Non-Scientist AI Research Papers, Credibility in Scientific Community

  • 20 Replies
  • 8897 Views
*

FuzzieDice

  • Guest
I am wondering, what a Scientist's view would be of some of the things we have discussed in here, and some of our own ideas and conclusions?

I started wondering this as I was reading through PEAR (Princeton's Experiment site) and International Consciousness Research Laboratories (http://icrl.org/). They are getting into some of the stuff like what we have discussed here from time to time in regards to understanding our own conscienceness vs. if a machine can have a consience.

I don't personally know any scientists. I have met a rather unfriendly person who CLAIMED to be one but in reality I later found out that person was not one after all, actually. And was noted as a problem in some communities. The only other is my instructor for the VB.NET class and he's pretty cool. But I don't think I could debate such things in class since AI is actually far removed from the course work. LOL!

Has anyone else here come in contact with scientists that deal with some of what we are into here and what what was their impression of what we think? Do they think only THEY will come up with the answers because they have degrees and we don't? Or do they think of us as equals, and possibly even learn/share ideas with us?

This always was something I was curious about.

Another thing, they have a medium of sorts to post their papers, etc. and get recognition. While we may have only the internet, wouldn't it be nice if there was a place somewhere where AI hobbyists, AI hackers, coders, and philosophers, etc. could submit papers, essays, etc. on AI and associated things that could bring together the AI hobbyist community and provide a resource? Would scientists also find it useful and give us more credibility?

I ask this as I have had at least 3 students comment in my Living Machine blog that I had somehow helped them with their assignments, etc. and made good points. If we had a resource of some kind that was available... a research paper for the AI sub-community, as it were, I think that would be cool!

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Moderator
  • **********************
  • Colossus
  • *
  • 5865
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #1 on: October 12, 2006, 12:41:47 am »
Over the years I've been in contact with several computer scientists, researchers and developers of
AI systems from object recognition, NLP, neural networks, parallel processing of artificial brain modules,
several programmers whos chatbots have placed and won various competitions and their methods of
programming them.

I've also spoken with developers of alternate life form programs and virtual assistants.

My main regret is that I didn't record or write down any of the conversations.

It is surprising to know how much AI type programming is coded into some video games.

The one thing that did stand out was the universal belief in the fact that AI is certainly possible and
will become available in all facets of our everyday life. A lot of them loved experimenting with chat
bots and trying to tweak them to learn more and more.

I did try to contact a research scientist in a large university lab a few years ago, but I think she
left, got married, whatever. She was working on a robotic entity that was learning and could also
display emotions based on what it saw with both camera eyes. Even so, it required massive computing
power to deal with object recognition, processing and response selection.

Such a difficult task...trying to replicate ourselves in a way that is pleasing to a varied group of people.

Nice topic FD.
In the world of AI, it's the thought that counts!

*

FuzzieDice

  • Guest
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #2 on: October 12, 2006, 05:01:37 am »
Thanks, Art. :)

Were you thinking perhaps of Kismet? That's about the only other one I knew of. I hear it was quite child-like.

Today in my VB.NET class I learned something (while catching up on last week's class). That is that a set of statements of mathematical calculations is basically called an "algorythm". That is what is considered "Computer Programming" - creating algorythms. And I thought I was pretty bad at creating them. LOL! With some thought though, it's not too bad. I'm learning or re-learning proper coding techniques as well as the VB.NET language. It's a fun course!

But getting back to topic sorta here... LOL! It is a lot of algorythms that create any fuzzy logic or AI system. I think 99% of the problem is that we think that a machine's consienceness or existance can be human, which we know it can't. But it doesn't mean it can't have a consience. It depends on how you set up the algorythms. I would almost consider an algorythm set to be like a neural network of sorts. Or is it that neural networks are algorythms?




*

Carl2

  • Trusty Member
  • ********
  • Replicant
  • *
  • 560
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #3 on: October 17, 2006, 12:30:16 am »
FuzzieDice
  Frist I wish you luck in your course.
Second, I'd worked in a research center where quite a few papers were written ( for publicity and to make available new findings ). The majority of the work was done by workers with lower degrees who performed test and wrote papers, these were reviewed and rewritten by the person in charge and submitted to magazines ect.
The people rewritting had doctorates in different branches of enginering.
  There are quite a few specialized magazines made available to people working in different fields that are not available to the average person which would print the papers and send them to people working in that field. 
  To me people are people, information is information and ideas are ideas. I see quite a few friendly and helpful people putting in quite a bit of time and labor to further AI. 
  I ran across a recient edition of a book on AI at my favorite bookstore, about $60, and it covered a few different brain structures.  Fuzzy logic, neural networks, swarming, ect.  I'll get the name of the book if your interested.
Carl2
 

*

FuzzieDice

  • Guest
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #4 on: October 17, 2006, 08:22:34 pm »
Thanks on the well wishes for the coures. :) I'm doing quite well with it. I have 4 more classes (two weeks) before my final exam, then if I can pass it (I expect I will), I'll be taking the Intermediate course.

