Cyborg AI Minds are a true concept-based artificial intelligence with natural language understanding, simple at first and lacking robot embodiment, and expandable all the way to human-level intelligence and beyond. Privacy policy: Third parties advertising here may place and read cookies on your browser; and may use web beacons to collect information as a result of ads displayed here.

Friday, July 07, 2017


Cross-fertilizing ideas with another major AI project.

Yesterday we were able to do some cross-fertilization of ideas in the HTM Forum of Numenta, about which in 2005 we wrote our only Slashdot story. Numenta is where serious AI enthusiasts are taking the laborious approach of reverse-engineering the neocortex of the human brain. Then Mentifex here swoops in and claims to have solved AI with a totally top-down approach to how the mind works. The Mentifex AI Minds are based on theoretical ideas of the macro properties of neurons, such as extending spatially and temporally over a putative MindGrid and having as many as ten thousand synapses with other neurons. The Mentifex Minds use neural inhibition to dislodge briefly topmost ideas in favor of ascendant other ideas. Since Mentifex AI is concerned mainly with neuron-based concepts playing a role in thinking, we reverse-engineer neurons only enough to create AI software that can demonstrably think and reason in English, German and Russian. We hope to poach some great minds who think alike from the Numenta project. It could take a thousand years to reverse-engineer the neocortex, and Netizens who get tired of waiting for such a bottom-up approach are welcome to try out the top-down AI that runs in Strawberry Perl 5. Our goal is to release basic AI software with sufficient intellectual functionality that individuals and teams, even if working in secret, will latch on to our existing codebase, reverse-engineer it, and create from it even better AI Minds than we tenues grandia are capable of.