Quantum "bandwidth" decays exponentially as a function of distance.

Where I come from cost of coding is irrelevant. The only cost that matters is the liability incurred by the codes existence.

This actually seems much closer to telling a nurse to fabricate the drug before administering it.

Bursting of the AI bubble can't come soon enough. Far too much tech bro nonsense, DRAM / GPU prices way too damn high while unhealthy levels of collateral damage accumulate.

There is too much value in the ability of corporations to custom train models on internal datasets for the whole world revolves around our centralized AI service thing these AI corporations are dreaming about to have ever worked.

To make matters worse generative AI space is stagnating. Going forward cost of local inference is declining at the same time distance in capabilities between centralized corporate models and publicly available weights shrink. Regulatory power plays have stagnated as the public tires of AI scare mongering and so they are out of options.

What is going on currently with the colossus stargates is a destructive desperate hail mary to brute force their way out of the massive hole the industry keeps digging for itself.

Any article talking about qubit counts should reference the computers capabilities in terms of logical qubits.

Any article talking about error correction should provide a scaling law for the required fanout of correction circuitry as a function of logical qubit count.

Any article failing to do either of these things is a waste of time.

You wouldn't know the facts if they hit you on the head. This is the source:
https://searchepsteinfiles.com...

From this you somehow manage to infer the following:

"So apparently Jeffrey Epstein wrote that Donald Trump sucked Bill Clinton's dick and Vladimir Putin has the pictures.

Stop, read that again, Google it, and just Jesus fucking Christ what the hell is wrong with our world?"

NONE of the shit you spouted above is anything even remotely resembling a "fact". FFS you even manage to get the speaker wrong. It was Jeffrey's brother who made the statement.

What makes me uncomfortable is coming to obviously absurd conclusions based on shit information and then proceeding to wonder "what the hell is wrong with our world".. Look in the goddamn mirror.

What the hell is wrong with YOU? Get some mental help or something dude... your rantings are completely unhinged.

LLMs are simple feed forward networks run in a loop. They make no use of "statistical analysis" nor is there a "knowledge base".

The just statistics statements are as useful as saying just autocomplete or just deterministic. These are completely meaningless statements that in no way address capabilities of the underlying system.

It's is shorthand for "it is" ... "change its underlying probability" not "change it is underlying probability". "adapt its output" not "adopt it is output".

This is gobbledygook. You are guessing and have no actual clue how the technology works.

More unhinged nonsense from a clueless clown.
https://en.wikipedia.org/wiki/...

How do you square these two things? For what other reason would Trump pull strings to furnish a pedo like Maxwell with a luxury prison cell while publicly floating the possibility of pardoning her? Why do it?

Any ideas why the Internet is full of complaints about shit compression and lag? Is everyone using their Quest 3 incorrectly? I distinctly remember the complaints rolling in at the time. Checking back they are still there.

The problem with wireless HMD displays is latency and bandwidth. Years ago I remember seeing perhaps even for a short time shipping add-on product for Vive I think it was that were at least using high bandwidth 60ghz radios. Now people are going with standard WiFi and bolting on crappy compression schemes to make it work. This is a step backwards.

So an HMD with an internal battery and onboard CPU/GPU weighs less than a similar HMD without any of those things? Yea sure that makes.... total.... sense.

Do you ever read what you write before clicking submit? You can't have it both ways. Either the work is done within the onboard GPU generates heat or the same work is done somewhere else generating heat.

If you are wireless streaming and complaining about heat from the PC then as a basis of comparison you need to do the same shit on the onboard GPU and compare the results.

If you run something that makes the whole room heat up on a remote GPU running that same thing on the HMD if possible which it is not would result in the HMD melting into a pile of goo.

If you run the same workload of the HMDs GPU on a remote GPU then it isn't going to heat up the room because the remote GPU will be sitting mostly idle.

Quest 3 is a shit experience compared to tethered HMDs for seated SIMs.

I have no doubt it can run PC games. Yet the device is not physically capable of running any PC game directly on the headset. I could not for example run MFS, DCS or Elite. My PC is barely capable of doing that and it has huge fans dissipating hundreds of watts. If the HMD could do that ... it would instantly melt into a pile of goo.

I'm sure it runs smartphone VR games just fine yet this is all quite a far cry from running any PC game.

While I've never used Quest 3 there are no shortage of complaints about shit quality of the compression.

This statement doesn't make much sense. VR is an outlier 99.9% of PC gamers don't care about. 99.9% of software on the game store does not support *ANY* VR HMD. All the HMD can do at best is run low spec flat games in flat mode.

I have no tolerance for app store monopolies and dependencies, DRM..etc. If it isn't available on GOG, open source or directly from the vendors I'm not interested. If an HMD requires a fucking account on steam to work at all I'm not interested. This is just as stupid as saying to use a monitor I need to have a Samsung account or to use a mouse I need to have a Razer account. No fucking way, no thank you.

These issues have in fact drawn significant attention and criticism over the years. I reject the disgusting appeal to popularity you are trying to invoke to cover for fundamentally anti-consumer and indefensible behavior.

You are misinformed. There is ZERO NLP processing in LLMs. The only "translation" is a simple lookup table that translates integer tokens to word fragments.

They read, they just don't have infinite attention to dedicate to vendor nonsense. The source of software isn't relevant. What software is allowed to do is what matters. Misplaced focus especially given the fact Google app store itself is full of malware only contributes to fatigue swaying attention away from what is important to what is not.

If an OS vendor really cared about what was in the best interest of the user they would never place the user in a situation where they face take it or leave it demands for privileges from app vendors. They would not allow blanket privileges assertions so broad and nebulous their utilization is effectively ubiquitous.