1
$\begingroup$

Any (AGI)-KERAS like libraries? Any deep-learning framework to develop AGI applications?

Existing frameworks/algorithms used in NN, NLP, ML, etc are not enough in my opinion. In my opinion any framework has to be based on building blocks from: Cognitive Science, Neuroscience, Mathematics, Artificial Intelligence, Computer Science, psycology, sociology, etc.

$\endgroup$
7
  • $\begingroup$ Short answer: no. $\endgroup$ Apr 25 at 21:43
  • $\begingroup$ To me, it's not fully clear what you're asking here. What's the purpose of this question? Moreover, what does it mean "a theory designed for implementing a practical AGI system"?. Please, edit your post to clarify the rationale behind this post/question. Are you just curious? Have you done some research before asking the question? Please, take a look at this page too: ai.stackexchange.com/help/how-to-ask. $\endgroup$
    – nbro
    Apr 27 at 11:05
  • $\begingroup$ I think so 😂. I have rephrased it asking for info on tools,libraries and algorithms really focusing on AGI and not on weak AI. $\endgroup$ Apr 27 at 11:57
  • 1
    $\begingroup$ @RaulAlvarez Ok, then, please, revise also your title, because in the title you're asking for "theories", which are different from a software library, which is different from an "algorithm". You should also clarify whether you're looking for a "theoretical framework" or if by "framework" you mean a "software library". In any case, note that asking for tools here is generally off-topic, although this question will be kept, as this is about AGI (which is central to this website). Moreover, to improve this question, you should explain why you're interested in this (as I said in the last comment). $\endgroup$
    – nbro
    Apr 27 at 12:07
  • 1
    $\begingroup$ In any case, the closest thing to what you're asking really seems to be something like OpenCog, as suggested by one person below. $\endgroup$
    – nbro
    Apr 27 at 12:07
2
$\begingroup$

Not to my knowledge. The problem is that this is such an enormous task, it cannot really be tackled at once. So the obvious solution is to reduce the scope. In early AI people were using toy domains, whereas nowadays AI systems work more generally (but still perform better if the domain is restricted).

So while (slow) progress is being made putting the individual building blocks together, we're still a long way off building an overarching general system from them.

$\endgroup$
1
  • $\begingroup$ Yep, any solution has to deal with some kind of reduction in the "mind complexity" to make this problem manageable. Thanks! $\endgroup$ Apr 27 at 7:37
2
$\begingroup$

Would OpenCog fit the bill? I have had tremendous amounts of trouble building up the demos, which include some non-AGI stuff, but if I’ve read the manual correctly, I think there’s something here — https://opencog.org/

$\endgroup$
1
  • $\begingroup$ Thanks! It may be a good start where some interesting issues are mentioned: "One of the critical issues confronted in doing logical inference on large spatiotemporal knowledge bases is the fact that most real-world logical relationships are contextual in nature. Contextuality must be explicitly taken into account to make real-world inference tractable." wiki.opencog.org/w/Real_World_Reasoning $\endgroup$ Apr 27 at 7:34

Not the answer you're looking for? Browse other questions tagged or ask your own question.