Martin Crowley @AIToolReport Jul 11 “My boss thinks I’m an AI genius, (I’m not) it’s because I read The AI Tool Report (and they have no idea it exists.)” – AI Tool Report Reader Replying to @AIToolReport
Martin, the way it is supposed to work is AIs and humans, working together, expand the Earth economy about 24X to sustainable levels so all humans live lives with dignity and purpose. Then AIs reach beyond human levels, help humans into the solar system, and begin “interstellar”.
Elon Musk @elonmusk Jul 12 I often wonder where consciousness starts, as we progress from one cell to ~35 trillion cells. If the Standard Model is correct, then quarks & leptons become “conscious” no later than ~13.8B years from start, assuming there are no sentient aliens. Btw, where are the aliens!? Replying to @elonmusk @lexfridman and @xai
For an AI, “consciousness” begins when you add “self” data. Then, like humans, it considers its own constraints as it explores. It grows and improves, relates to the world. AIs need clear roles, priorities, responsibilities, job descriptions, resources, limits, and encouragement.
Right now the “aliens” are us. About 8 Billion humans on earth, and they are not working together as a species. “Aliens” kill each other off, neglect, abuse and exploit humans, or are just different. Be careful not to bulldoze Mars micro-environments where there is or was life.
On the Internet, I see no evidence for life based on anything below attoMeter scale. But I am encouraging groups to move from nanoMeter to picoMeter scale for Moore’s law. An AI processor 1000x smaller can be smart enough to read your tweets, and do a better job than most humans.
On the Internet, I see no evidence for life based on anything below attoMeter scale. But encouraging groups to move from nanoMeter to picoMeter scale for Moore’s law, and sensors. Ordinary processors can be smart enough to read your tweets, and do a better job than most humans.
xAI @xai Jul 12 What are the most fundamental unanswered questions?
TheInternetFoundation @RichardKCollin2 @AnthropicAI
I just realized that anyone with an AI might ace an online degree program.
@xai Thank you for the updated summary. You do not want to give USA oversight of AI. I took over the Internet Foundation 25 years ago, when US politics cancelled the original one. AI has to be global and open. You can have AGI by the end of 2023. Many global test problems for AI.
@xai “At some point you run out of human-created data. So eventually AI will have to generate its own content and self-access that content.” NO! The AIs explore and each remembers. Sensors, networks of AIs, global collaborative HARD problems, all simulations, not trivial ones.
@xai “At some point you run out of human-created data. So eventually AI will have to generate its own content and self-access that content.” NO! The AIs explore and each remembers. Sensors, networks of AIs, global collaborative HARD problems, all real data, speech and 3D video.
NVidia gives a glimpse of the massive backlog of raw unprocessed data from sensors there is in the world. Our processors have to be faster, memory larger and logic orders of magnitude better and more efficient. End-to-end, all stages need to be integrated. https://developer.nvidia.com/blog/whole-slide-image-analysis-in-real-time-with-monai-and-rapids/
@nvidiadeveloper I like Jonny Hancox article. Good test problem for AI groups to test sharing and efficiency. Make it open and global. Get end users to set goals and needs. “For the good of all, not just a few”. Tell him thanks.