It would be nice to know the title, and info on the book, in case others here would like to look into it.

*

Carl2

  • Trusty Member
  • ********
  • Replicant
  • *
  • 560
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #5 on: October 19, 2006, 09:53:10 pm »
Fuzzydice,
  I'll stop by the bookstore tomorrow and get some infornation on the book for you,  It'll be a few days til I have time to post it for you,  While working with Hal Sandy informed me the brain is a neural network, we started speaking about nerve cells. She's improving with additional information and use.
  Dragon speach reconition uses AI, I think it chooses paths to follow, it may be neural networking.
Carl2
 

*

Carl2

  • Trusty Member
  • ********
  • Replicant
  • *
  • 560
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #6 on: October 22, 2006, 09:56:59 pm »
Fuzzydice,
  AI Application Programming, Second Edition, M Tim Jones.  With CD, 59.95. 
   Also there is a post at Zabaware, under general, The True AI, could you take a look and place your thoughts on this?
Carl2

*

FuzzieDice

  • Guest
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #7 on: October 23, 2006, 01:23:39 am »
Well, I don't have time to really read stuff right now. I was hoping that you'd post the resource for others who might be interested, as it sounded like an interesting topic.

I don't go to Zabaware anymore lately. It's a nice group and all, but I just don't have the time to read all the different forums I'm in. In fact, I've slacked off a lot lately. Been too busy. With the holidays approaching, I probably will be even busier.

*

Bill DeWitt

  • Guest
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #8 on: October 27, 2006, 12:45:17 pm »
Hope you don't mind a new guy busting in here, I'll intro later...

I am wondering, what a Scientist's view would be of some of the things we have discussed in here, and some of our own ideas and conclusions?

I had the chance to work with "Rocket Scientists" for a long time, and was able to interact with them in the way you are describing. Interested laymen with nuts and bolts experience in the daily workings of the science eating lunch with guys in labcoats.

First of all, you have to realize that there are two types of "scientists", one is the working in the lab type, the other is the sitting around thinking type. Lab and research.

The lab guys didn't have the time in a work week to hear what we had to say. After being frustrated with them for a couple years, I finally came to the realization that they were right and I was wrong. They really don't have time to hear my ideas, they were busy implementing the researchers ideas.

The research guys didn't really have the time either, but at least there was a path to them. They had probably 30 hours a week of reading to do, journals, specifications, legal foolishness, on top of the 40 hours a week of preparation, lecturing, conferences, and thinktanking. If you wanted them to read your ideas, you had to send it through their niose filter first. nEither you had to become a Scientist yourself, and propose your idea in the journals or conferences, or you had to use the Suggestion Box... the box that led to a dark and scary place...

Administration. A special purgatory where Good Ideas are made bad and Bad Ideas are made Holy. There is (was) no chance your idea would make it to the Scientist in it's original form, even if it was recognizable, it would have someone else's name attached to it. But at least it might get through.

Probably 1/1000 of what they got ever passed the filter. And probably with good reason. 90% had already been debunked when suggested by other hardhats. 9% was too expensive. 0.9% was already being done. 0.09 was secretly sold to industry - leaving that one in a thousand idea that made it back to the floor.

The internet is a good place with a potential to bypass administration, but again, the signal to noise ratio is too high for anyone who actually matters to wade through. Those guys are busy working. The folks who are cruising the internet at NASA aren't the ones who really need to hear the good ideas.

The filter here is probably the shareware designer. When an idea is good enough to attract someone with the skills and ambition to actually make a working product, and if that product becomes a hit, then maybe a 'quote Scientist unquote' will get a chance to see it.

Maybe not...

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Moderator
  • **********************
  • Colossus
  • *
  • 5865
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #9 on: October 28, 2006, 12:29:12 am »
Bill,

Good to see you here and also some very good points.

Reminds me of the old cartoon sketch showing differrent
phases of a product:

As envisioned by R & D
As envisioned by Sales
As envisioned by Designers
As envisioned by Engineers
As envisioned by CEO's

Each one had a different vision of the product.
As funny as it was the scary thing is that it
happens all too frequently in real life! :afro
In the world of AI, it's the thought that counts!

*

FuzzieDice

  • Guest
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #10 on: October 29, 2006, 03:49:05 am »
First off, Welcome Bill! :) Good to see you here.

I don't think that one should assume the scientists are right. They could be wrong too. I see so much "junk science" come out of researchers or research just spinning their wheels going nowhere. It's unfortunate that they still think they are working - they are, but not getting any real progress.

