Hacker Newsnew | past | comments | ask | show | jobs | submit | jki275's commentslogin

I run local models on my M1 Max. there are a number of them that are quite useful.


Preach.

Golang is the best language there is for most workflows that aren't bare metal embedded or have real time requirements, and this is coming from a 20 year+ C++ dev.


The NSA doesn't do serious work?


That wasn't the claim. Ability + interest + time + budget + ... are what makes a serious tool.


I can definitely write code with a local model like Devstral small or a quantized granite, or a quantized deep-seek on an M1 Max w/ 64gb of ram.


64gb is fine.


This subthread is about the Macbook Air, which tops out at 32 GB, and can't be upgraded further.

While browsing the Apple website, it looks like the cheapest Macbook with 64 GB of RAM is the Macbook Pro M4 Max with 40-core GPU, which starts at $3,899, a.k.a. more than five times more expensive than the price quoted above.


I have an M1 Max w/ 64gb that cost me much less than that -- you don't have to buy the latest model brand new.


if you are going for 64GB, you need at least a Max CPU or you will be bandwidth/GPU limited.


I bought my M1 Max w/ 64gb of ram used. It's not that expensive.

Yes, the models it can run do not perform like chatgpt or claude 4.5, but they're still very useful.


I’m curious to hear more about how you get useful performance out of your local setup. How would you characterize the difference in “intelligence” of local models on your hardware vs. something like chatgpt? I imagine speed is also a factor. Curious to hear about your experiences in as much detail as you’re willing to share!


Local models won't generally have as much context window, and the quantization process does make them "dumber" for lack of a better word.

If you try to get them to compose text, you'll end up seeing a lot less variety than you would with a chatgpt for instance. That said, ask them to analyze a csv file that you don't want to give to chatgpt, or ask them to write code and they're generally competent at it. the high end codex-gpt-5.2 type models are smarter, may find better solutions, may track down bugs more quickly -- but the local models are getting better all the time.


Bongino was a secret service agent..


There are literally hundreds of thousands of TS/SCI cleared government employees and contractors. There is nothing "exceptionally difficult" about any of this.


polygraphs are junk science.

I don't know who the other two are, but Bongino was already polygraphed and cleared for the Secret Service -- there's no reason to pretend that he wasn't cleared for the job. This article reads like a political hit piece and has no real grasp of reality. It also has statements that clearly betray its author doesn't really understand how security clearances work anyway -- most clearances don't require polygraphs, those are an IC and LE thing, and any OCA can grant any waiver they choose to grant. In this particular case if the AG wanted to review this decision she could do so as his boss, but it really makes no difference.

Given that polygraphs are, again, junk science, who gives a shit.


Then they could remove them altogether, but they aren't pushing that.

Your argument about the effectiveness of polygraphs is entirely irrelevant. The inconsistency and special treatment is the problem.


Completely agree, it's a waste to make a laptop screen touchable.

If there is really a touch "screen" requirement, I would put a screen under the trackpad -- that would be more useful for me at least. I think somebody tried that and it pretty much flopped though.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: