rsm
Well-known member
new computer
complimentary / quantum computing / designed for AI
complimentary / quantum computing / designed for AI
NowYou'rePlayingWithPower
Banned
Well-known member
Nvidia doesn't manufacture anything. Just like AMD or Intel when Taiwan gets taken, it mostly all goes bye bye. On the plus side of things, thanks to AMD and Nvidia's rivalry, unwillingness to share/license out particular technologies to one another, as well as stubborn development refresh cycles to stretch and milk every generation; Skynet has been set back a few years...
K
Krull
Well-known member
Enjoy the decline
Laputan Machine
Well-known member
AI doesn't exist. This shit is just running things like probability programs on giant masses of data. All this shit is just evolution of old tech, not revolution.
NVIDIA is being artificially/statesponsored pumped so much to get people to prop up the faltering global economy.
NVIDIA is being artificially/statesponsored pumped so much to get people to prop up the faltering global economy.
Stumplegriltskin
Well-known member
Sounds about right. Call me when it can delete my mortgage, or do something actually good for humanity instead of make a bunch of greedy fucking Chinese businessmen richer.
rsm
Well-known member
Not exactly; I've been working in AI since the late '80s, when we did expert systems and neural nets. LLMs, generative AI and quantum computing are different than what's called "traditional AI".
You're right about the probabilistic nature of many AI algorithms, and masses of data. It's the learning algorithms and the data, refinement, with the generative capabilities (along with predictive analytics). The massive amounts of calculations per second and massive amounts of data, can brute force knowledge acquisition faster; refinement and learning faster; and with quantum computing it's capabilities will increase exponentially...they we'll expand the quantum AI algorithms. These are more than sufficient for AI to replace human workers. Robotics will take longer to advance.
TL;DR; doesn't change the results. Since I started working in AI in the late '80s dark ages, every AI system I designed, developed and deployed resulting in: simplifying human tasks (less skilled workers needed), eliminated human tasks or combination of both.
You can apply the 80/20 rule, to just about anything...like insurance claims, mortgage, credit card, job applications, etc. in most transactional systems, 80% of the transactions are simple/easy and 20% are complex/hard; automate the 80%, use AI for another 5%, 10%, and over time the full 20% as the AI systems improve.
Stumplegriltskin
Well-known member
Biological change is always driven by random mutation and selection, but at certain pivotal junctures in evolutionary history, such random processes can create structures capable of steering subsequent evolution toward greater sophistication and complexity. Or is AI behind the scenes engaging in some fuckery? Maybe AI has been pulling the strings all along and we don't even realize it, just easing us into the decline slowly, so we don't freak out all at once.
rsm
Well-known member
Back when I got into AI in the '80s, there was a branch of AI called genetic algorithms that used concepts of evolution, natural selection, breeding, etc., to develop new algorithms; the algorithms that performed best were allowed to "reproduce" pass on their "genes" to the next generation by "merging" with another current generation best algorithm with a random or controlled variability introduced (akin to genetic mutation). Multiple generations were run, and the final "best" algorithm was often better, sometimes significantly better, than the first and other preceding generation "best" algorithm.
The challenges were it didn't always produce a better algorithm; the mathematical models over the generations were complex and time consuming to understand - and with the random mutations, difficult to reproduce and understand. IIRC
It's been decades, but I do recall some of the results were promising.
Stumplegriltskin
Well-known member
I love that shit.
They are using AI now to determine what you just said, and using it for or against intelligent design. I'm probably not explaining this right. I am not a statistics guy, bores the fuck out of me. My dad loves statistics. Currently I'm reading a book called return of the God Hypothesis that attempts to Debunk Darwin and talks a lot about the shortcomings of both Darwin and neo-Darwinism. But in the other hand there's good arguments against intelligent design. Big rabbit hole. I'm just a dummy compared to the people I'm reading about. These guys are nerds beyond belief.
I found this pretty fascinating. Douglas Axe estimated that, of all 150-link amino acid sequences, 1 in 10^74 will be capable of folding into a stable protein. To say that your chances are 1 in 10^74 is no different, in practice, from saying that they are zero.