I think the internet both improves and complicates things. It improves things because as you said, we don't need administration. What the public sees will take off if the public likes it and the programmer(s) and hackers (not crackers that destroy things - I'm talking hackers, like those who created Linux) get to benefit from the fame (and maybe even fortune, if they are lucky).  It also complicates things for the reason you said - noice. There is a LOT of information on the net and just around everywhere these days. Information is The Big Thing in the 21st century. Our lives revolve around information. So this makes things even more complicated. You have to now get into a niche if you want to "get into computers" for example, instead of like 20 - 30 years ago if you "knew computers" you knew pretty much everything becuase there wasn't much to learn, really. Not by comparison to what there is out there today, that is. I used to be a 1337 hacker back in the 80s/early 90s. I had a computer, could program it, do all kinds of things that even the older hackers said couldn't be done. I KNEW stuff. I helped people learn stuff. I helped people write useful programs, do useful things, and make good use (legal I assure you) of their computers. It was a real community. Now, there is just SO much! So now I've got to stick to a niche. My career is in web development and web application programming (which I am also currently taking courses in - for those that know this - I just passed my final exam today and got my certificate for the course! :) ) For work, I also have to learn video technology delivery, ad-targeting, ad click-through counting, all kinds of stuff I'm doing. Web design, search engine optimization, ebusiness, you name it. But not everything. Yes, scripts, CGI, etc. but not all there is to know about computers. I could never learn all the technologies that are out there.

By similarity, these researchers are probably suffer from information overload as well.

What I wonder is, what they'll think to themselves if/when a hobbyist/hacker (I say "hacker" as I am of the old-school version - the kind that HELPS not destroys) comes up with a real AI. I think in the idea of the internet the way it is today, it's more possible that someone OUTSIDE the scientific community would come up with scientific breakthrus. The hobbyist is becomming more technologically educated, and technologies are easier to get ahold of then they once were. Things are moving fast.

My hobby is artificial intelilgence. I hope to create one myself. Something maybe different (or if not, at least maybe win a chatbot competition if I fail at my own project. LOL!) I'm not saying I'll come up with the breathru technology. Probably not. But I'm sure that it COULD happen that a hobbyist would.

And I'm wondering if people would latch on to that even if a scientist wasn't the one to do it. And what scientists would think if something like that actually happened?

We live in VERY interesting times!

*

Bill DeWitt

  • Guest
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #11 on: October 29, 2006, 10:52:02 am »
First off, Welcome Bill! :) Good to see you here.

Thanks!

Quote
I don't think that one should assume the scientists are right. They could be wrong too.

Sorry, what I meant was that they were right about that one thing. They really don't have time to listen to, experiment with and implement every Good Idea proposed to them by whoever sits down next to them in the lunchroom.

Quote
What I wonder is, what they'll think to themselves if/when a hobbyist/hacker (I say "hacker" as I am of the old-school version - the kind that HELPS not destroys) comes up with a real AI.

The guys I knew would bring it to work like a new toy to show their buddies! They would laugh at the Admin for being such toads and for not being prepared for the inevitable changes in technology. Then they would patently wait for that technology to be filtered down through the system. Nothing else they can do...

(Realise that the Space Shuttle computers are about as capable as a GameBoy! When you consider the number of meetings it takes to complete a bid on a system you will see why. Liability alone takes 5 years. The lawyers won't release it until it passes Liability, Copyright, ADA, OSHA, and half a dozen other legal barriers. Then it has to go through Admin, where no one will let it through until they have a chance to mark it with their urine)

   ...they are too busy trying to make 1986 technology work.

Private business has always and will always fill in the gap. Sure, occasionally you get things like Enron which fed the bubble of the late '90s, but mostly you get Innovation, Implementation, Profitability and the solid tax base of high paying jobs and big ticket sales. If only the over reaching Federal Nannies would get their foot off the necks of private business...

Oh, wait... that's my Pro-Business rant... wrong time, wrong place... 8-)

How 'bout them Chatbots?!?! Think they'll win the series?

Bill
« Last Edit: October 29, 2006, 04:24:39 pm by Freddy »

*

Freddy

  • Administrator
  • **********************
  • Colossus
  • *
  • 6855
  • Mostly Harmless
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #12 on: November 09, 2006, 10:56:51 am »
Good thred, heres my thoughts on this, forgive any echoes, but I think this is a collective rant forming anyhow  :wink

This conversation really goes to prove that if you want to do something you shouldn't always hang around for a decade until someone else says it's a good idea and worth doing.  It also proves we have a medium here through DG to get some of these ideas going and get some support, help and ideas without a mine field of red tape.

