• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed

  • Powered by FeedBlitz

« Fundamentally Different | Main | Nanotechnology and Wealth »

August 22, 2005


Feed You can follow this conversation by subscribing to the comment feed for this post.

Tom Craver

Maybe a tool that continuously monitors everything we're reading and writing, and continuously updates a listing of stuff it has searched out and maybe even pre-downloaded for us - just click a link. And of course, following a link will cause it to search out and present a different set of links. It'd feel sort of like rummaging through your own memories, by a process of association.

It could learn to recognize terms you're likely not familiear with in a document you're reading, and look those up for you. So you're reading about the space program and you see terms like CEV and LSAM - and by the time you read those terms, the computer has realized you probably don't know them, and has highlighted them with a link to a useful reference, or perhaps even a definition it can pop up for you.

Chris Phoenix, CRN

Tom, that's a great tool, but you've designed a 90's interface onto it. We've got processing power to burn nowadays; why not have it look up every word in the document, using the context of the document to present the most useful definitions? Just mouse-over a word, and a mini-window pops up with information about it. Roll the mouse wheel for more or less information.

OK, that's still a 2000's interface. For a 2010 interface, don't use a mouse; use eyephones and gesture recognition for richer, more intuitive navigation. But we don't have the hardware and UI developed enough yet.


Mike Treder, CRN

You guys are both being way too conservative. The next paradigm will increase our ability for information storage and retrieval at least 100 times, and more likely 1000 times. We've been on an exponential curve for a hundred thousand years, and I doubt it will stop now.


My guess is we will have a close interface with a networked supercomputer-powered AI.

Tom Craver

Each wave builds on the last, but using a radically new idea to take over a formerly human task:

- "Sounds can have meaning" (but humans had to remember information)
- "Symbols can store information" (but humans had to copy it)
- "Machines can duplicate information" (but humans had to manipulate it)
- "Machines can independently process information" (but humans had to provide access to the information)
- "Networks give easy access to information" (but humans still have to...????)

Fill in the ???? and you've probably got your next paradigm shift.

Actually, within "network paradigm" there's been several sub-paradigm shifts. Clearly the linked Web was a major enabler - but humans still had to browse for what they wanted to know. Search engines took the Web from a toy to a resource - but humans still had to figure out how to search effectively. My own proposal above was basically to eliminate the human effort of figuring out how to search.

It's worth asking whether these sub-paradigms might not themselves be full-fledged paradigm shifts - sure they're coming fast and furious, but that's exactly what that table projects!

If we take that assumption, we might replace the last, "network" paradigm with the following:

- "networked computers bring information to any computer" (but humans still need to know where the information is)
- "Web protocol eliminates the need to know where information is" (but humans still need to search the Web for what they want)
- "Search engines automate web searching" (but humans still need to specify the searchs they want)

Which would bring us up to what I was getting at in my first post:

"Auto-search figures out what what will be of interest, searches for it, and presents it" (but humans still have to ????)

Possible next steps would be "abstract useful information", "structure collected information toward a specific task", or something like that.

Dan S

Direct computer-brain interface will be the next big step. Before it, several sub-steps will be taken:
- 3D search interface (perhaps concepts as nodes in 3D space and links as search vectors)
- Special language for manipulating concepts and visually construct search terms
-Virtual reality based search engine with VR-helmet and gloves

Mike Deering

Vista doesn't constitute a paradigm shift but an incremental change. The WWW goes 3D. Every website is a 3D space instead of a 2D space. Every website has a 3D relationship with every other website. But every IP address can have as much space as it has memory to display, so it's more like 4D.

michael vassar

Integrate and judge the information. Synthesize model worlds from it. Sounds like Cyc or other powerful AI tools could fit the bill. In a less high tech world, so could very effective and well designed market feedback systems.
Neurotech enabling people to evaluate their own evaluations might be even more important.

Mike Treder, CRN

I envision distributed evaluation systems: my avatars (AI's) go at an issue or problem from different points of view. They prepare a matrix of outcomes based on thoroughly run scenarios and show me a smart decision tree -- all within seconds, of course.

Dan S

Actually, everything you proposed is incremental change – just combination of existing capabilities. A new paradigm is something that you could not construct that way. What I am speaking about is completely different thing.
Considering logical structure no web site now is 3D. Most of it is actually 1D. Text is a linear combination of symbols. Necessary so, because upon invention of writing there were no other options. Web links add second dimension trying to utilize capabilities of a new media. But this is a crude way to it.
Linear writing systems will become history. But not just that. We all are capable of reading new media content – web pages, interactive computer programs, 3D simulations… But how many of us are capable of creating that content (‘writing’), utilize full capabilities of new computer media? Only a few. Computer revolution is a revolution in a way we think and it is not happened yet. Not until every computer user will feel absolute freedom in creating new computer content. And that *is* the new paradigm.

Tom Craver

Paradigm shifts occur when something important that previously depended on human mental effort, can now be done independent of humans.

And it's RESULTS that count - if something creates an actual fundamental change in how people do things, and in their effectiveness, it should count as a paradigm shift - even if it mainly appears to extend elements of previous paradigm shifts.

Writing "merely" extended language.

The printing press "merely" extended writing.

Search engines "merely" extend the web and computing paradigms - but using a search engine amplifies human mental capabilities in a way that web browsing and computer applications alone did not.

