What's everyone up to ?

  • 909 Replies
  • 128562 Views
*

ivan.moony

  • Trusty Member
  • ********
  • Replicant
  • *
  • 684
  • look, a star is falling
Re: What's everyone up to ?
« Reply #900 on: June 29, 2017, 08:37:35 pm »
Been busy lately with improving a metalanguage I'm working on, it is still not certain when I'll publish it. It looks cool, but I wonder if it can look even cooler. I had some plans with it including smart encyclopedia, smart induction machine that uses shared processor time to conclude new knowledge in science, smart knowledge extractor, assistant that solves problems in different areas of science, and the list goes on as I approach math, physics, chemistry, biology, ... It is never ending, but I have only one life to spend and it is not enough for all the things I want to do.

Until lately, I had a very strong motivation for my work in hoping that it would be used for easier developing of artificial food, so we don't have to kill to survive, but in the last year I got raced out (should I say luckily) by two startups named Memphis Meat and Super Meat. Guys behind those names did something extraordinary. They managed to grow a whole meat stake in laboratory, using from a few living meat cells taken from chickens and cows that walked out alive. Last thing I've heard lately about these very special projects was that they managed to lower the cost for producing a pound of such a meat to only $6000. I'm looking forward for cheaper meat that could replace traditional meat production.

I should be happy, finally my dreams are coming true, but somehow... you know, I feel a kind of bizarre...for about twenty years I have been drifted by a single idea of helping in similar projects by offering an artificial intelligence, but now, suddenly I'm useless. But I shouldn't complain, things are developing better than I expected, people actually care for animals, 2/3 of Americans would taste artificial meat, while 1/3 would completely replace traditional meat by the artificial one, some sources say. I should force myself to be thankful.

Those guys are doing fine even without me, I can finally relax and point my attention in another directions. I need to lower down my expectations and to take a more realistic pathways in my work. New times, new plans, I guess...

*

korrelan

  • Trusty Member
  • ********
  • Replicant
  • *
  • 662
  • Look into my eyes! WOAH!
    • Google +
Re: What's everyone up to ?
« Reply #901 on: June 29, 2017, 08:54:25 pm »
As good as their solutions are… what if they fail… then what?

Soylent, flying cars, numerous cures for the big C, ultimate antibiotics, omg, the list is endless… start-ups need money and promise the earth, they receive funding and then… vanish.

Who picks up the pieces Ivan? Who has the next good solution? Who is still working the problem?

Until you see it being widely utilized and making an actual difference… it’s worthless.

Look at Deepmind, IBM Watson, hundreds of companies and start-ups declaring they have solved the AGI problem space and we can all relax… the world is going to be a better place… do I stop researching… nope.  It’s just spin, bo**ocks… until you see it making a difference… nah! Part of the whole process is to make claims that will psych out the competition... its like two boxers before a bout... don't fall for it dude.

 :)
It thunk... therefore it is!

*

hakobian

  • Roomba
  • *
  • 5
Re: What's everyone up to ?
« Reply #902 on: July 01, 2017, 08:41:52 am »
Hi all/ O I seen my two of my favorite websites change shape and ideals.   i wasn't happy.  But after two years, there is much to read about.

Increased mechanation will improve the world.  A move towards awareness rather then privacy will further democracy.

My ideas are bad without backing up.

Technology is availaible.  It isn't about proving it but making it happen.

The world is your bot.

It's not AI though.

How big a machine do you need for ai?
« Last Edit: July 01, 2017, 09:23:16 am by hakobian »

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Global Moderator
  • ******************
  • Hal 4000
  • *
  • 4360
Re: What's everyone up to ?
« Reply #903 on: July 01, 2017, 07:38:23 pm »
I think that would depend on one's definition of A.I.
Chatbot - relatively small
Digital assistant - larger and more complex
AGI - Extremely large amount of resources and computing power.
In the world of AI, it's the thought that counts!

*

Freddy

  • Administrator
  • **********************
  • Colossus
  • *
  • 6051
  • Mostly Harmless
Re: What's everyone up to ?
« Reply #904 on: July 04, 2017, 11:01:38 pm »
6000th Post  :jump3:

*

Data

  • Global Moderator
  • ***********
  • Eve
  • *
  • 1263
  • Overclocked // Undervolted
    • Datahopa - Share your thoughts ideas and creations
Re: What's everyone up to ?
« Reply #905 on: July 05, 2017, 11:53:15 am »
Grats on the 6000  :waveJump:

*

ivan.moony

  • Trusty Member
  • ********
  • Replicant
  • *
  • 684
  • look, a star is falling
Re: What's everyone up to ?
« Reply #906 on: July 06, 2017, 12:01:03 pm »
I just invented a new text parsing algorithm. Should be fast. And should be able to do natural language processing. I think I'll call it "Scheduling Chart Parsing Algorithm". Beside that, along with already seen structured grammar input, I invented a new way to output a very compact and memory friendly parse forest.

(You don't want to know how many times I've rewritten my javascript parser over the last five years. I think I'll finally settle down with this one)

*

infurl

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 255
  • Humans will disappoint you.
    • Home Page
Re: What's everyone up to ?
« Reply #907 on: July 07, 2017, 04:00:34 am »
Some details to support your claims would be very interesting to see.

How does your algorithm compare with the Tomita GLR parsing algorithm?

*

ivan.moony

  • Trusty Member
  • ********
  • Replicant
  • *
  • 684
  • look, a star is falling
Re: What's everyone up to ?
« Reply #908 on: July 07, 2017, 04:31:21 am »
It is top-down and It is based on scheduling ahead parsing symbols(s) in corresponding places in the chart and advancing the chart pointer by one in a loop. Also supports the left recursion in the essence. Only the topmost symbols in sequences are written to the chart, while sub-alternations are calculated and forgotten as soon as ahead symbols are written to the chart. Reminds of Earley, but it's simpler. A few dozens lines of code brutto, I think.

Will publish it as soon as I implement it, promise :)

*

infurl

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 255
  • Humans will disappoint you.
    • Home Page
Re: What's everyone up to ?
« Reply #909 on: July 07, 2017, 05:14:46 am »
Thanks, I get what you're doing now. I wrote an Earley parser using Common Lisp some years ago. It turned out to be quite complicated but it worked very well until you started encountering a lot of ambiguity which slowed it down dramatically. I discovered that merging parse trees improved the performance dramatically and thought hooray I'd invented something.

Then a bit later after that I learned about GLR parsers and the Tomita algorithm and realised that the method had already been invented in 1986. The good news was that GLR parsers are way faster than Earley parsers, and much simpler too because they use shift-reduce, so I rewrote all my libraries around GLR parsing algorithms, still in Common Lisp. Subsequently, I rewrote everything in C and now I can parse arbitrarily large documents defined by complex context-free grammars containing millions of rules very very quickly indeed. All in all, I'm very pleased with the result.

It sounds like you are doing some good work there, keep it up. :)

 


Users Online

14 Guests, 2 Users
Users active in past 15 minutes:
ivan.moony, WriterOfMinds
[Trusty Member]

Most Online Today: 47. Most Online Ever: 208 (August 27, 2008, 09:36:30 am)

Articles