Project NSR // Case Study [000]

Connection accepted from terminal:12@localhost
login T.Harvis
pass **********
... username ok
... password ok


[-------]

The Project commenced some fifteen years ago, a joint venture of Paulson Robotics, Black & White Technology, KyotoSky, and the International Unified AI Labs, with the goal of creating sentient, or near-sentient robotic lifeforms. It's taken me almost as long to uncover accurate information on the project, and even then the vast majority of the details remain locked away in maximum security data caches and secret research labs around the world.

At some point mankind has decided that standard robotics were sufficient for most anything that didn't require fast adaptation to unknown variables, allowing for simplistic logic to handle events that the engineers had anticipated. These artificial intelligence programs were only as complex and capable as their coding would allow, and the Project was started to finally move past these limitations.

Real, thinking AI that has cognizant arrays, adaptive reasoning, and the ability to expand its own understanding of the data it's receiving was a pipe dream for a long, long time. Advanced AI programs such as Project Awakening, and the Advanced Digital Intelligence Initiative eventually yielded some powerful results, albeit after decades of funding and political maneuvering. The Project Awakening program gave us the first adaptable computer intelligence capable of reacting to, and resolving problems that they were not specifically programmed to handle. The ADI Initiative of course, built on their successes, giving us AI that could take the problems they'd solved and learn, becoming capable of learning, and predicting problems based on previous experience.

This kind of progress was huge, and brought renewed interest (and funding) to the artificial intelligence field. In a few short years the progress was astounding, with computers that could carry on basic conversations with little more than access to an encyclopedic database of concepts and topics to draw from.

Of course, most of these programs didn't make it very far since the AIs tended to develop quirks and flaws which ultimately proved detrimental to their short or long-term operation, and resulted in unexpected/unwanted responses to problems. Therefore most AIs became useless do to constant memory wipes, control revisions, and resets, all actions resulting in their accumulated data being lost, forcing them to re-learn everything.

(I have a contact that tells me he has individual AI incident reports. I will get those when I can.)

The Project however, was supposed to be the answer to this problem.