O Brave New Web
As the millennium clicks over from the aughts to the tens, one of the truisms that’s been circulating is about all the virtual things that weren’t just a decade back. Ten years ago there was no Facebook. No YouTube. No Twitter, tweeting. No iPods trilling with downloads. No Kindle, delivering books out of air. Soooo – if all these amazing toys have constituted themselves since A.D. 2000, what will roam the Internet dimension by A.D. 2020? According to observers and shapers of the cyber-world – a future-focused coterie that includes influential alumni and faculty of the University of Illinois – the Web is going to be bigger, better, faster, more connected, more convenient. It’s going to be everywhere.
It’s going to be wild.
“The Internet is a next-generation information transportation system – like a television/radio/fax system on steroids,” observed computer scientist and entrepreneur Roger Johnson ’65 ENG, MS ’66 ENG, PHD ’70 ENG. “It is the repository of information on what humankind knows about itself. But instead of being restricted – like the information in a library – it will be quite available.”
While in the Web’s first iteration users would navigate to static pages and read them, perhaps responding with e-mails and posting their own pages in turn, the far more interactive Web of today – nicknamed Web 2.0 – is a place where people link and blog and tweet and share and play and just hang out. Social networking has exploded into the stupendous popularity of Facebook (350 million active users). Online gaming has mutated into endlessly buildable, addictively immersive environments like “Second Life” (where an estimated 770,000 players populate a virtual world, having careers, building homes, traveling, performing, socializing and even earning actual money).
Now, in part through an evolving online architecture known as the Semantic Web, the next iteration is already shifting the focus back to the Internet itself. “Until now the Web has been people-operated,” observed science fiction writer Edward M. Lerner ’71 LAS, MS ’73 ENG. “The richer languages and data models of the Semantic Web will enable machine-centric access” – meaning computers will talk with other computers. Lerner, whose master’s from Illinois is in computer science, uses high-tech savvy to underpin his writing and has a fondness for futuristic worlds where rogue cyber-entities sometimes create mischief and chaos. In his 2008 novel “Fools’ Experiments,” a computer program develops awareness of other programs and decides to stalk and destroy them.
Anytime, anywhere – this evolution of the Internet is making disks and even flash drives the paraphernalia of the past.
“I think of Web 3.0 as a step toward an AI [artificial intelligence]-friendly Internet,” said Lerner, whose work extends through such multi-volume “space operas” as “Fleet of Worlds” and “Known Space” and includes novels written collaboratively with Nebula Award winner Larry Niven. “The technology envisions agents roaming the Web carrying out information-intensive tasks.”
In the good old days of Web 1.0, search engines (more quaintly known as spiders) went after search terms much as dogs fetch sticks – sticks that can come in bundles of hundreds of millions. Now – trained by semantic tools and stored data – the spiders are more like valets than pets. While a random search on the single word “Avatar,” made on Google in January, brought back 628 million Web results, the No. 1 hit showed the nearest theater screening the James Cameron movie of that name. Pretty good hunting.
A much more powerful generation of extremely bright search engines awaits in the online future. One such tool is already around – Wolfram Alpha, which can answer queries across subjects ranging from date, time and weather to socioeconomic indicators, regression analyses and DNA sequencing. By searching its own database – which grows daily more robust on input from a staff of researchers – the program can compute a Doppler shift, chart a genealogy and tell you what the weather was like on the day you were born. It can find and calculate nutritional information for recipes, translate words into Morse Code and solve math problems from simple algebra up to symbolic integration of formulas. With hundreds of capabilities, it appears to be the most powerful reference tool ever devised. It’s also online and free to the public.
“Alpha has a huge amount of information,” explained Theo Gray ’86 LAS, who, with UI faculty member Stephen Wolfram, founded Wolfram Research more than 20 years ago to market Mathematica, software for high-level computation. The company has now debuted Alpha as a way for prospective users to test out the complex, sophisticated capabilities of Mathematica. “Alpha knows what the data means, as opposed to a conventional search engine which doesn’t,” Gray observed. Searching with an engine like Google, he said, is akin to getting book recommendations from a reference librarian, while with Alpha, “we’ve actually taken the books out off the shelf and read them all and understood the formulas and digested the information that’s in there.
“The system is then able to use that information and synthesize it … and give you back the actual answer rather than just pointing you toward a book that contains the answer.” Though Alpha doesn’t search the Web firsthand, it’s clearly the precursor of future tools that will (perhaps to be named Beta and Gamma, among others?).