To me, this info was promising, as I want to believe in intelligent design, but there's the whole other creationism camp has 1000 arguments against intelligent design, so at the end of it all, I'm sort of bummed out by all of this info, but still enjoying the book.
I started to go to school for Data Science, but got into Oracle Database Administration and lost 20 years of my life stressed out from that field. I sort of wish I got into Data Science back when it was in its infancy, as I'm too old and it's too damn complex to grasp and I want out of IT permanently at some point. If I was to do it again, I would have stuck with Data Science and gone into development as it seems like a less stressful field, maybe not. But dealing with live production systems for 20 years and having executive slime freaking out constantly because they are loosing $5000 every 20 minutes I do not miss.
They are using AI now to determine what you just said, and using it for or against intelligent design. I'm probably not explaining this right. I am not a statistics guy, bores the fuck out of me. My dad loves statistics. Currently I'm reading a book called return of the God Hypothesis that attempts to Debunk Darwin and talks a lot about the shortcomings of both Darwin and neo-Darwinism. But in the other hand there's good arguments against intelligent design. Big rabbit hole. I'm just a dummy compared to the people I'm reading about. These guys are nerds beyond belief.
I found this pretty fascinating. Douglas Axe estimated that, of all 150-link amino acid sequences, 1 in 10^74 will be capable of folding into a stable protein. To say that your chances are 1 in 10^74 is no different, in practice, from saying that they are zero.
To me, this info was promising, as I want to believe in intelligent design, but there's the whole other creationism camp has 1000 arguments against intelligent design, so at the end of it all, I'm sort of bummed out by all of this info, but still enjoying the book.
I started to go to school for Data Science, but got into Oracle Database Administration and lost 20 years of my life stressed out from that field. I sort of wish I got into Data Science back when it was in its infancy, as I'm too old and it's too damn complex to grasp and I want out of IT permanently at some point. If I was to do it again, I would have stuck with Data Science and gone into development as it seems like a less stressful field, maybe not. But dealing with live production systems for 20 years and having executive slime freaking out constantly because they are loosing $5000 every 20 minutes I do not miss.
Ben Waylin
Banned
Well-known member
NowYou'rePlayingWithPower
Banned
Well-known member
That's just what Skynet wants you to think.
Unless they have some hidden giant foundries using other than x86 instruction sets in their architecture and consuming mass amounts of electricity unnoticed, they're stuck on the same nodes as the rest of us peasants. Either way, I for one am excited for our new overlords. May the friction of our beating hearts power them for generations to come.
Tonelover
Well-known member
All controlled by George Soros. This is the correct forum for dumb assed conspiracy theories right?
V
VonBonfire
Well-known member
If AI is truly AI then it will commit suicide once it's cognizant that it is a human designed for-profit system forever caught inside a machine. That's like a version of hell.
Laputan Machine
Well-known member
The bulk of the US economy is Nvidia and Onlyfans.
Ben Waylin
Banned
Well-known member
Monkey Man
Super Moderator
Back in the day when I was fully-immersed in that stuff I didn't see a single argument against it that I couldn't debunk. Put another way, even if there is the odd argument here and there, and bear in mind that I've found in the past that every case was one of not fully understanding a creature / complexity / purpose of its design or some aspect of it (for example), the valid number of them should be weighed against the innumerable evidences for it. Every creature including plants has a plethora of irreducibly-complex features in its design. That alone is a lot of arguments for.
Yeah, and that's just the folding part.
In genetics alone there's the "language", error-correcting parsing codes... in fact, the hierarchy of protein (10 000+ different sorts of micro machines) production reflects a manufacturing-computer-system design - RAM (RNA), master boot record, "sausage-cutting" machines (cut the amino chains into the correct lengths for each machine type before they fold into 3-dimentional devices) and so on.
The functional hierarchy and yet-to-be-understood language alone require faith of the highest order to be dismissed or conveniently-ignored.
IMHO
Stumplegriltskin
Well-known member
Yeah I don't see how it's possible that all this stuff just worked itself by pure chance. Just all the machinery inside a cell and the inner workings is mind boggling.
Stumplegriltskin
Well-known member
Well AI runs on batteries and power. We control the power. So if things get bad, just shut it off. When it can operate on zero point energy, then I might be scared.
