Monday, December 11, 2006

Web 3.0 - The Bridge To The Singularity


Web 3.0 is the advent of a brave new paradigm where we will interact and solve problems together through a network of A.I. assistants. This is Part 3 of “The Web 3.0 Manifesto” where I describe how we will all become integral cognitive nodes of the Internet, cooperating in real time to provide key knowledge, logic, and pattern recognition capabilities for each other in breathtaking new ways that were not possible before Web 3.0. I will use a story to introduce some of the core technologies and protocols that will comprise Web 3.0. In later articles I will talk about some of the artificial intelligence technologies and methods that will facilitate this new generation of the Web.

Preface To The Story

In the future, user and programmer alike will be integral participants of what is commonly referred to as the Cognitive Abstraction Layer or CAL. The CAL is a mixture of machine and human intelligence, supported and woven together by new Web 3.0 protocols and technologies, and designed specifically to facilitate problem solving between people acting anonymously in a dynamic cooperative manner. This new paradigm for anonymous cooperative problem solving will set the stage for greatly advancing the current state of our combined technological prowess, raising it to the level needed to begin work on the A.I. supercomputers that will mark the beginning of the Technological Singularity.

Note: by the time Web 3.0 appears we will use many different new interfaces to communicate with computers. However to keep things simple in this story, Jonathan talks to his computer and his computer talks back. Naturally, his computer’s name is HAL.

Story: A Day In The Life Of Jonathan Byte

[Scene opens: Jonathan Byte, an elite Web 3.0 programmer, is ready to start his day and we open the scene with him in his room ready to begin work.]

HAL: Hello Jonathan.
JON: Hello Hal, work configuration please.

[The computer wipes away the 3D World Of Warcraft 23 virtual playing field that fills the room in mid-air, made using the latest 3D projection technology, and replaces it with the Dynamic Marketplace interface to the Cognitive Abstraction Layer. Jon spreads out several floating 2D screens into different areas of the room to get a better glimpse of today’s offerings. He does this by using hand gestures with the help of the Nintendo Wii Reality Bridge, an interface much like the one seen in the science fiction movie, Minority Report.]

HAL: You’re in luck today. There are a large number of diagnostic requests in the marketplace today. Would you prefer locked or open bid offers?
JON: I’m feeling adventurous today. Show me open bid offers first, but only those that have a 90% probability of attracting no more than 5 competing programmers.

[In an open bid, any programmer can compete in real time for the job in a winner-take-all fashion. The first programmer that solves the problem or finishes the design task takes the specified payment with the others getting nothing. A locked bid offer is a request for programming help where the submitter commits to pay the subscribing programmer a specified fee, as long as the work is completed in the allotted time. Locked bids are lower risk and therefore pay a lot less.

Open bid jobs are what keeps Web 3.0 commerce moving at maximum speed and makes many programmers rich in a short time. By examining the history of each programmer before allowing them to compete, the CAL will ensure that the group of programmers competing for an open bid have disparate problem solving approaches to the specified task request. This guarantees that the degree of healthy competition between ideas necessary to find an optimal solution exists, while avoiding wasteful effort that is truly redundant and varies only in a trivial manner. With a little help from HAL the jobs that Jonathan usually prefers are brought to his immediate attention.]

JON: Nice one Hal! I see you’ve found some diagnostic job requests that require a level 10 visual pattern recognition expert like myself, but don’t require extensive domain specific knowledge. Yesterday I had a task that required me to learn about intercity air traffic coordination before starting the job. I lost nearly a half hour of billable time reading about that stuff!
HAL: I’m always happy to help Jon.
JON: Pull up the job labeled “ADX-Omega 3” for me please, and quickly since it’s an open bid.
HAL: There will be a 5-minute delay before the Hypergeometric Data Visualizer (HDV) is available and prepped.
JON: Screw that. Jack up my HDV account to premium level 5 to get the time down to 30 seconds. I’ll pay the additional cost.
HAL: Based on previous experience you have a 75% chance of winning this bid, do you want to confirm the HDV premium usage fee?
JON: Confirmed, now move it, the clock is ticking!

