Yes. I don't have any of my banking apps on my phone. Why do you need that? It's a bit more annoying to click all the "No I don't want to use your fucking app" popups when going to their website, but I can just go to their website in a browser on my phone in the rare instance I need to look up my banking information on the go, which is rarely or never. If I need to send money to a friend, I can use Zelle on the web.

I did have mobile banking apps on my phone at one point, and I just setup Magisk to ensure those apps couldn't check for root privileges. It worked fine, but on my recent phone, I saw no reason to install banking apps anyway.

Does your bank literally require a smart phone? Fuck that. Get a different bank. If you live in an nation where literally every bank will not give you an account without a smart phone app, there is something seriously wrong.

I just use a piece of plastic. I don't have to charge it. It doesn't need batteries. I can just insert it into any point-of-sale system. I really don't get the whole "pay with your phone" bullshit .... seems like only negatives with zero positives.

Steam at least allows people to run whatever on SteamOS and doesn't lock down the hardware. Too bad their Steam Box/Cube is being delayed by the RAM cartels. CXMT needs to come off the fucking embargo list.

You've obviously never flown an airplane in your life. There is SO much that has to be done. Pilots have to calculate their fuel, run through a heap of checklists, monitor oil pressure per engine on takeoffs and landings. Most pilots don't even use ILS for landing; it's easier for most of them to just land by sight. Planes can auto-descend, sure .. but anything less than ideal conditions (rain or wind) needs manual piloting.

It's worth nothing that every cargo delivery startup so far has been a massive failure, with even Amazon struggling to get things past the proof-of-concept phase.

Planes are nowhere near being automated. Drones? Maybe. Cargo planes? That would be a big stretch, but that would have to come first? Passenger planes? Not for at least 15 years, and probably never.

Are you a software engineer? They big money is airlines. Some of those pilots can get up to $200k with experience.

I love how Anthropic makes some statement about AI and war/surveillance, even through they have a partnership with Palentir. Such crazy bullshit from both of these brain destroying companies.

> fascist

That word is thrown around so much it means nothing. Jensen is a fascist?! That doesn't make any sense. Annoying, cringe, greedy, ... sure. CCP collaborator, to the extend of increasing his money and power .. sure, I guess .. but he was born in Taiwan! He should really know better than support the CCP in any way.

But to be fair to the man, he did found nvidia. I still remember my first TNT2 video card; so much better than dealing with the Voodoo2 and the VGA bypass. He ended up buying out 3Dfx too. It's sad we are where we are today.

It does not KNOW! Do you even know what an LLM is? It's literally a big blob of floating point weights from ever part of a word to every other part of a word (and embedding space). Through breaking words into tokens and running them through transformers with various blocks, it generates the next token. That's it. It has no knowledge.

It's not a "bug" because there's no real code flow that can be adjusted. The original human re-enforcement learning took thousands of human hours of people sitting in cubes clicking on the generation that was the least retarded. It's not easy to fix these machines.

We don't know if this person "prompt hacked" into fantasy, but these are complex systems and there is a massive lack of understanding as to what exactly they are.

5090 runs at 600w and is air cooled ... but most of these new data center machines do have central water cooling across all the racks.

I hope they don't die out. There isn't enough competition in the CPU market, or in the Fabrication market. It's bad enough they have some Chinese CEO that's totally cucked to Jensen Huan and are 10% owned by the US government. I miss Pat Gelsinger. That 13th/14th gen disaster really screwed up that entire company.

Because it really does not work that way at all. You can train on law databases for billions of compute hours and even add a 1000 hours of real lawyer feed back reinforcement, but at the end of the day, it's just billions of weights for parts of words to parts of other words. It's randomly guessing then next best word. A RAG store would at least point you to documents, but a lawyer needs to read those documents; not LLM summary.

Cases are complex and your answer shows you don't know how the random token machines really work. Neither do the lawyers.

The thing that bothers me so much right now is the absolutely insane amount of dependence people have on these random word machines. I know people who use them instead of search engines, or they'll read the generated results at the top of Google/DDG as if they're fact without question. There is little literacy onto how models really work, and even for developers who should know better, they still see the current generation of machines similar to the movie "Her" than the semi-deterministic feedback systems they truly are.

The trust people put into these things is frightening. If it isn't for the right attorney or judge that digs through and discovers these completely randomly generated court cases, cases like this could be made entirely with incorrect standing and precedent. There are people who do not use chatbots at all, or in limited capacity to generate images, video and segments of code they clean up. There are others who use them for everything from recipes to fitness advice to generating entire applications they submit for code review without manually reviewing and fixing the generation to match their code style.

When the crunch hits the LLM industry, some of these people will easily shell out $500/month to keep their bots, and it will require that much to keep some of these companies afloat. The lack of decent off-line models is troubling. Some companies will refuse to shell out the $1k~$2k/employee increases I think we're likely to see. This is a disaster brewing, and we're not ready as a society to deal with it.

Since we're finding some of these models can reproduce entire chapters of Harry Potter with 95% of the original, I can only imagine the actual models used by Anthropic/OpenAI are probably 400GB ~ 1TB in size. They're not telling us what's really in these models. I suspect they are huge, and at some point, it just turns into lossly JPEG/mp3 compression with weird realistic artifacts.

It doesn't really work though. Both companies are over-inflated and over-extended and both will suffer hardships. OpenAI is betting on their government contracts being more than their dwindling GPT subscriptions. I bet most of these people don't even have paid accounts. I've written about how you can't vote with your wallet before:

https://battlepenguin.com/poli...

There are a lot of hard, difficult problems that no LLMs can solve. So Stack Overflow today tends to have only very hard problems, with complex dependencies. The examples in my articles were fully self-contained with pretty thin dependencies and they were still both closed as off topic and deleted (I have yet to solve them either. I could try again with Claude Sonnet/Opus I guess, but I have heavy doubts the random code generation will be helpful).

The Great Decline of Stack Overflow happened long before the AI era. The place just became a toxic mess of rules and completely legit questions were shut down and deleted for no reason. I wrote about this last year:

https://battlepenguin.com/tech...