How will Yammer outsmart Jive Software

Search and not find

The internet brings the knowledge of the world to us on the screen. So the promise. Meanwhile, we are drowning in information. A search at the limits of receptivity.

Text: Thomas Ramge
Illustration: Thea Barkhoff



I. Information wealth

Tobias Arns has to give his opponents one thing: "They know how to use metaphors and negative terms to steer the public discourse about digital media in their direction." Arns means figurative terms like "data flood" or "information tsunami". One of his opponents is, for example, Manfred Spitzer, the brain researcher with a penchant for pointed interpretation of research results. A particularly perfidious linguistic image is the battle term coined by Spitzer of "digital dementia", which threatens our children acutely because they are permanently "carried away by data streams".

Tobias Arns, a man in his mid-thirties with dark glasses and a creative industry streak, knows how to handle data and words. After graduating as an IT specialist, he studied communication science and German language and literature. In his profile on the professional network Xing, under "I offer" it says: All Things Digital.

Arns does not translate "Information Overload", that semantic hit by the American futurologist Alvin Toffler, as Wikipedia does as "information overload", but as "information wealth". He felt this wealth "without restriction and in the literal sense of the word as an enrichment". He demands: "We finally have to turn the question around: What do we miss if we do not use the new variety of information with enthusiasm?" Now, however, it is part of Arns' job to sing the hymns of the digital revolution: He is the head of social media and mobile for the lobby group of the German IT industry, the Bitkom IT and communications industry association.

The second stanza of the digital song of songs goes something like this: We just have to learn to set our information filters correctly. We have to learn to ask the right questions of the giant digital machine. Then we become part of an information system with unlimited possibilities. The Internet theorist Clay Shirky summarized this in the formula: "It's not information overload. It's filter failure." In good German: Anyone who feels overwhelmed by information is just too stupid to use the technical tools with which the knowledgeable unearth the treasures from the mass of data.

II. Data paralysis

Carl Frech has a slightly different view of the digital world, although he has probably set his filters well. For more than 20 years, the communication designer has been dealing with the question: How do we pass information to users in an understandable way? For many years Frech was a board member at the Meta-Design agency, responsible for software development, interface design, online strategy and mobile technologies, among other things. Today he works as a consultant and also teaches at various universities. Frech has a very clear opinion on Shirky's statement: "To dismiss being overwhelmed by the exploding range of information as a lack of competence of predominantly older people in dealing with new media is stupid and arrogant."

The professor of information design observes an interesting contradiction among his students, who were the first generation to grow up with the Internet. "On the one hand, they present themselves as virtuosos of social media. On the other hand, they hardly dare to come up with their own thoughts. Because they click through the network from source to source and think: Everything has already been thought and done has been written down. What else should I contribute? " Then, according to Naughty, they used "copy" and "paste".

The problem for the generation born into the digital world is therefore much bigger than one might think. Frech diagnosed his students with "an upstream depression in the digital perma activity". In other words: Too much Facebook not only makes you unhappy because the messages from "friends" make you realize your own incompleteness, as sociologists at the University of Utah recently demonstrated in a comprehensive study. The infinite research possibilities are also a problem for the searcher: The reference systems to more and more information paralyze the difficulty of distinguishing important from unimportant, relevant from irrelevant, which grows with the amount of information. Data analysis becomes data paralysis. Frech calls the episode the "recycler mentality". In the oversupply of information, we, whether young or old, quickly scrape together apparently relevant things. "The sense of quality often falls by the wayside."

The bottom line is that the data overload hits modern people twice.
• First, information pours down on us in ever larger amounts and at ever shorter intervals through e-mails, feeds in social media, messages in chat, breaking news on news sites or status reports in professional networks. Experts call this wave "push communication".
• Second, we need more and more time and competence to find the information that is relevant to us. If we are actively looking for them in "pull mode".

The common thing about these two waves of information is that they break in on us simultaneously. In order not to drown in them, we need two different life rafts. For push communication, these are filters and the ability to disengage from communication at the right moment without missing anything really important. For pull mode, it is search competence combined with more intelligent search systems.

