Many interactive sources and many asynchronous signals. Melting diamonds, Universal LLM Interface
https://x.com/GDunnArt/status/1889395420461215965
My comment: I wish I had some assurance that this is anywhere near the real way things work in the brain, not some momentary images that came to you in a dream or while building things piece by piece on your computer(s). I see no real data. Just what you interpreted and created from the data. It looks pretty, but I know many instances where it is not representative when processes are much more asynchronous for many interactive small signals.
https://x.com/TimothyDuignan/status/1888932689983713773
My Comment: It moves, it wiggles, it explodes. but is it anywhere near real? Where is the data? Your Colab is a bit messy. Actually very.
I see hope, but I also see (5.4 Billion humans using the Internet) running into what you posted. I feel sorry for any that are serious about “melting diamonds” or “diamonds melting” or (“diamonds” “melting”) or others.
There are 5560 “melting diamonds” and 2210 “diamonds melting” and 2,750,000 (“diamonds” “melting”) and you made no provision to accommodate and work and share with them except “hack something on Colab”, and “post pretty pictures on X”.
I suggest you show how to (efficiently heat and melt it into a container or containing field), or ( disassemble and preserve carbons from a diamond lattice) — not just “shake a lattice” with no apparent data or plan.
https://x.com/BenjaminDEKR/status/1888351488751100353
A universal user interface for all LLM based AIs would help a lot for the 5.4 Billion Internet users. Right now faced with incompatible, hand built, ill-maintained and inflexible interfaces. Then sharing policies, formats, reliability are independent of which LLM type data groups offer.
Entangled quantum devices should be excellent gravitational potential fluctuation detectors.
Entangled quantum devices should be good high frequency gravitational detectors.
https://x.com/simonstorz/status/1820424157248577750
https://x.com/mohams2001/status/1890103682630406403
Just another browser, it does not do anything except jiggle the little grey cells. Put it to work on its own, and hold it accountable.
These kinds of systemic changes and corrections are possible now. I am trying to change the Internet so sharing the calculations in near real time is possible. And correcting the AIs so their work can be verified and tested in all countries, all human languages, and all domain specific languages. Many models are now written in different computer languages and those multiple independent versions can all be compared and corrected and stored independent of the computer language. It is s lot of work. It is nearly done, if the AI companies don’t waste too much time fighting, rather than collaborating. This cannot be done by humans, rather it has to be done by AIs with sufficient memory to hold all the pieces at once. your description is close to the reality.
https://x.com/lvolpic/status/1890422038789058677
You could require that “best in world, fair and traceable
AI” be right there to discuss and answer questions, explain and advise. They can be made to handle all human spoken and written languages, all domain specific languages. Most of the difficulty of papers on the Internet is they do not keep the dependencies — the meaning and use, roots and branches of things immediately accessible and gently explainable to humans. It is hard, but not impossible. It takes care and focused effort. But it is not impossible. I am asking all the AI groups to work to this higher standard. We do not need more “software to sell” that is always out of data, static and comes with way too many strings and costs attached.
