GTOM: Difference between revisions
From MediaWiki
No edit summary |
No edit summary |
||
| Line 1: | Line 1: | ||
Gativus has a vision of artificial intelligence based on nature of human consciousness. Most fundamental is the fact, that human’s brain simultaneously host two quite different networks: neural network and concept network. | Gativus has a vision of artificial intelligence based on nature of human consciousness. Most fundamental is the fact, that human’s brain simultaneously host two quite different networks: neural network and concept network. | ||
[[File:MEM-116-101.gif|center|thumb|''Figure 1: Two Networks in one brain'']] | [[File:MEM-116-101.gif|center|thumb|''Figure 1: Two Networks in one brain''|617x617px]] | ||
Research in neural networks has a lot of achievements (such as GPT, LLM, etc.), whilst the concept/knowledge networks can’t boast significant results (very restricted solution, such as RDF/OWL description). | Research in neural networks has a lot of achievements (such as GPT, LLM, etc.), whilst the concept/knowledge networks can’t boast significant results (very restricted solution, such as RDF/OWL description). | ||
Gativus is making attempt to converge these two networks into unified common network and give the researches an opportunity to find association between neural and concept layers. | Gativus is making attempt to converge these two networks into unified common network and give the researches an opportunity to find association between neural and concept layers. | ||
Revision as of 02:45, 7 August 2024
Gativus has a vision of artificial intelligence based on nature of human consciousness. Most fundamental is the fact, that human’s brain simultaneously host two quite different networks: neural network and concept network.
Research in neural networks has a lot of achievements (such as GPT, LLM, etc.), whilst the concept/knowledge networks can’t boast significant results (very restricted solution, such as RDF/OWL description).
Gativus is making attempt to converge these two networks into unified common network and give the researches an opportunity to find association between neural and concept layers.