III. Attention management

"A third of people say: there is no data overload. Two thirds say: The information world is collapsing right now. Neither of these helps," analyzes Craig Roth, an expert in communication and digital collaboration at the US IT company. Consulting Gartner. "We have to regain control over our communication. That can only be done through better attention management," he said.

Attention is a finite good. Digital communication systems, from e-mail programs to social networks, are programmed in such a way that they attract as much attention as possible. The human sender behaves the same way in digital space. He uses the opportunity to send his messages comfortably, quickly and (almost) free of charge, regardless of the loss to the recipient. The pre-setting of man and machine in digital space is set to send, and as much as possible.

Now the recipient could ignore the messages. In our private life, says Roth, we are making increasing use of this option. We don't suffer any material damage if we don't look on Facebook for a few days. At work, on the other hand, it is not so easy to disconnect from the flow of information - especially not in hierarchical structures. This is where, according to Roth, the real problem begins: "The number of e-mails is increasing. The expectation that the recipient will also read them remains the same. The limits of the receptivity have long been reached, however."

Good management must first recognize "that we are dealing with a problem of both a cultural and a technical nature". That data flooding is not a natural disaster against which there is no protection. Rather, it is one of the tasks of CEOs and IT managers in companies to change communication and technology in such a way that employees do not drown in information.

This includes binding rules for sufficiently long response times, economical use of cc-mails or uniform marking and filing of the really important content. On the technical level, however, Craig Roth sees the bigger levers. This in turn includes the use of modern software, for example for working together on texts or for scheduling appointments and intelligent deletion processes. The latter automatically remove outdated data and thus ensure that "the digital haystack does not grow immeasurably".

So-called social software, which transfers the logic of social networks like Facebook to companies, can also make communication more efficient. This is surprising at first, because company-internal social networks such as Jive or Yammer initially create a new channel with a new flow of information that employees then have to keep an eye on. According to Roth, the gain in efficiency lies in the fact that the senders no longer expect all recipients to read everything. "The system is designed so that the really important things are picked up, discussed and posted on by others." In plain language: If the data stream is unmanageable, we at least create an environment in which the unimportant messages are sent and the important messages fight for attention based on the principle of resubmission.

This does not solve the problem of a more targeted search in the exponentially growing amounts of information. Good filters, human or mechanical, or a combination of both, can keep useless information away from us. However, they do not help us to find the essentials.

IV. Demonstration of power of knowledge

We're heading for the San Francisco skyline. The skyscraper is getting closer and closer. Stefan Keuchel, the press spokesman for Google Germany, taps the control button to the left and we fly through a canyon towards the Golden Gate Bridge. At least that's how it looks on the head-high panorama screens that surround us in the foyer of Google's German headquarters. He says: "The skyscrapers are half real, half animated. This is our further development of Google Earth."

In the canteen, Keuchel would like to talk about what he calls the "future of search", thus giving a different framework to the question of data overload. He begins the conversation with a sentence that a press spokesman for Daimler, Siemens or Beiersdorf would probably never say: "We don't think our product is good today." That doesn't sound modest though. He continues, "The perfect search engine knows exactly what you mean when you ask a question. And it delivers an entirely satisfactory answer that is relevant to you."

The company owes its breakthrough to an idea as simple as it is brilliant that one of the two founders, Larry Page, had as a student. In search queries, Google not only counted the number of relevant words - called syntactic search in technical jargon. The patented "Page Rank" algorithm counts and weights the links that refer to a website - and thus leads to more relevant search results. So far, so well known, from the long search hit lists with links marked in blue, which Google shows us with every search and which we then obediently follow before we find what we are looking for or come across things we did not suspect. that they exist, but they are still relevant. Or just not find anything and give up in frustration because the abundance of offers does not seem to provide a relevant answer and the feeling creeps up on us: We are too stupid to search.

Stefan Keuchel picks up his cell phone and demonstrates what technology can already do today. "How high is Cologne Cathedral?" He asks. The cell phone answers: "157 meters." Keuchel says to the pocket machine: "Show me pictures of the Berlin TV tower at sunset." Dozens of photos appear on the display that made the hearts of Berlin city advertisers beat faster. Keuchel opens a laptop and Googles an unsolved Sudoku puzzle. He finds one on the website of "Zeit Online". He photographs the puzzle with his mobile phone, taps "Solve" with his right index finger, and the - presumably correct - answer numbers appear immediately in the previously empty Sudoku fields. The Google spokesman smiles cautiously. The demonstration of power of knowledge continues.