The programming techniques – semantic tagging, annotation, algorithms and categorization – that render Alpha’s database so searchable are also those used in building the Semantic Web. An ambitious venture to classify information and enable search engines to target and cross-reference that information, the Semantic Web has the potential to offer answers to very specific questions by penetrating the enormous proliferation of data that layers the Internet. At present its enactment entails a long, painstaking process – a human process, being carried out by people who have Web sites around the world – of line-by-line reading and standardized categorization of texts and images, supplemented by the programming work known as metadata, which links descriptions of the texts and images to other texts and images. (Though it is sometimes called Web 3.0, the Semantic Web is just one component of the future in cyberspace.)
Dan Roth, a researcher who leads the Cognitive Computation Group at Illinois, wants to speed up the search process by teaching computers not simply to read terms and tags and recognize connections but to comprehend language itself. The task is both monumentally difficult and stupendously interesting because so much of language is about context and ambiguity.
“If I write a program that can process news – for example, The New York Times – does that mean it can read blogs?” explained Roth, who also leads the Multimodal Information Access and Synthesis Center at Illinois, which researches artificial intelligence for the U.S. Department of Homeland Security. “Even for people this can be difficult. There are different language conventions, such as a lack of capitals [letters]. … There are the same problems involved as in moving from language to language.”
Roth’s program, at present, can understand text at a sentence level – recognizing, for example, that in the phrase “peace of cake” the word “peace” is misspelled. His work is proceeding on the theory that when a computer has ingested enough text and images (including Semantic Web-style annotated material), that computer will be able to make inferences from other text and images. And already “it’s doing things that I did not program it to do,” Roth observed, noting that “when you feed a computer an infinite amount of data, you cannot predict its behavior ahead of time.” Ultimately, he hopes to enable computers to read anything, anywhere on the Internet.
With 25 billion indexed Web pages, that’s a lot of territory to cover.
Anytime, anywhere – this evolution of the Internet is making disks and even flash drives the paraphernalia of the past, replaced by online data repositories that range from word-processing and document storage to photo-sharing sites and social networking groups. Housed in mega-computing resources, such as the National Center for Supercomputing Applications at Illinois, high-end databases allow researchers to share information about everything from children’s literature, plant names and the Mississippi River to astronomy, earthquake engineering and honeybees – to name just a few of the resources at the U of I. The University Library offers access to huge online archives of fine art and literature, which open new research possibilities. For example, through semantic tags, “you could find out the words that Jane Austen avoids, by comparing her work with work by other authors of the time,” noted John Unsworth, dean of the UI Graduate School of Library and Information Science.
Not only are data and software wicking up into cyberspace, the management of computer resources themselves is headed for the virtual stratosphere in a phenomenon called cloud computing. A “cloud” is a network of resources, housed on servers and enabled by the Internet, that provides storage and services on demand – like a utility, where customers draw off a grid not of watts but bytes, paying based on how much they consume. At Illinois, the computer science department hosts a Cloud Computing Testbed supported by industry partners including Yahoo, Hewlett-Packard and Intel.
Also Web-fueled are ventures such as GeoVector, a company with which Johnson is working to create what he describes as “a mouse on the surface of the Earth.” Using satellite and Web technology, GeoVector applications will enable cell-phone users to retrieve on-the-spot information about their whereabouts, getting the names of streets and structures by simply pointing and clicking. Johnson even envisions Web-connected binoculars that will allow visitors to national parks to read the names and histories associated with the scenery viewed through its lenses, playing simulations of, for example, what happened to Custer and his men at Little Big Horn, overlaid on the actual view of the historic battle site. The potential is “very, very wide,” said Johnson, for applications of such overlays upon the world. These he defines as “augmented reality” – a real-world view merged with virtual imagery.
Not only are data and software wicking up into cyberspace, the management of computer resources themselves is headed for the virtual stratosphere.
