nperez

7 ms·

1.
by nperez 5d ago
Immediate first impression is that this tagline should go. If the project brings something valuable to the table it doesn't need to shit on other permissively licensed open source projects to make a case for itself
2.
by nperez 10mo ago
The HN audience is overwhelmingly aware of the issues around right to repair and data collection, so there isn't much reacting to do there - I assume there's already near-unanimous agreement that it's a good thing to educate
3.
by nperez 11mo ago
Seems like a more organized way to do the equivalent of a folder full of md files + instructing the LLM to ls that folder and read the ones it needs
4.
by nperez 1y ago
I feel like this sort of thing will be referenced for comic relief in future talks about hysteria at the dawn of the AI era. The article actually contains the sentence "The machines aren’t just taking over our work—they’re taking over
5.
by nperez 1y ago
Whether web or native is better is hardly relevant to the core of this issue IMO, which is about fundamental rights to admin our own devices. Having to make a network request to fetch an external resource every time you want to run code on
6.
by nperez 1y ago
Agreed 100%. When you work on an app every day, it all makes sense to see the cool features flash by, but you need to design for people who don't have a clue what your app does.
7.
by nperez 1y ago
I'm not a modeler but I've tried it a few times. For me, modeling is a pain that I need to deal with to solo-dev a 3d game project. I would think about using something like this for small indie projects to output super low-poly ba
8.
by nperez 1y ago
Kind of fun to get into a brutal insult battle with. Hope I didn't violate any TOS with with that one.
9.
by nperez 1y ago
No degree. I've been working the full stack for almost 15 years full time, including recently learning to train various types of gen AI models. There are still orgs that are rigid about their requirements, but I'm a mid-30s guy at
10.
by nperez 1y ago
It's inevitable because it's here. LLMs aren't the "future" anymore, they're the present. They're unseating Google as the SOTA method of finding information on the internet. People have been trying to do t
11.
by nperez 1y ago
I'm not going to disagree because greed knows no bounds, but that could be RIP for the enthusiast crowd's proprietary LLM use. We may not have cheap local open models that beat the SOTA, but is it possible to beat an ad-poisoned S
12.
by nperez 2y ago
I think I'm mostly in agreement with your points. I think a significant part of the downfall of DEI was deliberate bad-faith behavior from those who actually oppose equality, but there are also things to be learned about how DEI progra
13.
by nperez 2y ago
I think there's still appeal in the underlying (very) basic ideas of trying to create a workplace that's comfortable for everyone. It's being rolled back quickly because that's what influential rich people want, and beca
14.
by nperez 2y ago
I think information and culture/fashion both have a lot to do with it. Pre-social media, you could get drunk and embarrass yourself, and forget about it by the next day. Now everything is recorded. Information about alcoholism is easie
15.
by nperez 2y ago
Not a lawyer but my opinion is the law shouldn't be used to compel anyone to provide unpaid services. With that said, I still think it's terrible for the community. If other package repositories like npm, PyPI, and others decided
16.
by nperez 2y ago
The preamble of the GPLV2 begins with: "The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change
17.
by nperez 2y ago
Maybe an unpopular opinion, but it sounds like they're doing this the right way. I want to be able to run a local model using something like VLLM or FastChat, then be able to call it from a context menu. No obnoxious toolbar taking up
18.
by nperez 3y ago
That's hilarious. Someday we'll all be working QA, just scanning over AI output for issues, like manufactured goods passing by on a conveyor belt.
19.
by nperez 3y ago
I began my career in a startup where pretty much everyone else but myself was an Ivy grad, with many coming from FAANG internships, and I have very high opinions of them. We all were very driven and all had things to teach each other. I don
20.
by nperez 3y ago
Reminds me a lot of the boat I'm in - based on his CV it looks like he's self-taught and started around the same time I did. This market right now is killing us, and I've got to wonder if AI tools are filtering out self-taugh
21.
by nperez 3y ago
There's not enough info on how this works. I need to understand the full user flow for both parties before considering signing up. Will I be giving others an alternative email address, like what Apple does? Or do they also need to sign
22.
by nperez 3y ago
Not sure if this is the primary reason, but it should be enough to throw off anti-spam systems that are looking for repeated posting of the same content within a short timespan
23.
by nperez 3y ago
I'm not a lawyer, but I believe there was a court ruling that says AI generated works cannot be copyrighted. So you could use them, but couldn't stop anyone else from doing what they want with them
24.
by nperez 3y ago
For those not in on the origin of the joke, Next.js introduced "server actions" which became a huge meme / debate topic on Twitter - for example, https://twitter.com/peer_rich/status/1717609270475194
25.
by nperez 3y ago
Is the versioning scheme only in .001 increments? I know version numbers aren't indicative of progress, but I still find it fascinating that a 26 year old project is just releasing 0.260
26.
by nperez 3y ago
I've seen businesses tank because of search algo changes long before this. I see it as something similar to reliance on a third party API - you have to be prepared for the rug to be pulled on you at any time. This is why it makes me an
27.
by nperez 3y ago
I've worked on a project that was intended to replace Excel spreadsheets with a web front end. It was presented like we were moving our processes out of the stone age - as if we were replacing a pen-and-paper process that "doesn&#
28.
by nperez 3y ago
Ah, 175b it is then. That's actually kind of encouraging that an open source trillion+ param model isn't needed to get there. A 180b Falcon model very recently came out https://huggingface.co/tiiuae/falcon-180
29.
by nperez 3y ago
I'm not sure, I've tried looking it up and couldn't find an official answer. But that site says GPT3 is 175b and GPT4 is 170 trillion. That's insane if true Edit: and this one says 1.7, idk https://levelup.git
30.
by nperez 3y ago
ChatGPT uses a 1.7 trillion parameter model IIRC. I wouldn't expect a 70b model to compare well. There are medical finetunes that might be better to try for this, like https://huggingface.co/wanglab/ClinicalCamel-7

More ›