He calls up the brand eins website on the laptop. He clicks somewhere, and the entire page, including all sub-pages, is translated into Arabic. The translation is certainly not yet perfect, but every Arabic reader will be able to grasp the meaning by and large. Keuchel continues to smile, this time a bit self-deprecatingly, and says: "Now comes a big-pants announcement: Google has solved the problem of multilingualism in the world." He speaks again on his cell phone: "Where is the next train station?" The mobile phone says in a British female voice: "Where is the next train station?" Keuchel says: "Over there, around the corner." This time the device answers in a German female voice: "Over there, around the corner." It works in Chinese too. Unfortunately, neither of us could check that.

A digital everyday all-round assistant should be able to do all of this in the future. Expertise in answering everyday questions, if you will. At least that seems to be Google's current vision. Another current product is an app called Google Now. She gives the frequent traveler HSV fan and lover of the band The Prodigy Stefan Keuchel, who lives near the port of Hamburg, answers to questions he never asked. For example, that it is time to check in online for the flight to Brussels or to set off earlier because there is a traffic jam in the direction of the airport. The app is connected to Keuchel's online calendar. It gets information about the flight from the connected booking system, the traffic jam information from traffic information pages. The app shows the expected HSV team line-up for the upcoming home game and informs about when The Prodigy will next play nearby and how much the tickets will cost.

Unfortunately, the Google spokesman can only demonstrate the climax of the search for the future of search by video. There are still no "Google Glasses" at the German headquarters. These are interactive data glasses that users can use to stroll through the real world. You can ask questions orally to the glasses. The answer then appears in the field of vision, calculated using search algorithms in a data-recorded world. In the commercial it looks like this: The seeker gets up in the morning and looks out the window. The glasses show him the current temperature and the forecast for the day. She tells him at the entrance to the subway that the train isn't running today, and at the entrance to the bookstore, where he can find books on music history. Stefan Keuchel sums up: "A good search recognizes the context in which the user needs information. And it is our job to display the information in such a way that the searcher does not feel overwhelmed."

In an interview in 2004, the two Google founders Larry Page and Sergej Brin tried to think the future of search through to the end. At that time, they came to the conclusion that computer intelligence will at some point find its way directly into people's heads, presumably with the help of an implanted chip. Man only has to think a question, and the machine feeds the answer directly into the brain. The young founders did not think through the ethical questions to the end in this interview. They preferred to indulge in fantasies of omnipotence. A global corporation of a previously unknown kind could be built on these. The Google glasses, if they should work as well as in the promotional video in the near future, would be the everyday version of the omniscient man-machine.

V. Data, words, meaning

The seeker of the present does not wear data glasses. He wants to get more relevant answers faster when he searches for information on the net. As a technical solution to this problem, the vision of the "semantic network" has haunted the IT world for around ten years. The inventor of the World Wide Web, the British Tim Berners-Lee, made it his second life mission.

Semantic network means that thanks to intelligent notes on a second level (so-called metadata), the content of websites can be interpreted by machines. A semantic search engine could provide a concrete answer to a halfway complex question. So she couldn't just say how high the Cologne Cathedral is or when Angela Merkel married. When asked "What is the semantic network?" automatically deliver a coherent, concise and generally understandable answer. And ideally take into account the previous knowledge of the seeker. This machine is not yet in sight.

How far the semantic search is in German can be seen in a building with the floor plan of a hockey stick. It belongs to the Hasso Plattner Institute for Software System Technology in Potsdam, HPI. In the futuristic stairwell in the foyer, sayings from historical great thinkers show the young generation the way to knowledge. On the top floor you can read: "One learns astray. J. W. von Goethe". Two floors below, the scientific staff of the federally funded Mediaglobe project show what their machine can do.