[As part of Web 3.0, one of the most important underlying services is the Hypergeometric Data Visualizer or HDV. This technology is a breakthrough in program flow control and data traffic analysis. The HDV global servers can breakdown the N dimensional flow patterns of any software program, even if the application exists across a large number of different Web 2.0 style servers and user computer systems, and then convert the application quickly to a series of 3 dimensional moving displays digestible by the human eye and mind. Payment for access to the servers is negotiated in real time via the Dynamic Marketplace interface.]

HAL: The HDV is ready; do you want the usual colored wire frame display with filtering of non-relevant web elements and automatic motion freezing during important data traffic collisions?
JON: Yes Hal, thank you.

[Jon has chosen to do a diagnose and debug job today instead of a design task. A new set of screens spring into mid-air around Jon. It’s a wire frame representation of the program the user has assembled during complementary design sessions with their A.I. computer assistant. It resembles a giant city rendered in wire frame view like that in a graphic design package, except this city is alive with color-coded motion at a speed suited perfectly for Jon’s brain and visual pattern recognition capabilities. As per Jon’s personal taste, faster moving data traffic is being displayed as blue packets and slower ones are shaded red with a gradient of colors in between. These colored blobs of light whiz briskly along the wire frame city’s connecting pipes. With the help of their A.I. assistant the user has done a decent job of molding the application they desire, but there’s a problem. Under certain critical conditions the application fails. This is where Jon comes in.]

JON: Interesting. Hal, do you see the congestion in the blue area at Zeta (4, 7, 92) where the data packets are coming in from the yellow pipes?
HAL: Yes I do.
JON: Simulate a traffic load that is double the amount currently coming in from the yellow pipes in that sector.

[Immediately the activity from the yellow pipes speeds up and the packets change from a soft green to a troubling bright blue. The receiving orange spherical nodes, which represent servers processing the data, are pulsating in an unhealthy manner indicating that they can’t keep up with the traffic. What Jon doesn’t know, and doesn’t need to know, is that the yellow pipes are live sales data coming from cash register terminals in Spain that feed servers in the user’s Italian headquarters. It is now obvious that a surge of traffic coming from those feeds, due to periodic major sales promotions, is more than the company’s servers can handle. Jon will never know this, but the user is an Italian manufacturer with retail affiliates in Spain.]

JON: That’s it! When there’s a peak surge of traffic from those feeds the application’s receiving servers overload. Quick, recommend a doubling of capacity for those servers with a general suggestion to the user to stagger the data coming in from the yellow pipes in that sector.
HAL: Done! I’ve locked the solution and the user’s A.I. assistant has confirmed the solution as viable. You won the open bid and you just made $3000 in 15 minutes. I believe that’s a new record for you Jon.
JON: We make a good team Hal.
HAL: Thank you Jon.

[End Of Scene]

Notes On The Story

Many simplifications were made to portray this story. The data being represented could have been shown in a topological map style, or a 3D audio soundscape representation augmented by touch (haptic) peripherals, or any number of different ways. Rather than data traffic packets it could have been CPU threads, parallel execution paths, database replication patterns, or anything else. In addition, the real life data traffic patterns perceived by Jonathan would be far more complex than the simple e-commerce scenario given.

From The Abstract To The Concrete And Back Again

An important linguistics task performed by Web 3.0 A.I. assistants is the conversion of abstracted concepts back to the correct domain specific language that is appropriate for any particular user. Jonathan’s recommendation to the user regarding “the data coming in from the yellow pipes” would be translated into the appropriate concepts and ideas that pertain to the user’s actual application. For example, his suggestion to “stagger the data” might result in the user’s A.I. assistant suggesting that the Italian company should schedule upcoming sales events with enough time between them so as not to overload the point of sale equipment. Semantic Web software agents would handle this translation task but these software agents would not be the ultra-powerful A.I. systems that we will someday have when the Singularity arrives. Web 3.0 will make sure we don’t have to wait that long. They too would be assisted in real time by human elements that would be shunted in to handle the difficult parts of the translation, with the salient parts of the linguistic translations catalogued for later usage in future machine to human conversations.

