user_7832

While I like the premise, the opening text:

> Ah, Manufacturing Consent—the font that doesn’t just scream “serious journalism,” it calmly editorializes it in 144-point Fraktur while subtly nudging your worldview.

Unfortunately feels too much like an LLM output for me.

(Which is probably post-ironic in this context?)

Bonus suggestion: get that internet archive extension for your browser that automatically sends every page you visit to the IA (if you're ok with the privacy side).

You don't have to worry about having your copy easily accessible, and as a bonus it's also available for everyone else in the world.

Idk why (apparently) no one has mentioned it yet.

Forget consumer TVs.

Get a commercial "display".

That's it. That's the solution.

Pros: dumb display. Longer warranty.

Cons: more expensive.

Because, of course. Why'd they be the nice guys?

I've had so many conversations trying to use (for example) disclosed CVEs to gain access over my devices, and the lovely guardrails refuse to answer, thinking I'm targeting someone else.

And I suspect even when you later clarify your position in the chat and "convince" Claude, the report's probably gone and already flagged your account.

So (if my understanding is right) Claude makes and stores/sends a report after every (suspicious, per anthropic) message, and enough reports get flagged?

I've never been more interested in running my LLM locally. Thanks Dario.

Thanks, I haven't used the gemma models in a while. Last I remembered, their small models weren't bad but obviously showed signs of being "small". (Funnily irl I use Gemini 3.1 pro more than all other models combined).

Here's the link btw (also for my own reference):

https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/blob/main/Qw...

That's a very good point I hadn't considered, and I think you are right.

Apparently the ud-q4_k_xl version of qwen 3.8 27b is quite better than the q3 and not much larger. At least per r/localllama. (I'm also trying to find a good local model).

It's crazy; some years ago (many?) there were so many Alroids projects. I even wanted to buy a few tins of mints just to use the boxes (not that I dislike the cinnamon ones - or the quick rusting not withstanding).

I remember, there used to be powerbanks, buck-boost converters, joule theives, a ton of simple but cool projects.

I wonder where that vibe died. (Or if it transformed into something else?)

Don't worry, unless it's like GLM being used, every AI tool will refuse to backdoor the code.

Just like you just need to ask the LLM to "make it bug free".

(At this point even I'm more sure if any of this is sarcasm.)

Summary of things that Apple has done: (copied from another thread)

1. Apple is the largest buyer of flash memory in the world

2. It used the leverage to force nand manufacturers into deals that ensured it was the most favored customer and no other customer got a better deal

3. $0.80 of every dollar consumers paid apple for memory upgrades went to apple's profits, only $0.20 were spent on supplies. Apple was making big money by squeezing not just the manufacturers but also its customers

4. Apple would order memory at already rock bottom prices and refuse delivery without communication, leaving manufacturers with loads of unused inventory which forced them into voluntarily lowering prices again

5. Apple schedules negotiations so that competing manufacturers would run into each other and dragged these evening meetings past midnight without dinner as a form of psychological tactic

6. To enforce this it forced expensive and lengthy audits on memory manufacturers to look into their books to reveal how much competitors were paying and how much each manufacturer was outputting to identify the struggling companies so that they could be squeezed even harder

7. Apple once threatened to have sandisk's ceo "crawl on his knees" when sandisk sold to higher paying customers instead of apple. When western digital required a fee for capacity reservation, apple sent a picture of the date written on a whiteboard to western digital and promised "payback"

8. As revenge, apple demanded that western digital pay them "testing fees" if they wanted their products considered, without any purchase commitment. They then got the security to escort the western digital executive out to humiliate him

9. Apple's spending on memory has quadrupled since the rampocalypse as manufacturers have rejected apple's past negotiation tactics. Apple was caught off guard because they were drunk on their leverage power over memory manufacturers and didn't expect them to push back. Memory manufacturers are now "gleeful" at apple's state

> Of course, model size doesn't capture all variation. Models that are trained later might have new data or new techniques or new architectures that earlier models lack. You can see this with models from all providers - Sonnet 5.5 outperforms Fable 5.1 in ways, and Gemini 3.8 Flash outperforms 3.1 Pro in ways.

