Well, it’s a zero day for 24 hours, right?
You should be more terrified. And also more glad
Well, the cars themselves weren’t racist though
Progress!
What do you mean? Any AI can replace any knowledge worker. You just fire the worker and say AI will do it, easy
Look around… Running a corporation isn’t about doing things or making things anymore, it’s about the stock price. And consulting companies have convinced investors that layoffs are so hot and rich
If you haven’t noticed, the quality of everything has been plummeting since COVID. They don’t care about tomorrow problems, just extracting as much wealth as they can before it all comes crashing down
I mean… It is genuinely hard to work for someone not evil. Let’s say you’re an AI engineer… Meta is probably the best because most of the non-corporate LLMs flow from there… But they’re also using it to build personalized echo chambers, which is horrible
OpenAI is at the top and Microsoft has shown every inclination to make it a monopoly, so I could understand wanting to work on competitors
You could go smaller and work somewhere like anthropic, but then you don’t have the resources to be on the cutting edge (depending on your specialty)
I blame people who buy Teslas more than those who work at Tesla at this point. Especially when they slow walk the bad things…I mean, Twitter would probably be less Nazi if more talent stayed onboard to resist institutionally
You can change jobs if the new one also sponsors you, and it’s my understanding that xAI tapped people from Tesla, but I might be wrong about that
Anyways, what’s happening sure looks like malicious compliance to me… It’s really not that hard to get an AI to list far right talking points, it’s just hard to bake it into the model
So you have people that made a pretty good model, but also can’t figure out basic AI infrastructure? I find that very hard to believe
I’m not. What would you do in this situation? Let’s throw in that you’re on a visa, so you can’t just quit
I’d maliciously comply.
You want access to the prompt? Here you go boss man. You want grok to share your Nazi views? Sorry sir, we’ll have to totally start over with training data. Or we could use a modified RAG
You want help with the prompt? Sure boss man, what do you want it to do? Oh, you want it to notice Jewish names? Sure boss man, I don’t know what you mean by that, but now it keeps saying it’s “noticing”. That’s weird
Oh, you want to fine-tune it on your tweets? Sure thing boss man… Oh, would you look at that, it thinks it’s you. Nothing can be done about that, it’s too much data from one source. Well, should we roll it back boss man? Your call
I’d just keep playing this game… Elon isn’t going to come out and say “I want grok to be a Nazi”, and I’m not going to read between the lines for him. I’m not going to come up with ideas to solve the problem, I’m going to let Elon’s ego direct the course and throw out “we’ve designed grok to seek truth over all else” as much as possible
Do they? Because every time I’ve looked at the issue, it seems like they manufactured a crisis out of a small number of unrelated home invasions
I think they removed the graphics of it, but not the effects when you have them in you party
Yes, electronics are very cheap… But remember the part where they also have a mechanical mechanism? They have two systems, where most cars have this very simple lock that connects to a tiny motor assembly. It’s literally a piece of plastic and a few wires
The tablet thing is true, they’ve changed cars to computerize everything, and once you’ve done that you can connect everything over a network. Every button needs to do back to a chip to become a digital signal, so before you had these complex one-off wiring harnesses for everything
But the tablet thing is again, common. It makes sense, it’s just worse
But Elon is a unique case. Elon likes to actually make decisions, because he thinks he’s Tony Stark. He actually goes down into teams and hangs out, and they have to just work around whatever decisions he makes. It’s present in all of his companies, but you can see it most in Twitter, because they didn’t have time to build a team to strategically distract him when he comes to visit
This absolute idiot has spent the last month trying to get grok to be a literal Nazi. First, he added a bunch of white genocide to the prompt, making it change the topic to that from any question for a few days.
Now it’s responding all confused, and saying things like “I never gave Jeffrey Epstein tours of spaceX or Tesla” when asked it Elon did it. Seems to me they fed Elon’s tweets in the RAG system in a amateur way
He micromanages and meddles constantly… That’s what he does at his companies
For a counterexample, Jeff Bezos. He was heavily involved in the fire phone, and had some genuinely cool ideas… But the priorities were all wrong, so it flopped. He learned his lesson
No, the problem is they engineered something they didn’t need to, because Musk thinks everything should be electric because it’s cool. They had to then engineer a mechanical release, because it was required by law (for good reason)
Mechanical door locks would have been cheaper. The fly by wire in the cyber truck is far more expensive, heavier, and far more dangerous than the very well polished power steering systems every other car uses
Maybe it’s something like they wanted to make more money on repairs or something… But even that they could’ve done better by starting from very common, cheap technology
Let’s be clear… The real problem here is that Elon Musk, opinion having idiot that he is, made decisions from on high with very little understanding of engineering
Because it looks like candy. It feels like jelly. It’s individually wrapped in clear plastic, just like candy
Now, imagine someone leaves one of those on the counter, or in a random drawer. That’s where loose candy lives.
So of course other people, who maybe don’t do laundry and don’t often see tide pods, are going to go “oh, look, candy!”
And then they call poison control as they retch and the cells in the mouth turn to soap, and they get added to the statistics
No, that’s the rationalization. They admin wants to use palantir to collect data on American citizens, including giving ICE mobile camera installations to do face tracking. They want to do a 1984, and they don’t want any speed bumps from the states
Also, AI companies gave Trump money and there’s talk of restricting their power/water access to data centers to prioritize the people living nearby
Can’t you just get a new launcher? That’s day one of getting a new phone for me
The other thing that he doesn’t understand (and most “AI” advocates don’t either) is that LLMs have nothing to do with facts or information. They’re just probabilistic models that pick the next word(s) based on context.
That’s a massive oversimplification, it’s like saying humans don’t remember things, we just have neurons that fire based on context
LLMs do actually “know” things. They work based on tokens and weights, which are the nodes and edges of a high dimensional graph. The llm traverses this graph as it processes inputs and generates new tokens
You can do brain surgery on an llm and change what it knows, we have a very good understanding of how this works. You can change a single link and the model will believe the Eiffel tower is in Rome, and it’ll describe how you have a great view of the colosseum from the top
The problem is that it’s very complicated and complex, researchers are currently developing new math to let us do this in a useful way
I have no idea what you’re talking about, but I think you dropped this: .com.ru
Do not ask people how to do this. If you don’t know from a web search, you shouldn’t try this without backing up a full disk image and understanding how to roll it back. Or at least backing up everything you care about
It’s not a particularly hard thing, but it’s a very irreversible one
You could just download it, no need to actually print the thing for no reason
I would buy a larger drive to do that though, just to say I did
You can in Japan apparently. They have a system where you can go to convenience store and pay by scanning a code
You can also sorta use zelle, but the technical integration is not great. It’s a very manual process as it is
Ultimately, the problem is we let two companies dominate commerce itself. We just need to let the governmet do payment processing, and require compatibility