
113.
by nico 11mo ago
> Claude often ignores CLAUDE.md > The more information you have in the file that's not universally applicable to the tasks you have it working on, the more likely it is that Claude will ignore your instructions in the file Claud
114.
by nico 11mo ago
> Claude basically disregards your instructions (CLAUDE.md) entirely A friend of mine tells Claude to always address him as “Mr Tinkleberry”, he says he can tell when Claude is not paying attention to the instructions on CLAUDE.md when C
115.
by nico 11mo ago
Can relate. My inactive google ads account all of a sudden got banned. No explanation except some generic link to their terms of service. Appealed, got automatic denial, no reason given. Have retried multiple times, same result
116.
Ask HN: Are you maintaining an app using Rails 4.x or lower?
2 points by nico 11mo ago|0 comments
117.
by nico 1y ago
It’s wild that consumers need a piece of cutting edge technology, to have a fighting chance against corporations taking advantage of them
118.
by nico 1y ago
That’s a very interesting take. I hadn’t really considered evolution I guess if you really wanted to start from scratch, you could figure out how to evolve the whole system from a single cell or something like that. In some ways neural netw
119.
by nico 1y ago
Would love to see an architecture that learned more like humans. Start with just imitating one letter, then a few more, than some syllables, then full words, then sentences, etc. Progressively adding on top of previous knowledge Also, it’s
120.
by nico 1y ago
I just built a logistic regression classifier for emails and agree Just using embeddings you can get really good classifiers for very cheap You can use small embeddings models too, and can engineer different features to be embedded as well

More ›