Is this really so/could you give a rough example please? I've not used a lot of Claude or ChatGPT, but 3.1 Pro always feels most grounded and intelligent. 3.8 Flash (in my very limited testing) isn't bad, but it feels less nuanced.

(Fwiw I don't use it for code, but ask maths and real world logic based questions.)

I agree that it's a job, but.. for most of humanity, people lived and worked in the same spaces with the same humans. The home was a unit of (pre) industrial production. Your neighbours were your partners, or at least involved in general economic activities with you.

Our human brains unsurprisingly still default to that, and it is quite understandable imo.

From what I remember, most of these studies found that these guys were not diagnosed (previously).

Very similar, except I live in a very large city... but it's not an anglophile 1st world country, so tr list is full of streets and unis and roads. "Hey do you know your <location> features <location station>!?" lol

I misread mortality as morality and was exited for a second :‘)

Fun fact, now that you're here: adhd is severely over represented in prison populations. Several times the baseline general population rate.

Fwiw I think you can run 3.8 flash (next?) on your setup with SSD streaming.

There's this video (1) that I found, and my understanding is that flash next can be very efficient wrt ram needed, by using a "dictionary" (or something similar).

1 - https://youtu.be/IH8XmxiwliQ?si=lJPvYzi3mHNsiw8r

> The actually sane default is to treat it like a junior trainee you would never trust with rights to anything critical

In my experience (mainly claude and Gemini argon, less so with gemini 3.1 pro), it's a bit worse. A junior dev won't hit "yes, delete" so confidently and think "I can just remake it" and proceed to ~~lie through its teeth~~ apologize 2 seconds later.

I agree that reply was unnecessarily rude, but... idk man. I'm a huge cheerleader for Google [1], use 3.1 Pro via AI studio, I'm on a Pro plan... and I can't access this model? Not even a "4 chats per hour" type thing?? I mean, I have video generation access for some reason (no good for me), but Google's policies are baffling to say the least.

Ofc I know this isn't your fault in person, but I really wish Google would get its shit together.

...As a long time tech enthusiast I should probably stop being so delusional. Heh (cries internally)

1: lol I know who cheers for a trillion dollar company but I'd rather not cheer for OpenAI

It's good if you're a tech guy in silicon valley with a good salary who used to upgrade every year to the latest pro max.

It's terrible if you're a labourer in India who buys phones under 10k INR (~100 USD) when your salary is 20k INR per month.

Unfortunately there are way more in the latter category than the former, and most of them won't know what ram is. They'll just be paying more and getting less going forward for a few years.

(Source, the salaries and phones I've seen of the staff at office. And we're in a major region you've likely heard of, not some tiny rural village.)

Tangential but I've been low key dying to tell someone - just today I FINALLY hit 100% usage on my paid claude account.

I really don't know why, but it feels very satisfying. It's like the LLM telling me "good job buddy, you've worked hard enough. You deserve a break now."

(For context I write a lot of text per message it sends. I'm not asking it to one-shot an app to solve world cancer, I'm just working with mathematical stuff with a lot of back and forth.)

And I think Alexa is a proper name too in some places?

How'd you do it with a vpn? Doesn't it still need your facebook account to be from the US?

For a company of Apple's size? Yeah it's the difference between pocket change and pocket change.

If browsers changed that much so suddenly, a lot more people would have big problems.

Technically proper HDR is supposed to do exactly that. Doesn't matter if it's a pass or an image on a website.

True, but there's a very good business reason, and it's a slightly unethical/scammy one. Just like here.

> just rent access to a Mac online

The friendly librarian/community centre person sure as hell isn't going to be aware that that's an option.

Understandable and quite possible, however that seems like a relatively lazy thing for Apple to do.

I'm not entirely sure what's the use case for most of these cards, but I'd hardly be surprised if this kind of decision didn't affect usability and accessibility for a lot of iPhone users, just because their venue guys don't have a mac.

And it's not like apple doesn't have the resources to make this software run across different types of computers. Iirc itunes and bonjour are (still?) very much developed by Apple for windows.