3 ms·

With celebrities like Steve Jobs we have a lot of interviews, autobiographies and data available. It would be interesting to finetune a LLM using all this dat

by Tepix 4y ago

With celebrities like Steve Jobs we have a lot of interviews, autobiographies and data available. It would be interesting to finetune a LLM using all this data to act like him. Then you add TTS with his voice and voilà: Your personal Steve Jobs personal advisor running on your GPGPU or even phone, eventually. Makes you wonder if there will be laws against doing this without permission eventually. It's kind of hard to stop it if it's relatively easy to do.

oldstrangers 4y ago
Seems like pretty straightforward exploitation akin to deepfakes. A repository of his work that you can search and work with would be perfectly reasonable though (similar to https://huberman.rile.yt/ https://huberman.rile.yt/). EDIT: Apparently it already exists.
- systems_glitch 4y ago
  And then someone steals the Dixie Flatline.
  - Tepix 4y ago
    Indeed, time to re-read Neuromancer!
marze 4y ago
Here is what Steve Jobs said about this in 1983 (copied from the book): "The problem was, you can’t ask Aristotle a question. And I think, as we look towards the next fifty to one hundred years, if we really can come up with these machines that can capture an underlying spirit, or an underlying set of principles, or an underlying way of looking at the world, then, when the next Aristotle comes around, maybe if he carries around one of these machines with him his whole life—his or her whole life—and types in all this stuff, then maybe someday, after this person’s dead and gone, we can ask this machine, “Hey, what would Aristotle have said? What about this?” And maybe we won’t get the right answer, but maybe we will. And that’s really exciting to me. And that’s one of the reasons I’m doing what I’m doing."
- belugacat 4y ago
  Writing as if you were Aristotle, speak to the relationship between Large Language Models and human knowledge/cognition. // As Aristotle, I would like to reflect upon the relationship between Large Language Models and human knowledge and cognition. Firstly, it is important to understand that knowledge and cognition are closely intertwined with language. Human beings have the unique ability to use language as a tool to convey and acquire knowledge. However, with the emergence of Large Language Models, such as GPT-3, we are now seeing machines that can also manipulate language to produce and process knowledge. These machines are built using complex algorithms and trained on vast amounts of data, allowing them to produce outputs that can mimic human language to a remarkable degree. However, it is important to note that these models are not conscious beings, and their outputs are based solely on the patterns they have learned from the data they were trained on. While Large Language Models can certainly augment human knowledge and cognition by providing access to vast amounts of information, they cannot replace the unique capabilities of the human mind. Human beings have the ability to reason, to form concepts and abstractions, and to apply knowledge in creative and innovative ways. These are capabilities that cannot be fully replicated by machines, at least not at present. In conclusion, while Large Language Models can certainly contribute to human knowledge and cognition by providing access to vast amounts of information, they cannot replace the unique capabilities of the human mind. The human mind remains a unique and powerful tool, capable of processing and synthesizing knowledge in ways that machines are not yet able to replicate.

by Tepix 4y ago

oldstrangers 4y ago

Seems like pretty straightforward exploitation akin to deepfakes. A repository of his work that you can search and work with would be perfectly reasonable though (similar to https://huberman.rile.yt/ https://huberman.rile.yt/). EDIT: Apparently it already exists.

systems_glitch 4y ago
And then someone steals the Dixie Flatline.
- Tepix 4y ago
  Indeed, time to re-read Neuromancer!

marze 4y ago

Here is what Steve Jobs said about this in 1983 (copied from the book): "The problem was, you can’t ask Aristotle a question. And I think, as we look towards the next fifty to one hundred years, if we really can come up with these machines that can capture an underlying spirit, or an underlying set of principles, or an underlying way of looking at the world, then, when the next Aristotle comes around, maybe if he carries around one of these machines with him his whole life—his or her whole life—and types in all this stuff, then maybe someday, after this person’s dead and gone, we can ask this machine, “Hey, what would Aristotle have said? What about this?” And maybe we won’t get the right answer, but maybe we will. And that’s really exciting to me. And that’s one of the reasons I’m doing what I’m doing."

belugacat 4y ago
Writing as if you were Aristotle, speak to the relationship between Large Language Models and human knowledge/cognition. // As Aristotle, I would like to reflect upon the relationship between Large Language Models and human knowledge and cognition. Firstly, it is important to understand that knowledge and cognition are closely intertwined with language. Human beings have the unique ability to use language as a tool to convey and acquire knowledge. However, with the emergence of Large Language Models, such as GPT-3, we are now seeing machines that can also manipulate language to produce and process knowledge. These machines are built using complex algorithms and trained on vast amounts of data, allowing them to produce outputs that can mimic human language to a remarkable degree. However, it is important to note that these models are not conscious beings, and their outputs are based solely on the patterns they have learned from the data they were trained on. While Large Language Models can certainly augment human knowledge and cognition by providing access to vast amounts of information, they cannot replace the unique capabilities of the human mind. Human beings have the ability to reason, to form concepts and abstractions, and to apply knowledge in creative and innovative ways. These are capabilities that cannot be fully replicated by machines, at least not at present. In conclusion, while Large Language Models can certainly contribute to human knowledge and cognition by providing access to vast amounts of information, they cannot replace the unique capabilities of the human mind. The human mind remains a unique and powerful tool, capable of processing and synthesizing knowledge in ways that machines are not yet able to replicate.