The Virtualization Of Human Knowledge In Real Time

The important point is that once we “virtualize” the ability to represent web program execution and data traffic patterns, and then shunt in human intelligence at the very moment it is required to do the complex pattern recognition work computers currently can’t do, we will exponentially increase our ability to help each other solve problems and reduce the amount of unnecessary duplication of effort currently plaguing us in our pre Web 3.0 world. When the technology pioneered by Amazon with the Mechanical Turk service is enhanced by A.I. and is built into the very fabric of Web 3.0, it will advance our collective problem solving capability to a vastly more powerful level. This will allow us to truly leverage the skills and talents of every individual and distribute those skills at light speed on a global scale.

Global Leveraging Of Human Talent

Right now there are web designers that are forced to program and programmers that are forced to do user interface work. This is work that neither party wants to do and that could be better handled by the other. It’s not hard to extend that premise to millions of different jobs everywhere across every industry group there is. In addition, all of us are burdened and bogged down with the scheduling, management, and coordination of work between people; much of which is an unnecessary duplication of effort and requires slow painful interfacing during cumbersome meetings. Web 3.0 will radically eliminate and revolutionize much of that. We’ll still have meetings, but they will be much more fun and creative. We’re people, we will always like to meet.

One of the huge benefits of the virtualization of software development and the modularizing of human intelligence will be the automated cataloging and reuse of solutions. Since all parts of the development process will be codified, the Cognitive Abstraction Layer can record the vital statistics of each problem solved and keep track of which people and skill sets were involved, what recommendations were made, and what solutions best suited the problem. Therefore, when a similar problem is encountered again, these solutions can be offered and tried out in both simulation and real time to be provided to other users that encounter the same problem. In addition, job notifications will be automatically routed via their A.I. assistants to the people that have the most relevant skill sets.

Closing Words

The new distributed problem-solving paradigm described in this article will be a quantum leap in the reduction of duplicate effort currently occurring in our civilization, further accelerating exponentially the current rate of technological progress. As I type this document my Word processor is making advanced grammar and structure recommendations in real time. In the near future a physics teacher, a poet, or a legion of other people with different skill sets will help me create my next document, most of whom I will never meet. My A.I. assistant will present their knowledge to me at the exact moment it is needed and in the correct context.

I leave it to the reader to see how these core Web 3.0 advancements can be transposed and applied to many other fields, well outside that of software development. In addition, a quick search on the Web regarding new computer interface methods will turn up exciting new interfaces to spark your imagination such as brainwave scanners and more. In later articles in this series I will take a look at how artificial intelligence will evolve to enable these advancements and provide the supporting and binding elements of a brave new Web. I do not know how long it will take us to get from Web 3.0 to the first A.I. supercomputer that will mark the start of the Singularity. That is a question for someone with a much bigger brain than mine to answer, someone like Ray Kurzweil or David Gelernter. I suggest you watch the debate on machine consciousness they recently had to see these two great minds at work.

Search Blogs For More Information On:

Bookmark This Post!

BlinkList | | Digg it | ma.gnolia | RawSugar |
reddit | Shadows | Simpy | Spurl | Yahoo MyWeb


At 3:40 AM, Blogger Ian Parker said...

When Alan Turing invented his test he was hounded to suicide. Nowadays if you are gay tu puerde buscar un cerillo with online dating. To pass the Turing Test and/or do precision searches the meaning of words has to be better defined.

Web 3.0 is not a route to this. In fact as far as I can see it is just an expert distributed database. Oracle already does this. If you were to use Ocean Store you could simply create a fairly basic database that did all the things Weg 3.0 is supposed to. Note that in 3.0 there is NO facility for a spider to create links from the meaning it can see. No, Web designers have to supply all the data to an expert system. This to be is no improvement on what we already have.

To me the basic requirement is good renderings in a second language. If you have a 2L and look at a translation you can tell immediately whether progress has been made or not.

While you "buscas un cerillo" while stations have elasticity "La estacion de ressorte" it is idle to speak of any singularity.

At 10:53 AM, Anonymous Anonymous said...

Hey, can't figure out how to contact the author/owners of this blog, I'd appreciate you taking a look at the blog I am maintaining as complementary reading for an undergraduate class at Rutgers University called "Cybercities".

I would appreciate being on your blogroll if you found the content useful.


Post a Comment

<< Home