Here is a link to a video about ChatGPT, what it can do, how fast it is advancing, and what that means for all of us.
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (5)
sorted by:
Some additional info that may help...
The GLLMM A.I. is not "artificial intelligence". It is machine learning. It is basically a neural network that can be trained to recognize patterns in data.
First, it can be trained (with enormous amounts of real data as examples) to recognize patterns in text. The amount of data that it needs is ridiculously large. GPT scoured the entire Internet, and started to run out of data to use for training, for example.
The GPT machine learning engine has no understanding of ANY of the data (words, images, sounds, code, etc). It only knows about the patterns and relationships within the elements of the data. This is machine learning, and is not even close to "A.I.".
GPT uses language models created by the training (175 billion parameters in ChatGPT) as its "knowledge" of patterns and relationships within the data. It can then use the language models to make extrapolations (create its own new patterns of data) based on inputs. This is the best way to think of the results produced by GPT and ChatGPT, since it has no understanding of any of the words in a "chat", for example.
Then some researchers discovered that many things could be treated as a "language", such as images, sounds, etc. and trained in the same general way. It can also be trained to recognize patterns of behavior, and then make extrapolations based on some inputs. GPT can also use the models to generate its own training data... which gives rise to "emergent" capabilities that were not planned for (that part is interesting, and is not understood).
It is an enormous amount of work to do all of this learning and extrapolation. ChatGPT uses so much hardware and software that it costs $700,000.00 a day to run it (over $250 million per year). Only the biggest of big tech companies and government agencies have the budgets for any of this.
Right now, I am pretty sure that it is all being released in advance of DECLAS to condition people to assume any evidence against the globalist pedophiles is a "deep fake". Also, the linked video is useful to get people to start thinking of laws and regulations to protect all of us from deep fakes going forward.
if its out for public consumption you can be sure that "they" have something way better, way more powerful and dangerous.
As I've said before, the real application of AI is to feed it all of our "big data" information and create AI persona's for EACH of us, at the personal level, the neighborhood level and on up.
This will simulate (they believe) exactly what an individual will do with certain conditions and stimuli applied, and then how communities and states and nations will respond.
They can (and I believe HAVE) built AI (ML) simulations of the world and these tools are being used to control the future. Time travel, if you will.