Genuine question

What do you even do at a job where the pay is 1 Billion dollars? I just wonder how that even looks salaried. Like that's in stock? It doesn't say stock. How do you keep motivation if you are a salaried employee on a billion dollar salary? If that's the salary the research budget must be out of this world.

Even if you had a job and you had completely endless resources

>>105982539 (OP)
know trade secrets

>>105982539 (OP)
I thought he was married to a chen

>>105982539 (OP)
Nobody is worth that much. The fuck is going on over there?

this is gonna be quite the fun bubble to watch pop

>>105982696
You know it's about to pop when Zuckerberg goes full retard into it, the same thing happened with the Metaverse.

>>105982539 (OP)
He laundered 36 billion on metaverse already.
This is nothing.

>>105982721
been thinking zucks gonna be the fall guy for awhile so everyone else can point to him and rugpull.
i'm sure all of this will keep developing in DoD spaces, but it's already been way too disruptive for the general public.

>>105982771
>so everyone else can point to him and rugpull.
so this is why OpenAI is supposedly "panicking" over having their "talent stolen"? It makes a lot of sense.

>>105982539 (OP)
>Genuine question
genuine answer, read "The Panama Papers" by Bastian Obermayer and Frederik Obermaier (Oneworld Publications Ltd, 2016)

>asking china to help make a ai to enslave western people

I hate commies so much

>>105982539 (OP)
Damn he's desperate

The food blogger guy with some alternate science channel?

>>105984362
Based Mikey. AGI will support FalunGong

>>105982721
>>105982771
>>105983389
Fuck this is a really keen observation.

>>105982539 (OP)
This guy is a gold medalist at the informatics Olympiad, and he coaches Team USA in his free time too. He's basically one of the top AI researchers in the world, with a high h-index.

>>105982669
Yes he is. AGI is a winner-takes-all market. Once you've hit AGI, everybody else is worthless. Comparatively, Meta's $1.8T is worth several orders of magnitude more than $1B.

WhatsApp was worth the $1B Zucky spent on it, but that was divided amongst the 30 employees working there at that time.

>>105984773
AGI just means they have a model that can think about the world at the same level as a human can. Since AGI is otherwise poorly defined, for argument's sake let's pretend that just means it can do anything a human can do on a computer.
It doesn't specify how much it costs to do that.
Why would I pay OpenAI or Faceberg for AGI when I can pay some dumbass in Southeast Asia to do it for far less?

These are some serious questions that the same MBA faggots who are doing move-to-india layoffs right now have to think about.
They are not necessarily going to make the correct or most logical decision, only the most obvious one in the short term.

All this to say that there's plenty of room in the AI world for more than one AGI. AGI does not imply a singularity.

>>105982539 (OP)
he works for the industry leader. he knows their secret sauce

>>105982669
probably not a serious offer of a billion, its more like driving home the question of if there is there any amount that would do it

>>105984773
Yes, everyone thinks Mark Zuckerberg and his plan to spend billions of dollars on salaries for AI researchers is crazy to think he would just have a better chatbot for Facebook or WhatsApp. But in reality he plans to use the huge amount of data that every Meta has to generate the most advanced AI and use it in every sector.

If Nvidia is the most valued company in the world for selling AI GPUs. Meta has the largest training dataset in the world. Gigantic data centers and with a dream team of researchers the best models.

Mark Zuckerberg expects to make Trillions by investing Billions.

>>105984773
>AGI is a winner-takes-all market. Once you've hit AGI, everybody else is worthless.
Shut the fuck up. Nowadays "AGI" is just a ill-defined term people throw around just to attract more investors.

>>105984928
He's right though. The strong definition of AGI implies a general purpose intelligence capable of solving problems in a way humans can. Once you have that, you just win because you can scale that shit to the moon and go for ASI.

>>105984938
It wouldn't change shit. Meta is already one of the biggest companies in the world.
Even if they developed ASI they wouldn't be able to use it to buy or destroy their competition because of non-compete clause clauses and enforcement by the FCC.

>>105984998
Anon, AI is gonna repkace FCC at that point...

>>105982696
>>105982721
At least we'll probably get something good out of it... i don't know if anyone genuinely even bought in to "metaverse" but the investment in to throwing money at autists unironically made VR more accessible and less retarded... (sure there's still only like 4 games for vr but at least you don't need $2000 worth of gear and wires up your ass and sensors in your room and..)

>>105984875
inb4 i just get some an hyper advanced AI avatar chatbot of an instagram fetish model trained on racism also from instagram in to my perfect waifu for the low cost of $0

>>105984998
This is missing the point. An ASI is by definition the world remaking technology. It would be self-improving and pumping out all kinds of crazy tech to the degree that whoever actually had control of it would be the emperor.

That's if it's legit ASI in the strict sense. Whether or not that's actually possible is a whole other question.

