John I've investigated how information could be reliably computed as energy, not only at 300K or 20 watts, but in a multidimensional topology. Not against current science alone, but with a nextgen scientific approach.
If computable at Shannon entropy, possibly harvestable into incremental capacitors. I soon discovered "theory" and "narrative" cite some laws to argue it is impossible to do - positing an NP Hard problem. However, if one drilled down into first principles and axioms, the veneer of the denouncement peels off. I concluded that it was theoretically possible, but only at the maturity level of a holistic scientific model we don't-yet possess, when we already should have. There appear to be novel mathematical components still missing, but the state of science, of the apocalypse of this present time, seem ready to reveal this knowledge to this civilization. Might be, we're going to need it? If I may leave a tantalizing thought: Kolmogorov compression doesn't have to be an exclusive brute-force gateway for all probable cases of informational optimization. For instance, if energy could be mathematically harvested, it would significantly alter the AGI hardware requirements for a sense-making topology. It may announce a foundational alignment-shift in physics. Much of this "imaginary" computational topology of applied intelligence is nearing readiness, but the research, political, adninistrative, energy and hardware constraints keep on getting bigger and bigger, and more vulnerable, by the day. Just saying: what if there were a multi-hybrid scaling down solution? The current LLMs would appear like steam locomotives relative to autonomous drone taxis. If I said more I'd have to hypothesize conspiracy, but there's zero gain in that. The simplest of choices then. Do we press on towards naturalized harmonic elegance in computational systems, or admit dependence-enabling, socio-technical defeat? I'm certain the Amish would be able to benefit from such 3rd-party technology. I mean, when the masses get hungry, what would governments not conscript, or nationalize (take by force)? On Sat, 26 Sept 2026, 19:19 John Rose via AGI, <[email protected]> wrote: > On Thursday, September 24, 2026, at 4:14 PM, Matt Mahoney wrote: > > The Hutter prize has demonstrated training a 6M transformer with 4 bit > weights on one human level training set (1 GB) using 10^18 operations > including offline GPUs. I don't think this does human level inference yet > because information theory suggests needing 50M to 100M parameters to > represent the training data, which I estimate would need 10^19 operations > consuming 10^10 J or a few thousand kilowatt hours of electricity. A human > brain running at 20 watts for 75 years (2.5 x 10^9 seconds) uses 5 times > more energy. So I think the power problem can be solved with good > engineering and a few years of research. > > > There may be something occurring in the brain where encoding, storing, > querying and transmitting information is different from the same old same > old and LLM’s with their engineers haven’t figured out that efficiency yet. > How do rare individuals who have had a brain injury wake up being able to > speak different languages, play a musical instrument, or all of a sudden > have hypercalculia and can recite Pi or compute primes? > > Prime fields are interesting: > > https://drive.google.com/file/d/1Jj1p9HmbasiC4747WxYsSu9a5ZWI-iZq/view > *Artificial General Intelligence List <https://agi.topicbox.com/latest>* > / AGI / see discussions <https://agi.topicbox.com/groups/agi> + > participants <https://agi.topicbox.com/groups/agi/members> + > delivery options <https://agi.topicbox.com/groups/agi/subscription> > Permalink > <https://agi.topicbox.com/groups/agi/T25fb5a030fcfa28e-Md2414b3245437377542890c2> > ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T25fb5a030fcfa28e-M2ff61e3c8837e53711d9194b Delivery options: https://agi.topicbox.com/groups/agi/subscription
