new year resolution 2012

2 views
Skip to first unread message

YKY (Yan King Yin, 甄景贤)

unread,
Jan 2, 2012, 1:09:45 AM1/2/12
to general-in...@googlegroups.com
Forecast for ~ next month or so...

1.  The online book is badly in need of revision, but I didn't have the time.  I will add a to-do section to the online book, summarizing what is to be revised, as a quick remedy to this problem.

2.  I've just worked out more details of the inference speed-up algorithm and distributive inference / message routing problem.  Will make a YouTube video of it.

3.  Will make another YouTube video about the politics of AGI.

4.  I'm considering to use a graph-based logic (similar to Opencog's Atomspace, and this guy Mark Rodriguez is also working on a graph database with similar aspirations).  Hopefully the transition will only involve minimal changes, and even simplify things further.

That's the short-term to-do.

For the coming year:

1.  Find an application area for Genifer that genuinely help solve people's problems.  Opencog is working on a generic game engine.  I'm inclined to avoid that area, and find something more original =)

2.  Find some partners to get involved in development again.  Our last attempt has stalled and currently our project is back to single-person (me).

3.  Find some beneficial ways to work with Opencog.

PS:  Last month, on IRC, I kicked Seh out of the chat room because I thought he made some negative comments.  I looked at the log again today and found that I was completely mistaken.  I apologize for that -- was paranoid again =[

KY
-- 
"The ultimate goal of mathematics is to eliminate any need for intelligent thought" -- Alfred North Whitehead

Matt Mahoney

unread,
Jan 2, 2012, 10:40:53 AM1/2/12
to general-in...@googlegroups.com
2012/1/2 YKY (Yan King Yin, 甄景贤) <generic.in...@gmail.com> wrote:
> For the coming year:>> 1.  Find an application area for Genifer that genuinely help solve people's> problems.  Opencog is working on a generic game engine.  I'm inclined to> avoid that area, and find something more original =)

Yeah, I would do that first. It will help clarify what else needs to be done.

I posted on OpenCog a list of things that I think AI ought to be able
to do. Something like this:

Language:
- Predict text with human-level performance.
- Answer questions in natural language.
- Predict user interest in email or web pages (relevant/spam).
- Pass a Turing test.
- Solve crossword puzzles.
- Solve math problems and prove theorems expressed in natural language.
- Write, test, and debug programs.
- Classify jokes as funny or not funny.

Vision:
- Read scanned text, including captchas.
- Match pictures to captions.
- Identify when two photos are of the same person or object.
- Write a script describing a video.
- Identify human emotions from video of facial expressions.

Hearing:
- Recognize speech.
- Synthesize natural sounding speech.
- Recognize when two songs are by the same artist or of the same genre.
- Recognize common sounds (a barking dog, breaking glass, etc).

Robotics:
- If embodied with an arm, obey commands to pick up, place, throw, and
catch objects.
- If embodied with wheels or legs, obey commands to navigate over
roads or complex terrain.
-- Matt Mahoney, mattma...@gmail.com

Reply all
Reply to author
Forward
0 new messages