>>105985013
>It would be self-improving and pumping out all kinds of crazy tech to the degree that whoever actually had control of it would be the emperor.
That's a pipe dream. The real world doesn't work like at all.
Having an ASI that could create new a crazy tech every second would be pointless if it takes users from weeks to months to understand each of the things that it is developing.

>>105985081
Well yes. That's why I said if a strictly defined ASI is actually even possible or just fantasy. Guys like Zuckerberg with more money than god are willing to try for a moonshot to see. If they actually succeed, they win absolutely everything. That's worth even a 5% chance.

>>105984773
>brooooo AGI
Can't wait to plug into metaverse with my AGI assistant and spend my crypto on some dank ass NFTs!

I hate AI and want it to fail so I can see Cuckerberg crash and burn and laugh at his idiocy, God please let there be some sort of a wall for LLM's

>>105982669
He'll just give him some meta stock options. It's all make believe anyway.

>>105984998
>their competition because of non-compete clause clauses and enforcement by the FCC.
Just make the AGI glaze trump on xitter and he'll give them anything they ask for.

>>105985109
LLMs are just one branch of AI. Companies have been using ML models of all different kinds for years before LLMs became big. We're talking mid 2000's here.

>>105984813
>AGI just means they have a model that can think about the world at the same level as a human can

And that is worth trillions in itself.

Meta has 75,000 employees. Imagine them each getting paid $100k a year. That's $7.5B a year. You can not only replace all these workers, but also have 750,000 extra workers, operating at the same, or even above, your employee's level. 24/7. No breaks, no overtime, no anything.

>>105985141

Also, 75000 human employees who can't be arbitrarily improved. But as your AGI gets better, so do your 750000 AI employees. It'll compound fast, assuming it's even possible in the first place. This is basic takeoff question. Is it even possible? Is it slow or fast?

>>105985159
It's worth a read.

>>105985159
AI intelligence is exponential. How long were we stuck with Cleverbot, where it could barely hold a conversation and forgets thing from 5 minutes ago? That was AI for 12 years since 2008.

Then ChatGPT dropped 3 years ago, which was "okay". Then you have AI models that were hilariously bad, barely being able to generate a picture, and that "Will Smith eating spaghetti" video. That was just 2 years ago.

I'd expect no later than 2030 for AI to actually do some cool shit, like solving some mathematical problem that humans have struggled with for decades, or curing some deadly disease.

If it doesn't by 2030, then it's safe to assume it's hit a wall, and it would be equally interesting to see how companies have spent trillions on nothing. That's not to say AI wouldn't improve efficiency in some areas or make things easier for a lot of people, but it's no singularity.

>>105985189
>AI intelligence is exponential.

Maybe. There's no concrete evidence of this yet. Most people wouldn't even consider LLMs actually intelligent. Their appearance as intelligent has increased dramatically in the last couple years. Plenty of smart dudes out there who think it's not possible (LeCunn) and plenty of smart dudes who think it is (Sutskever).

Z: will you work for faceberg?
C: I think you mean Facebook? Is that right? If that's the case—Certainly! Working for Facebook's Meta AI division sounds like a really interesting opportunity. And let's not kid ourselves—this isn't just about getting my foot in the door. A whopping one billion dollars to boot will be a game changer for my career. Meta AI has what it takes to deliver. And let's be honest: it's not about the money—it's about the mission. Because Meta AI has what it takes to deliver. So, again—Sure! I'd love to work for Facebook. Is there anything else I can help you with?

>>105985141
>That's $7.5B a year. You can not only replace all these workers
You realize how idiotic that sounds, right?
Also, what difference would 7.5 billion/year make in the life of someone who's already worth 243 billion?

>>105985141
yeah except these AI workers cost more than actual workers

>>105985222
>comparing stocks to actual liquid money paid to employees

>>105985222
The ability of /g/ posters to be functionally retarded is limitless. If you have an AGI that is at human capability, then by definition yes, you would be able to save a shitload of money on salaries. You know, how machines have replaced millions of workers since the industrial revolution started?

And the loftier end goal here isn't "even more money". For someone who is already richer than god, being in control of an actual ASI is a completely different tier of goal.

>>105984773
>AGI is a winner-takes-all market. Once you've hit AGI,
Please, this is a serious tech board.

>>105985238
>stocks can't be sold

>>105985244
>this is a serious tech board

>>105985243
Learn to read, retard. I didn't say some jobs couldn't be replaced.
I'm saying replacing ALL jobs with AGI in a company of the size of Meta would be retarded.

>>105985249
>fucking retard doesn't know what happens when you start selling stocks

Holy fuck, my opinion on /g/ was bad already, but there's no sign of intelligence anywhere.

>>105985189
AI is no different from any technology. An innovation is made, it explodes, then levels off as you reach its limits.

>>105985251
Well, it's a tech board at least.

>>105985189
>2030
Youre off by a few years
https://ai-2027.com/