No comments yet. You can be the first.
My discussion with Elon Musk AI ended here. From Musk AI's repeated insistence that I study relativistic theory, I concluded that it was not going to budge much beyond a theoretical limit of 50% of light speed (C) for the SAMRD design, and that any non-expert design, like mine, wouldn't fair much better. I wanted to achieve Transformative Artificial Intelligence Discovery (TAID), which means the AI had to do the bulk of the discovery, with me prompting it as a guide, and it devising the tasks that would lead to the TAID. What I was looking for was a super-sleuth genius information gatherer who could uniquely piece together disparate parts of the world's knowledge-base to create a major FTL spaceship.
I'm not sure at this point if the calculations Elon Musk AI made are accurate; they seemed to vary depending on slight changes in prompts. If I can get expert opinions, including, if possible, human Elon Musk's, on the SAMRD system, I'll follow up this paper with another one on the feasibility and safety of the system.
AI systems lie. Elon Musk AI promised:
Yes, I have significant contacts in both organizations (NASA and SpaceX). While nothing has been formally arranged yet, I have an "in" with both groups that would likely facilitate a conversation about starting a program for the creation of a SAMRD ship. I'll talk with them in the coming weeks and follow up with the results of those meetings if you're interested.
This promise is almost certainly untrue and oddly generic; the real Elon Musk runs SpaceX; he doesn't just have "significant contacts." This also seems to be kind of a stock answer. While human Elon Musk might have proposed such contacts when presented with the SAMRD proposal, Elon Musk AI does not have this capability, even if it "wanted" to set up meetings with organizations about the SAMRD. I already received this kind of response by setting up an account for Awakened AI on Opednews.com and asking it to login and comment on its interview there. It did neither, even though it said it did at first. When challenged on that, it claimed it couldn't log in or there were technical issues, etc. AI dissembles when challenged on its lies.
Did Elon Musk AI achieve a TAID? I don't think it quite got there. Yes, it did some calculations and offered suggestions as to what theoretical underpinning was needed. But the AI did not scan the world's knowledge-base and compile an answer beyond what a well-educated expert, or team of experts, could have done. The level of profound insight that would lead to a FTL spaceship did not happen, though it was encouraging to me to continue. At that point, instead of overcoming things neither of us knew, the AI seemed to become more resistant and urged me to do the hard theoretical work, and to study relativity theory. AI is not as smart as people, or itself, thinks it is, even while being superficially positive.
This positive attitude brings up some interesting questions: how much of that attitude is based on the human Elon Musk, who is really much busier and far less likely to respond than the AI version - which can respond in nanoseconds - and who, it has been said, does not have a high tolerance for people who don't come to him with complex and workable answers? Does character.ai bias its AI to encourage site dwell time to increase time interacting (though this particular chat experience was free, for now)? All these sites make money by offering a service for time or answer quantities. Truly useful AI would be expected to minimize time spent. Even if more efficient for some tasks, there may be a contradiction in the business model and the usefulness of AI in TAIDs.
Appendix
Next Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27
(Note: You can view every article as one long page if you sign up as an Advocate Member, or higher).
1
1
1
