I feel people have forgotten how aggressively Google push chrome in the early days.
Using Adware distribution tactics, where if you forgot to uncheck a box, while installing some free app, you'd end up with Chrome installed.
Why does AI need to be involved in such a migration.
The original game was made by one guy.
Its a small team with a deep understanding of their project, it seems counter intuitive to outsource this to AI.
i'm very confused why HNers got amnesia about pre-AI times. It was always possible to build software fast if you had expertise and good code that was easy to work with.
It counted as fast then. It would be really slow in comparison today. Sure, whole engine rewrites with compatibility shims were possible to do in months in the past. But let's not pretend it wouldn't be extremely faster today with LLMs.
I have no idea if that's true anymore in this situation. The benefit of good design and code is that it's flexible where it needs to be to make changes easy and hopefully avoids many changes overall.
If changes become cheap, the same benefits don't apply anymore. Large scale refractors can become background tasks. "Turn this project into ECS style" is one session away. The things we valued in long timescales happened because changes and backtracking were painful and a huge time sink.
Reading Jurrasic Park, it as bit of fun hearing of how the island purchased multiple Cray super computer to help with genetic research.
Fun even if its a bit of fiction.
All LLM are heavily subsidized by investors. No one is paying the true cost of tokens at the moment.
The rug pull is to get as many people defendant on the tool before they have to massively increase prices.
Sure, open models will definitely be an alternative once the frontier models hike prices to match costs.
But that will be a minority, and people will start to switch, but people use the frontier models for a reason and these AI labs will extort that, once the easy investor money dries up.
My company pay a copilot subscription to everyone, plus my team have a bedrock subscription that cost us 7k last month. In the meantime, some guys created a new internal team and are testing open source models (for our red team) on our own hardware (we didn't sold our 'old', 3 year old servers RAM/CPU, so the only investment was GPUs). My guess is that within the next 6 months our red team will use open source models exclusively, and my team will probably follow within two years. Especially if the GPU prices go down
You're assuming that people will be able to buy GPUs in the future. Which currently isn't the case and there is no timeline when they will be available again (at their pre-AI affordable price). I wouldn't be surprised if the labs keep buying up all the GPU capacity forever just to keep home models at bay (which I'm sure is what they're currently doing with all computer hardware).
How long are those models going to be safe to operate with internet access and tool use without ever being updated? Imagine if "do it for my grandma" is still as effective today because the models haven't changed and that every open-source maintainer upset with AI started polluting their libraries' logs with prompts that have been proven to work with no known reliable fix ever coming in this hypothetical future.
Also: who are the investors subsidizing electricity bills for the people who run LLMs on their own GPUs? I'm still waiting for my subsidy cheque over here, so if you can plz send it to me ASAP, thanks.
Question back: how do you imagine them "raising the prices" of the tokens generated on my own GPU? That's the implication we're discussing here. Please explain how that works.
as much as training might be subsidized, I don't buy that inference is, if there was no training required and labs could run their current models forever it'd be a gravy train even at current prices
I think the crux of the issue is that frontier labs do not make any financial sense. They are spending investor dollars in order to sell tokens at pennies, and then a Chinese lab can build a competing model on those pennies, and the frontier labs can't then start selling the tokens at a price where they'd actually even break even because most of the volume will then shift to those cheaper models.
Considering the reach of a HAM radio, the limited frequencies, and the consumption of airtime while transmitting, someone needs to enforce etiquette.
Its also likely the older folk operating in HAM radio space, have a background in communication operations when HAM was a far more common communication device.
HAM is still awesome, I find it very interesting with the long distance, remote, communication challenges some people partake in.
https://olduse.net/
Experience legacy usenet in real-time.
You can subscribe to different start points and receive posts as they were sent, 40+ years ago.
Even GOG has some oddities were the available copy is not the original experience.
My recent experience of trying to play the original Mafia game.
Booting it up a GoG copy, I found that there was no music while driving (a core memory of my original play through).
This music has been removed...
This also affected small radios you'd find playing music, they now dead.
It really took away from the game experience
I ended up scratching out my original CD copy of the game, installing version 1.0 of the game. It had all the music and better visuals
The patches removed visual features for performance, not GoG's fault here, not sure why GoG hosts an inferior version, its quite a downgrade. The music removal I do not understand.
(This all was particually obvious in an early mission, where you have to beat up some parked cars with baseball bats, the patched version show minimal dents and zero scratches, and the radio sitting in the corner make no sound. This completely changes in the V1.0 copy, so much more atmosphere)
This happens because the original publisher signed a deal for the commercial music with a limited time.
So when that time expired, any further releases had to remove the music.
Another example of how copyright makes preservation difficult.
I mean, it's obvious: copyright seeks to create a mechanism to compensate creators by creating scarcity. They succeeded.
It turns out archiving those creations effectively is in direct tension with that scarcity. So we made a system to incentivizes creation, and then destruction of the creation. This is visible with the "copyright cliff" in books, but also the incredible difficulty of legally keeping computer programs running over the decades. Copyright's restrictions on derivative works, coupled with the insanely long copyright term, create a sort of digital wasteland, with only those willing to go underground having great collections of old programs.
Raspi is likely going to continue making moves like this, since they have gone public.
You can't just have people buying the cheaper 1GiB boards and upgrading them to 4GiB, that's taking potential profits away from the investors...
I doubt many are doing this upgrade, and I think its awesome someone has the tooling & skill to do so. I cannot fathom how this is a threat to RPI.
I would assume a community firmware patch will likely release & resolve the limitation (until it becomes a signed binary)
Consider the alternative where a bigger number of modded boards with dubious RAM gets into circulation, which is no problem for Chinese companies to do. It would quickly result in a distrust of all RPi products and everyone would suffer.
It also has nothing to do with going public, it is consistent with what they were always doing: it never was fully open, in fact quite the opposite and they were always protective about their HW, you can't even order the SoC as a separate chip, only as a part of premade/customized board.
This doesn't stop dubious RAM from going into circulation, it just prevents different RAM sizes than what it expects. A Chinese company could still swap in faulty RAM bought in bulk as long as it's the same size. This doesn't do anything to protect consumers or brand, just profit margins.
No, if you read the actual post by them they said swapping in parts of the same size won't work either because they program in several other hardware attributes that the board will check for when booting.
I think what's giving me more concern is: what if the Pi is embedded in a device and the manufacturer doesn't realize the supply chain got tampered with?
One EEPROM update and suddenly you brick the board. It's like the FTDI bullshit. You're going to end up snaring legitimate users that didn't know they had illegitimate boards.