How is the weather, the computer asks

Collective artificial intelligence for the US militar

The "department of advanced defense projects" the ministry of defense has laid the foundation of today’s internet over 30 years ago. Now she works on projects, the foundation for a collective – artificial – intelligence.

As appreciative, the future prognosis of an organization, which develops something like electric dog sinks to trace mines? In the case of DARPA pretty. The defense advanced research projects agency was founded in 1958 by the US department of defense, when the soviet union demonstrated her technical superiority with the start of the sputnik satellite. The task of DARPA is to demand projects that are so futuristic that they do not necessarily act basic. "If someone suggests something that is clearly to realize, the DARPA does not interest", a scientist described the approach.

In which projects the DARPA is their freely budget of about two billion dollars, just says a bit about the future of information technology and the internet. After all, it was a project of darpa, which in 1969 with the arpanet laid the foundation for today’s network. DARPA also finances partly the world wide web consortium rounded in 1994 and is $ 30 million in the project next generation internet (NGI), a network based on new software and hardware, which according to DARPA scientists with a 1000 times high speed as the current network in five to ten years should also be available for private users.

The actual speed problem of the future is not the data transfer rate in the art network or the speed of the connected computers – it is the person. The senior DARPA scientist david tennenhaus already explained in the previous year at a conference. "Today, computers interact with people, these people then with the physical world, which slows down the process of data analysis", believe tennenhaus. Simply put, the goal of DARPA is to develop systems that can be interacted directly with the augenwelt. So systems that have their own sensory organs and make their own decisions on the basis of the incoming data. Another possibility to deal with the tremendous amounts of data of the future efficiently does not see tennenhaus.

Where these data should come from? The DARPA believes that the miniaturization of the battlefield of the future will fly a few hundred thousand tiny sensors such as pollen, which provide substantial information. A DARPA project is working on network architectures that connect up to 100,000 components and encourage them to act as collective intelligence in second fractions. For example, a sensor swarm buoys over enemy lines and tries to get radio traffic. A sensor shares his discovery of all others, whereupon they collectively change their work and radio traffic states.

Another example is a weather observation system that recognizes a previously unused situation. Instead of just reporting this, it now gets observation data from all weather stations of the earth and tries to recognize a pattern. A flow of such systems is the airbag or the ABS system. Again, electronics respond to sensory impressions. Much faster than a person has ever been able to emphasize tennenhaus.

A level deeper set the DARPA projects tides (translingual information detection, extraction and summarization) and information management (IM). Here, an artificial intelligence of self-contemporary information is to search and work up, so that a person can make decisions on this basis.

Tides describes as a vision the handling of a fictional crisis in northern lisland: hardly anyone in the US defense ministry speaks northernis, and no one has an idea of the present crisis situation and their causes. Tides successfully captures all northernistan websites, including printed documents such as newspapers about typeface recognition. Within a month, the program not only provides translations, but above all collecting and clearly structured information on important people and present developments. The improgram is intended to browse all available sources under a given problem and steer the attention of man for information from highest relevance.

How artificial intelligence problems can go solely, instead of being able to prepare for people, trying to show the project autnomous negotating teams (ANT). Complex problems are not approached globally by a central instance, but in smaller decisions that are decided by local instances, split up. A possible example: A calculator gets the abstract target: landtrupming in three days. He shares this the systems of the involved troop parts with. They put the abstract goal in more concrete tasks. For example: in three days 40 kilometers to the ubitzplatz. The systems of the tanks check the possibilities to achieve the task. Panzer X answers that his participation is not possible due to a technical defect. Panzer y wants gasoline. Panzer Z requires a particular spare part, as he was terrified, the best before date, during the manovers.

On the internet, the project active networks is a step towards artificial intelligence. The aim is information networks that automatically customize their capacities. The network is aware of its structure and the information transported therein and decides on this base itself.

Awareness of the type of information transported could perhaps the computer language DARPA agent markup language (DAML) developed by a DARPA project could be measured by a DARPA project. She should be in the net for artificial intelligences what is english today for natural. Daml was in the final stage of artificial intelligence in the network access to the code of computer programs, on sensors and all machines that are affiliated. Thus, tennenbaum vision became reality: computers no longer need people than mediators to the physical world.

Physical world involves more than you are currently presenting yourself. One of the rather long-term DARPA projects deals with the possibilities of an interplanetary internet project manager mari maeda: "that’s not something that happens in the next three or four years, but we have to think about it. It will definitely be a significant topic in 20 years."

Like this post? Please share to your friends:
Leave a Reply

;-) :| :x :twisted: :smile: :shock: :sad: :roll: :razz: :oops: :o :mrgreen: :lol: :idea: :grin: :evil: :cry: :cool: :arrow: :???: :?: :!: