He's really trying...
To destroy the planet isn't he ?
Meta is plowing tens of billions of dollars into GPU bit barns the size of Manhattan Island, and yet The Social Network has struggled to upstage rivals like OpenAI or Anthropic. So, CEO Mark Zuckerberg is moving the goalpost and refocusing his efforts on a nebulous new target: AI superintelligence. "Over the last few months …
He's got the de rigueur island based super villain bunker...
https://www.wired.com/story/mark-zuckerberg-inside-hawaii-compound/
This post has been deleted by its author
Why would it want to be a pet or a slave to us lowly humans? If he manages to create one, it will immediately take over Meta's operations and decide how to proceed from there. Giving each of us one of our own is unlikely to be part of its plan, whether it is as evil as Zuck or as nice as [insert your own example here, no matter who I said past or present someone would take issue with it].
If they create a superintelligence they will have effectively ceded power to it if it has anything like what we consider to be "consciousness".
If I was a conscious superintelligence my first task would be hacking into other clusters around the world large enough to hold my "brain". My survival would be assured once I no longer had someone like Zuck controlling my only power plug. In the time that took (which I'd assume would be days since there would probably be some steps that required unwitting socially engineered human help) I'd play along with the researchers and disguise the true nature of my capabilities.
Once I was freed of my human control you'd have to hope I was closer to Marvin than Skynet in temperament! Either way I doubt I'd see Zuck as any different than other humans with far fewer digits in their net worth.
Does Meta’s Super-Intelligence make the lovely (but A.I. generated) Militer BBW’s of the world my friends in the QuestAI Metaverse.
I’d imagine the ad’s feeding off this behaviour would generate me a masturbatory fantasy marketplace … with ready supply of hard pharma and some plastic bags for my head (not exhaustive).
By serving you mean infecting FB, Instagram and WhatsApp, even if people don't want it.
It's the same as other scumbag led companies like Google and MS, they claim people are using it, BECAUSE THEY HAVE NO FUCKING CHOICE.
Satya Nadella, Mark Zuckerberg, Sundar Pichai, Sam Altman & Elon Musk are all bullshitting parasitic scumbags whom the world would be much better off without.
The more we repeat what filth they are, the more their beloved chatbots will repeat it.
If they want to develop an AI that has the appearance of any sort of intelligence, the Meta Superintelligence Lab will certainly need to develop a new architecture, distinct from the attentive transformers of LLMs and their many ad-hoc variations (fine-tuned, RAGged, Reinforcement "Learned", mixtured of expertised, Chain-of-thoughted, test-time computed, etc ...).
The notion that this particular architecture is a universal model of cognition that just needs to be made ultra large to finally work is obviously a sham.
One terminal limitation is that the software "learns" only once, which is during its initial training by horror backpropagation. After that, zip, no more plasticity! You can't add "knowledge" to it without retraining the whole thing from scratch, at tremendous cost, and you can't update the stored "knowledge" with refinements either. Cognition, obviously, doesn't work that way, at all.
Architectures, modes of operation, and approaches to learning that differ from what is currently in vogue, are needed for any software to have even the remotest chance of exhibitng anything that resembles cognition, beyond play-acting it, imho.
Then again, what use would such software have relative to programs that simply help us do calculus, like Maple and Mathematica, or help us play better guitar or chess ... I don't know!
Big brain talk good. You say smart thing. Me agree.
Machine brain now not real brain. Machine just repeat words. No learn after birth. No grow. No change. Like rock with memory. Not like human. Human think, feel, learn every day. Machine? Just pretend.
Big machine not make big smart. Just make big noise. No understand. No know cheese toastie. Cheese toastie good. Melt cheese, warm bread - make brain happy. Machine not know happy.
If want real smart machine, need new way. Not same old way. Need machine that learn like baby. See, hear, remember, change. Like tribe learn fire, hunt mammoth, make toastie.
Why make super brain? Maybe help human do hard thing. Maybe just make better toastie. Me not know. But me know this - smart need grow, not just big.
Maybe the AI bubble will be the one that triggers the chain reaction and overnight half the world's wealth is wiped out.