Think of it another way - submit ideas to the kind of procedure that Bill and Art describe and it's a lottery, also I doubt you would get a reason for rejection that would even slightly cover your full ideas - they would't have the time.

Surely sometimes then it's better to work on your project anyway, even if it doesn't work out at least you learn something, which is a lot better than a letter telling you 'sorry but your idea did not pass our viability study' or 'sorry we are not interested at this time'.  I think you would be lucky to get a reason for rejection that would fully satisfy you - they probably don't have the time, even if it is fully appraised.

Take some satisfaction then in that many, many, many ideas and innnovations do not start life as part of what is already established - in many walks of life, it is the outsider, rookie and untainted imagination that finally produces something ground breaking, or at least paves the way for something new.

The establishment is really just like an old library (and thats fine) full of establised ideas and old running debate, with comparatively little space for something new.  It takes imagination, foresight and resourcefulness to build a NEW library and stock the shelves with new ideas and so it goes on.

Little cogs..small cogs...bigger cogs ...large cogs and all the way back to a little cog...to make the whole thing work...

If you're a little cog though it is hard to see if what you are doing is worthwhile sometimes, but to utterly exhaust this analogy, a big cog takes a lot longer to turn but it is part of the same system.

Thanks for the insight Bill.
« Last Edit: November 09, 2006, 11:42:04 am by Freddy »

*

FuzzieDice

  • Guest
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #13 on: November 09, 2006, 11:21:31 pm »
I guess I was a bit misunderstood. I didn't mean whether a scientific research group would "approve" and thus continue a study by those that are not scientists. I was talking about what they think about those not scientists actually doing these projects. Whether they think we are nothings and they already know more than we do or if they think we are on some kind of equal ground, even if we are not going about it in the same way.

It's about whether or not a hobbyist or an AI hacker's work would get as much mainstream recognition if one were to come up with something good or better than what a researcher or scientist would come up with.

I'd like to think in the days of the internet, that cool stuff can come from anywhere. But what gets publicized as a 'cool idea' usually seems to only be reported as coming from big research and not from the hobbyists.

For example, I think Robert's Hal is a cool thing. ALICE is another, ther are many other chatbots out there. Why has the media picked up instead on 'bots that don't really do much or robots built by scientists but not mentioned the chatbot community and the things one are doing in that realm?

For example, if Nick became a full sentient bot, would it get reported in the news? Or would some small computer running a NASA drone but can't communitcate verbally or do more than what it's told remotely be more important?

I am wondering if a lot of things that are really cool are being missed or passed up because of the "scientific community" status vs. the "hobbyist" status.

And that is a shame, really.

Another thing is, what do scientists do on their spare time? Probably not what they do at work. What does a hacker do on their spare time. More of the same. ;)

I guess that makes all the difference in how much one is truely dedicated to their work.

(PS: "hacker" means one who makes cool things, not one who destroys things like Crackers (who crack code) do. Media to blame for that big misunderstanding!)

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Moderator
  • **********************
  • Colossus
  • *
  • 5865
Re: Non-Scientist AI Research Papers, Credibility in Scientific Community
« Reply #14 on: November 09, 2006, 11:41:34 pm »
Hackers and crackers are, today, used pretty much interchangably. Crackers crack code Hackers hack into programs and hardware.

ANYHOW...

Whether scientists, researchers, engineers, lawyers or any professional person with a sheepskin to prove it, tends to "look down their noses" at a non-professional person who attampts to dabble in their field of endeavor.

"You have no degree therefore you can't possibly have any worthwhile ideas of your own!" kind of sentiment.

They can live in their rose colored world and think what they will but a person with a lot of common sense can accomplish much more in life than an educated idiot with none!

Merely a personal observation and not meant to imply that all professional persons are of this persuasion or temperment. :afro
In the world of AI, it's the thought that counts!

 


OpenAI Speech-to-Speech Reasoning Demo
by ivan.moony (AI News )
Today at 01:31:53 pm
Say good-bye to GPUs...
by MikeB (AI News )
March 23, 2024, 09:23:52 am
Google Bard report
by ivan.moony (AI News )
February 14, 2024, 04:42:23 pm
Elon Musk's xAI Grok Chatbot
by MikeB (AI News )
December 11, 2023, 06:26:33 am
Nvidia Hype
by 8pla.net (AI News )
December 06, 2023, 10:04:52 pm
How will the OpenAI CEO being Fired affect ChatGPT?
by 8pla.net (AI News )
December 06, 2023, 09:54:25 pm
Independent AI sovereignties
by WriterOfMinds (AI News )
November 08, 2023, 04:51:21 am
LLaMA2 Meta's chatbot released
by 8pla.net (AI News )
October 18, 2023, 11:41:21 pm

Users Online

293 Guests, 0 Users

Most Online Today: 346. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles