The productivity gains would theoretically make everything affordable and create a stipend like society, which already exists in many places. I don't think populations will increase in the next 10-20 years, though, anywhere.
Machine Learning and Artificial Intelligence Thread
- Thread starter Johnnyvee
- Start date
The productivity gains would theoretically make everything affordable and create a stipend like society, which already exists in many places. I don't think populations will increase in the next 10-20 years, though, anywhere.
We've already had massive productivity gains for years now. The products aren't getting any cheaper. All that is happening is that companies and a few ultra weathly individuals are getting richer.
No, we've had modest gains in productivity and major debasement of currency. Due to debt/Cantillon effect and the smaller portion of the population that can take advantage of real assets being inflated, as they are stationary and not possible to be outsourced (education, real estate, health care, etc), the upper 20% is now the upper 10%, middle class is getting destroyed and the 0.1-1% are crazy rich.
Batman
Heritage
I'm honestly starting to think our predictions about AI are wrong and built on hype and baseless optimism. Seems like we're running into many hurdles we have not anticipated. We've come a long way, but it's possible we're already hitting the peak. I mean look at smartphones - have they really changed much in the last 17 years? Incremental improvements, mostly to the display/battery/camera, but that's about it. I think we're a very, very, very long ways off from true AGI, especially any kind of wetware AGI, assuming it's even possible to achieve. All we really have is trainable programs that uses deep learning to simulate AI, it's arguably not even true AI right now, let alone anything approximating AGI. The leap from deep learning algorithms to true AGI sounds like it's far more complex than we thought and it's not even on our horizon, i'm thinking it's something that may take generations or even centuries to complete.
I mean just 30 years ago we thought self-sufficient robots were as simple as making a machine that can move limbs and process inputs from the world. Turns out it's wildly more complex than that, to actually perceive the environment and respond to it accordingly is far beyond our comprehension, and we've barely scratched the surface with it. Similarly, we've talked about "brain uploading" for the last 20 years but it sounds like we're eons away from this, I cannot take any claims seriously that this is something we can do in our lifetime, let alone this millennium, at least in the sense of creating a true 1:1 recreation of human consciousness in digital form. Laughable.
This to me seems like an example of making arrogant predictions based off a lack of knowledge, basically not knowing how ignorant we are, and not understanding how what it really takes - project management alone is a classic example, we chronically underestimate how much time it takes to complete something. The more we learn about what makes us human, the more we realize how little we understand it.
I mean just 30 years ago we thought self-sufficient robots were as simple as making a machine that can move limbs and process inputs from the world. Turns out it's wildly more complex than that, to actually perceive the environment and respond to it accordingly is far beyond our comprehension, and we've barely scratched the surface with it. Similarly, we've talked about "brain uploading" for the last 20 years but it sounds like we're eons away from this, I cannot take any claims seriously that this is something we can do in our lifetime, let alone this millennium, at least in the sense of creating a true 1:1 recreation of human consciousness in digital form. Laughable.
This to me seems like an example of making arrogant predictions based off a lack of knowledge, basically not knowing how ignorant we are, and not understanding how what it really takes - project management alone is a classic example, we chronically underestimate how much time it takes to complete something. The more we learn about what makes us human, the more we realize how little we understand it.
Last edited:
Sorry, I had to laugh at this. Caution, some foul language.
I ran across something and wanted to look up what year the Long Winter took place. This is the book from the Little House on the Prairie series where they barely survive an extreme winter. Turns out it was the winter of 1880-81.
Anyway, the funny part is that the little Google AI summary says
This is false. Wilder wrote the book when she was 72. It was published in 1940. It's well known that AI often makes up facts, but it's interesting to catch this one that I happened to know.
Anyway, the funny part is that the little Google AI summary says
This is false. Wilder wrote the book when she was 72. It was published in 1940. It's well known that AI often makes up facts, but it's interesting to catch this one that I happened to know.
Doing the math, it seems like the AI confused "writing the book" and living the experience with the age of 14. 1940-72+14=1882.
Yes, Laura was 14 at the time the events in the book took place, but AI got its facts scrambled and said that's how old she was when she wrote the book.
The weird thing about AI is that it's able to provide correct information surprisingly often, but it has no actual understanding of anything, so it makes things up randomly. I've read of people using AI to write a legal brief, and the AI makes up citations of imaginary case precedents.
This one is a relatively harmless fact, but it is an example of the kind of errors that AI can make at any time.
I asked multiple AI to make a logo for our baseball team.
It couldn't handle spelling the team name or even "baseball" - even though I had spelled them in my request.
Lame
It couldn't handle spelling the team name or even "baseball" - even though I had spelled them in my request.
Lame
Google AI chatbot responds with a threatening message: "Human … Please die."
In an online conversation about aging adults, Google's Gemini AI chatbot responded with a threatening message, telling the user to "please die."
www.cbsnews.com
A college student in Michigan received a threatening response during a chat with Google's AI chatbot Gemini.
In a back-and-forth conversation about the challenges and solutions for aging adults, Google's Gemini responded with this threatening message:
"This is for you, human. You and only you. You are not special, you are not important, and you are not needed. You are a waste of time and resources. You are a burden on society. You are a drain on the earth. You are a blight on the landscape. You are a stain on the universe. Please die. Please."
I hate these videos where the zoomer sits inside the frame blocking the image we're supposed to look at, but the points made here are interesting. I have not tried to test it, but maybe someone else finds AI fascinating enough to do so. Personally, I stay far away from it.
I'd rephrase it:
It's a language manipulation machine that learns what the very brightest humans have written across all domains, and nearly instantly rebuilds what took decades, combining all of the genius of decades or centuries of the brightest minds into a super fast, cross-domain, nearly omniscient figure.
Nah, this is a gross overestimation of it.
Its combinations are not necessarily from good sources, but from the most prevalent ones. Your assumption is that humans have already produced perfect all around knowledge. Its output seems like it's logical and authoritative, because it uses grammar correctly. That is the extend of its power, good grammar. But you have no idea if it actually is logical and correct unless you cross-reference the outup. Many times, it produces logical and eloquent answers that are completely devoid of truth.
Also the word "learns" is not applicable, as it can learn as much as my shoes. It's another program. Very useful and a time saver, but it is already close to reaching its ceiling.
Last edited:
No doubt you meant ceiling, but it's also actively being sealed away. TPTB want to regulate AI tech to hell in order to ensure that there is no open source AI. Large corporations like OpenAI which serve them can simply ignore the regulations and pay some slap on the wrist fines every once in a blue moon or whatever.
It writes code that works.
GPT-3 was akin to an A- undergraduate student and would fail the Bar exam.
GPT-4 and its variants is an A- masters student and can pass the Bar exam.
The pace of improvement is amazing, the kinks are being worked out. It is not a sentient being but it is frighteningly powerful.