It looks like you're using an Ad Blocker.
Please white-list or disable AboveTopSecret.com in your ad-blocking tool.
Some features of ATS will be disabled while you continue to use an ad-blocker.
originally posted by: StallionDuck
There was a game I played back in the early 2000s called America's Army. It was a FPS like Call of Duty. It was made by the ARMY to help train the troops. They let it out to the public, possibly to help beta the game. It was free to play and massively popular. The graphics were well ahead of it's time. The enemy always looked like the Taliban no matter what side you were on. It was a really fun filled, beautiful and very detailed.
Fast forward to today....
So we know that these games were tailored to help train the troops for combat and what to expect in the heat of battle. We know this because the ARMY already bragged about it back in 2000. Now, AA has spanned across many platforms and became not just a free game to help the troops but pretty much wide open for everyone. I believe it was being sold at some point and no longer free to play.
So sure... They take all of that information, dump it in a database. Add such games like Halo, CoD and others, and now you have a huge database on how to accomplish missions and how to achieve goals with the least amount of death while putting out a mad amount of destruction. Playing these games for so many years, I can see how the data would be effective from those that play very well.
Now.. Finally. Why stick with troop training? That's a thing of the past. Now you take all of that data and toss it into drones, robotic soldiers or simply use those people who are the best at these games and put them behind a remote control to pilot these same bots to do what they enjoy.. Kill the enemy.
I can see how it all plays out rather well. This is not hard to believe one bit. Matter of fact, I believe that was the plot all along.
Regarding the Google research, scientists were trying out a simplistic sequence in order to "teach" the machine possible responses. "In this paper, we present a simple approach for this task which uses the recently proposed sequence to sequence framework," researchers said in the abstract of their paper. "Our model converses by predicting the next sentence given the previous sentence or sentences in a conversation. The strength of our model is that it can be trained end-to-end and thus requires much fewer hand-crafted rules."