A founding member of SAI Technology – a Silicon Valley company specializing in mobile electronics for defense and intelligence – Johnson spent time at Illinois both as a student and faculty member. (He currently sits on the UI Alumni Association Board of Directors.) Johnson came of age in the labs of the UI College of Engineering, where he worked in the late ’60s and early ’70s on the early online learning system PLATO. Those were heady days at Illinois, days that sci fi writer Arthur C. Clarke saluted by making Urbana the birthplace of Hal, the rogue computer that inhabits “2001: A Space Odyssey.” This was the beginning of the Illini techie begats, a genealogy that numbers such stars of Internet innovation as Microsoft’s Ray Ozzie ’79 ENG (instant-messaging and e-mail), Marc Andreessen ’94 ENG (Web browsing) and the PayPal triumvirate of Steve Chen ’99 and Jawed Karim ’04 ENG (who together went on to create YouTube) and Max Levchin ’97 ENG (who has since devised that security box with the squiggly figures you need to retype to gain access to certain Web sites). That more geniuses will spring from the electrical and computer engineering and computer science departments at Illinois, as Athena from the head of Zeus, seems certain. What’s less clear are the new innovations to unfold. Back in mid-20th century, who would have foreseen the computer as a place to buy new shoes and run into old friends?
In those dawn years, Johnson said, “a lot of the motivation for the World Wide Web initially was to assist researchers in major universities and major government science centers in sharing information quickly and conveniently.” What started as point-to-point communications among “a thousand or so large computers” has grown, Johnson observed, into “a single network that shrouds the world.” Like a river system or a bloodstream, that network is penetrating with ever-greater depth and detail into human life, pulsing from desktops and laptops into cell phones and netbooks and even onto chips embedded in pets, to keep them from getting lost, and in people, to monitor their health.
For Johnson, a challenge that looms in the coming cyber-world will be the sheer proliferation of devices with unique Web addresses. “Where we’re headed is not just humans holding gadgets,” he said. “There are more and more pieces of equipment – cameras, sensors, robots that are part of this network – pieces of memory buried in the network. As large as or larger than the number of humans on the Net will be non-humans – pieces of equipment, from chips embedded in your body to things that observe individual processes. These are all unique addresses. The question will be how to accommodate them.”
Among the more auspicious Web-driven innovations sci-fi writer Lerner foresees are prosthetics for the brain, including neural interfaces to the Web (the blink of an eye or the slight movement of a finger allowing a user to navigate), implanted computer chips, contact-lens computer displays and human/nanotech hybrids. “And,” he pointed out, “I’ve written about all of them.”
Ever wary of Web-based entities, though, Lerner raises trustworthiness concerns to do with, for example, radio frequency identification devices (RFIDs). “There may be data gathered about me – and searchable through the Semantic Web – that I have no interest in sharing,” he noted. “Consider the RFIDs in my E-ZPass transponder, MetroCard ID, passport, clothes and maybe someday the currency in my wallet. And think of Semantic Web agents mining that history of everywhere I’ve been.
“To trust agents to do many things on our behalf – like make medical decisions – we’ll need, I believe, a much more secure Web,” he observed. “Without much more robust authentication, many-layered authorization systems and hacker-proof databases, lots can go wrong. Identity theft might go beyond impersonating you and me to impersonating our autonomous Web agents.”
“One thing that we have to attend to – a huge regulatory issue that we’re just starting to look at,” professor Roth likewise cautioned, “is trust. Trustworthiness.
“My kids get most of their information from the Web. They study from the Web. … So how do we develop and improve trustworthiness?” he asked.
What measures, for example, will help users assess the quality of the information they encounter on any given Web page, when sites vary from the meticulously edited stories of The Christian Science Monitor to the groupthink of Wikipedia? How do users protect themselves from the hostile viruses and spyware lurking in the Web environment?
And beyond is the Web itself, a network that connects the minds of billions of people around the globe, creating a super-entity that itself could be likened to a giant brain.
“System scientists and sf writers often point to the theory of emergence, in which complex behaviors emerge from a large number of simple items. Ants in an anthill. Workers in a bureaucracy. Neurons in a brain,” Lerner explained. “Will intelligent behavior emerge out of an aggregation of computers tied together by logic?”
It’s a question that may never be fully answered. With unique URLs surpassing 1 trillion, the Web now has 10 times as many places in it as the human brain has neurons. What human brain is going to figure THAT out?
Endlessly evolving, self-organizing, expanding and more and more intelligent, the Internet has become a new dimension, a dimension of virtuality, operating alongside time and space to contain and spit out information and (with some help from the UPS guy) stuff. The Web is connecting the world ever more closely to itself while broadening life’s scope at the same time. As the network clicks over and expands, its emergence will be understood and shaped in many ways and many places – including, undoubtedly, by alumni and faculty of the University of Illinois.