Nick Rennick

I think that the next paradigm shift will be far more revolutionary than those mentioned above. Mike Treder's comments show how each new paradigm enables humans to do less and less, as the new paradigm replaces a previous function of the human brain. He also asked what it is that computers cannot do currently that we need to, and explained that this will be the next paradigm shift. Computers with the internet still require us to THINK, to tell them what to do and direct their actions to useful means.

The paradigm I speak of is AI, where computers no longer need us to direct their actions. Our current information technology is equivalent to a robot, with incredible powers that are limited to what its human user can tell it to do. Our brains can only move so fast and understand so much, so it is us not the computers that are limiting our understanding. When computers are able to think for themselves, they will become far more intelligent than human beings, and probably eventually replace us as the dominant species.

Although this does raise significant moral and ethical issues, it is inevitable. If it is theoritically possible then it will become a reality at some point in the future. However, for computers to match the human brain's capabilities it would require much more advanced hardware and software, so we probably wont be seeing "Strong AI" for another 50 years or so.

Mike Treder, CRN

Nick, I agree that AI will represent Paradigm VI.

As computer storage capacity continues to increase explosively, it will soon outstrip our human ability for effective retrieval of information (if this has not happened already). Fortunately, before long we will enjoy the assistance of artificial intelligence that will not only provide access to needed answers, but also will be smart enough to ask questions we could not even formulate. I don't think this will take 50 years, though; I expect it within the next 15 to 20 years, if not earlier.

Of course, reliable storage and easy access to information -- no matter how much or how fast -- by itself is not enough to power the rapidly approaching technological singularity. Information quality (i.e., accuracy and relevance) and the ability to make creative, innovative, productive use of information are equally significant.

But if we assume that tomorrow's artificial intelligence will be smart enough to quickly and easily sift out quality data, and that the AI will also possess creativity at least the equal of our own, then it seems certain that technological and social change will indeed occur at a pace faster than we can presently imagine.

Tom Craver

It's important to distinguish between the technology development that enables the paradigm shift, and the shift itself. Recording human thoughts was the paradigm shift - written language was the technology. And the paradigm shifts improve human mental power or leverage, allowing us to accomplish more.

Major shifts require major new technologies, but also build on the previous shifts. Look at the problems remaining in the current paradigm - the next will solve some of them.

So - what still frustrates or limits you about the Web? For me:

- Getting any group to settle any controversial issue seems impossible - egos and vastly different experiences/backgrounds get in the way.

- There's so little practical benefit - the web isn't much help for thinking through and solving complex problems as a "group brain". People don't seem to brainstorm and plan new businesses online, or design new logic circuits, or do anything else requiring detailed planning.

Tom Craver

Wikipedia may be a partial exception to "nothing complex gets accomplished by groups online".

Chris Phoenix, CRN

Catching up on this discussion... Tom Craver wrote:
"Networks give easy access to information" (but humans still have to...????)

My immediate answer was: ...still have to know what they're looking for.

The subsequent posts said much the same thing. The next paradigm may be computers that feed us information that we know to look for. This would go far beyond extrapolating our searches and clicks in order to push content. This would be more like: your computer shows you something interesting that you'd never have thought of looking for and never told it you were interested in.

How could this work without telepathy and precognition? By data-mining for concepts and making connections between different domains. We've been bemoaning overspecialization for decades, while recognizing that there's no way a human could learn all of science anymore. Well, no human could... but a computer might get enough of a grip on it to notice worthwhile interdisciplinary collaborations.

In fact, I read several years ago that this was already being worked on, looking for connections between the keywords of papers in seemingly unrelated fields.

Once this kind of capability exists in even embryonic form, once people realize that it's possible and oh-so-valuable, they will try to extend it. There will be a ferment of efforts from which will emerge a common conceptual language. And within a few years, it will be de rigeur to put useful information into that system somehow, just as today anything that's not on the Web doesn't really exist.

I think I'm describing what the "semantic web" wishes it could grow up to be. But the semantic web is being pushed, not pulled emergently. We'll have to wait a while for the useful conceptual interchange format to emerge from data-mining projects that don't even exist yet.

Once we have this language and these projects, we'll be well on our way to a computer system-of-systems that in some sense "knows" a lot about the real world. It won't be "real" AI, of course, any more than expert systems are "real" AI. (That was sarcasm.) But it'll sure be useful.


Tom Craver

I'd like to see "WikiCalc" : Web software that would allow people to set up (and review and modify) little calculator programs, that others can link together to create more advanced calculators.

Eventually all data would be web published in a wikiCalc friendly format - tables that can be pasted into a WikiCalculator, or active wikiCalc data generators that can simply be linked to others.

Yeah, a lot of bad calculators would be made by amateurs, but each such instance would motivate experts to provide a more correct version, or at least annotate the calculator with a description of its limitations.

Finally, you'd want to be able to link to a specific calculation, or even better pop up a window to a calculator in something you post on a discussion board or on your own web site, so anyone can see your calculations as well as your results.

Chris Phoenix, CRN

Tom, that's a great idea! It should happen.

I was just reading an energy report that might as well have been presented in furlongs per fortnight. No two tables used the same units.

I'm also thinking that WikiCalc could extend into simulations and visualization.


The comments to this entry are closed.