The HPI employees have stored 10,000 hours of GDR documentaries from the Defa archive on a server somewhere in the data cloud. Intelligent software analyzed the soundtrack using voice recognition and noted key catchwords - with the associated time in the film. The characters shown in the picture, for example names of interviewees, were automatically recorded. All new data was linked with the existing one from keyword catalogs.As a result, if you enter the search words "Erich Honecker" and "Brother states" at Mediaglobe, you no longer have to sift through a stack of video cassettes quickly to find the right film about the former head of state of the GDR. Instead, he receives a list with exact minutes. In the screen window next to it he finds a transcript of the soundtrack. If the sequence is relevant, he can of course start it directly with a click.

Christoph Meinel, the head of the HPI, is a globally sought-after expert in semantic search. A conversation with him reminds you: data flooding does not have to be a digital phenomenon. Technical terms go into complex sentence structures. The speed of speaking increases parallel to the enthusiasm for a certain sub-aspect. It is the task of the listener to discover the essentials in the jumble of words and not lose sight of the big picture. It is not easy to separate the essential from the insignificant, because everything somehow appears essential.

A roughly filtered excerpt from a two-hour conversation with mathematician and computer scientist Meinel about data flooding looks something like this:

1. The basic problem is not new. A Ptolemaic scholar also felt himself to be in 250 BC. struck by the wealth of information in the Alexandria library.

2. Nevertheless, the world has changed due to the explosion in the amount of information. The acceleration of information and communication through digitization has no historical precedent. So we don't have a historical blueprint for how to deal with one of the many aspects of this phenomenon, data inundation.

3. From today's point of view, there will be no technical miracle weapon, no super algorithm, no super search engine that gives us all the answers to all questions in bits and pieces of the desired size.

4. The sum of the tools brings the approximate solution. This can be a semantic film sequence search, as with Mediaglobe. Or in the case of lecture videos, an additional bar on which the speaker's slides are faded in (and of course also semantically evaluated). It can be an application that doctors can use at the bedside to look up which therapy is likely to help with this clinical picture - and which searches medical databases in no time at all. Or a social network component in a digital learning environment for students.

5. Used intelligently in a bundle and situationally, these tools reorganize the knowledge of the world.

At the end of the conversation, Meinel comes back to the historically unique acceleration of information and communication. He says: "We don't need to believe that we will get this change under control in one generation. The uncertainty in handling data will be with us for a while."

If we humans cannot cope with the data overload ourselves, will we build increasingly intelligent machines whose attention knows no bounds thanks to storage capacity and computing speed? They don't have to be connected to a chip in the brain, as the Google founders Page and Brin predicted. The quiz computer "Watson" from IBM opens the door to the digital search-and-find world of tomorrow. He understands the meaning of the complex questions of the quiz master of the US show "Jeopardy" about everyday life and the way things are. And he answers them faster and more precisely than any human super-brain, based on the database of a few newspaper years and Wikipedia.

Isn't the question of data overload, as the super-smart Watson shows, one that is essentially aimed at the relationship between man and machine?
Christoph Meinel has to think for a moment. He says, "The semantic decoding performance of Watson is impressive. But it remains an auxiliary application for a context with very precise rules, namely the quiz rules of 'Jeopardy'."

Knowledge, i.e. the connection and interpretation of information, only arises in the head of the human being. "For the foreseeable future, thinking will not be able to take away any machine from us."

Conversely, this means that the limits of receptivity remain those of the human brain, the most complex product of evolution. And not the limits of the digital machine, the most complex product of technical innovation.

The currently most ambitious project for building a search engine with semantic technology is called Wolfram Alpha. If you enter the search term "semantic web", you will unfortunately not get an answer at all. Instead, he will be asked to leave his email address. The machine forgets to say "please". But at least it provides a reason why we should leave our contact details: "To show interest." Whatever this could mean.

More and more programs are transferring the idea of ​​Facebook, Twitter or Wikipedia to the needs of companies. They are intended to promote a constant exchange of information between employees, especially across departmental boundaries. Some of these networks are strongly based on common social networks in terms of their mechanics and interface design. Others focus on collaboration tools and collective project management functions. Well-known social software solutions include - in addition to Jive and Yammer - Atlassian Confluence, Blue Kiwi, Microsoft Sharepoint, IBM Lotus Connections and Socialtext. Google makes some collaboration tools such as "Docs", "Doodle" or "Calendar" available to companies and private individuals free of charge. Here, the user pays with their data.