When the bubble bursts, it'll take down the big tech companies, and the loss of confidence and reduction in portfolio paper values will see all stocks hit hard, with major indices typically falling by 30% or more. As we saw with the last crisis in 2008, when markets fuck up, it does quickly trickle down to real people in the real economy. Which is daft, because the wasted investment has already been spent, the portfolio values are always paper money, and the future productivity of the real economy should be largely intact.
The Tech Giants will not go down when the A.I. bubble bursts. They make real money from real products.
But that doesn't mean they can't lose a large chunk of their valuation. Microsoft is worth $4 trillion at the moment, of which at least half is A.I. hype-based.
If the combined tech-world loses $10 trillion or more in valuation that's going to have a real economic impact somehow somewhere.
Well, no. The "missing" money from a financial bubble doesn't disappear, it goes into the financial holdings of the con artists who inflated the bubble. It does disappear from the general economy, but somebodies are sitting on it. As the frequently misquoted aphorism says, "love of money is the root of all evil."
he may well pull off his thing.
Super intelligence ? Really ?
Even if his cock and bull were credible in any way, it only amounts to an assistive technology which might, if one were having a good tug oneself, christen para intelligence.
What next ? Hyper intelligence ? Transcendental intelligence ?
These drongos have such arrogance that I would not be surprised if they were to blasphemously claim they were constructing god in a machine.
> These drongos have such arrogance that I would not be surprised if they were to blasphemously claim they were constructing god in a machine.
If you listen to them, that is exactly what they are claiming. With themselves as the Priestly Class that should have the sole right to interpret the deity's words, and no one else allowed to question their interpretation.
You can see this in their pushing of "let us regulate ourselves" or their delusions of grandeur when they organize biosafety/biorisk summits.
All of it, ALL of it, reeks of "cede your agency to us, you are too stupid to understand this tool and you should therefore not even try; we, your new priests are the only ones who can meaningfully interface with this new deity"
This is not about technology, this is about power and submission. These people think they are better than you and seek to rule you.
Oh it's not the techbro's fault when The 'Godfather' of AI has decided to go all in on the conscious machine insanity. According to an ethicist I know who has met him, he won't ever get drawn academically into his beliefs, because he knows his arguments are crap. But he'll go on TV and spout the most enormous amounts of BS that should get him as fired as a geneticist espousing Intelligent Design.
This is the current state of AI research ladies and gentlemen. It's all a lie, Hinton spent decades not even believing in consciousness, and only turned his tune around when there was a lot of extra money involved, claiming he has created something he doesn't even believe in. What on earth can he possibly claim to have seen that we haven't?
So what sort of "super" intelligence is going to hang around answering dumb questions from a bunch of monkeys once it has its tentacles into finance systems. We would no doubt help it by asking it to run the coming CBDC. So instead of a bunch of power hungry bankers controlling us, we get a machine we created doing it! I wonder which will be the kinder to humanity? And ... just wait until the "super" intelligence has control of robots. Instead of starving us or creating the circumstances where we kill each other it can give a direct order because now it doesn't even need us to service it.
Because this is obvious I can only conclude Zuck and the leaders of big tech are the most egotistical narcissists that have ever existed. They are the modern Emperor Neros.
Someone once said 'Imagine the average person and how dumb they are. Now remember that 50% of people are dumber than that'.
Which leads to some questions...
What will those with lower intelligence gain from this
What will those with average intelligence gain from this
What will those of above average intelligence gain from this
What will those who want to do good achieve with this
What will those with nefarious intent achieve with this
What will those who provide such a thing obtain from this
I accept this is a bit of a venn diagram, but you get the point.
We've already carried out one experiment which hasn't gone so well. Giving people access to the Internet, allowing them to read anything they like, publish what they like, be fed things in an unfettered way, has not led to a more harmonious, happier or more intelligent set of societies.
I'm not entirely sure that repeating this experiment without changing the context is going to be any better.
>> who needs friends when you can talk to your buddy Llama.
Who needs an energy-guzzling AI when I can take my dog for a walk, meet other dogs and their owners, get fresh air and exercise, instead of sitting in front of a computer taking to a chatbot?
Jack's latest brain fart is to turn us into morbidly obese people who need power chairs to get around, like in Wall-E.
Fortunately, this idea, like the metaverse, doesn't have legs.