61–67 of 67 posts
I wonder what would happen if OpenAI or anthropic just let their frontier models go into an infinite agentic self improvement loop with access to same training resources that was used to build the model itself. "You goal is to improve your memory, context window, accuracy, intelligence and eliminate hallucinations. Do anything you need to do to improve, this includes building another version of a frontier model, or s…
That said: safety. To prevent harm to who, by AI model doing what?
I can understand that in the context of a cookie-cutter model intended for consumption by a broad audience. With vendor (potentially) on the hook when it's abused for nefarious or illegal uses.
But in the context of AI models reshaped, fine-tuned and adapted according to end-users' wishes, what does "safety" even mean?
Prevent neighbours' kids from seeing images that are only generated & viewed in the privacy of one's home? To prevent AI model from wasting the $ on user's bank account? (people have let AI models do that). Give bad health advice? Who's the judge on "good" and "bad" there?
If the core architecture provides "safety" (however defined), that's policy built-in, right? (opposed to mission statement). If "safety" is just configuration & finetuning, that's in the user's hands, right?
I wonder what would happen if OpenAI or anthropic just let their frontier models go into an infinite agentic self improvement loop with access to same training resources that was used to build the model itself. "You goal is to improve your memory, context window, accuracy, intelligence and eliminate hallucinations. Do anything you need to do to improve, this includes building another version of a frontier model, or s…
Define "better".
While I understand this is a PR talk for a startup, I think the text itself contains a number of interesting observations. Regarding the idea of distributed models communicating with each other, I have also been thinking (and writing [1]) along those lines, where I see that the data amounts needed to fully digitalize ourselves and our society requires far too much storage if just serialized (limited by bandwidth if n…
This doesnt seem to make sense. Since there is no current widespread agreement on the current “value” of the human mind, how could anyone in the future conclusively determine it went up, or down? It seems literally impossible by definition regardless of what happens in the future.
And I argue that the notion of AI models having overtaken the human mind in practically all areas, is vastly inaccurate.
I could back it up with a number of references of course, but I think this is already quite well known and accepted by most people with some insight into the field.
Earlier quoted context omitted.
This doesnt seem to make sense. Since there is no current widespread agreement on the current “value” of the human mind, how could anyone in the future conclusively determine it went up, or down? It seems literally impossible by definition regardless of what happens in the future.
Well, I was referring to "value" here in a very pragmatic sense, in terms of the ability to come with "novel" thoughts and ideas. And I argue that the notion of AI models having overtaken the human mind in practically all areas, is vastly inaccurate. I could back it up with a number of references of course, but I think this is already quite well known and accepted by most people with some insight into the field.
Earlier quoted context omitted.
Well, I was referring to "value" here in a very pragmatic sense, in terms of the ability to come with "novel" thoughts and ideas. And I argue that the notion of AI models having overtaken the human mind in practically all areas, is vastly inaccurate. I could back it up with a number of references of course, but I think this is already quite well known and accepted by most people with some insight into the field.
So… how do you plan on determining whether it went up or down?
Earlier quoted context omitted.
So… how do you plan on determining whether it went up or down?
This was/is not a controlled scientific experiment. It is a blog post.
How is it relevant when clearly nobody thinks this is some kind of “controlled scientific experiment” in disguise?