This post is about a US Air Force report about future technology which seems to foretell a well-known plot point in science fiction, the autonomous device that has been "insufficiently verified".
At various times, groups within our government attempt to give direction to the research and development that they are doing over the next decade. This is the sixth version of the USAF report, the first being issued in 1945 and co-authored by Dr. Theodore von Karman and General Hap Arnold.
In the words of the current authors:
"Technology Horizons" is neither a prediction of the future nor a forecast of a set of likely future scenarios. It is a rational assessment of what is credibly achievable from a technical perspective to give the Air Force capabilities that are suited for the strategic, technology, and budget environments of 2010-2030.
A link to the current version of this report, issued in 2010 is at the bottom of this post. There are a variety of very interesting things in this report but here are three statements near the beginning which I paraphrase here:
1. The USAF must pursue the use of autonomy (e.g. autonomous devices and systems) in an aggressive manner in all areas of its operations and work, far beyond what is currently being done today. Autonomous in this case means airplanes and equipment operating without humans aboard or directing their actions.
2. But the science and technology of "verification of this autonomy", in other words, how you know it will do the right thing, is far behind the science and technology of the autonomy itself.
3. And this is even more urgent since there are other countries who are far less picky about verification, and who are much more OK about things occasionally not working. Not working might mean blowing up the wrong building, or dropping on the wrong person, for example.
Thus, research and progress in the area of "verification of autonomous systems" so you don't blow up the wrong thing/person is the highest priority issue addressed by this report.
Now this is of course very funny as any student of science fiction or movies about the future is well aware.
The field of SF is filled with autonomous systems that go crazy or do the wrong thing with disastrous results. See for example Colossus: The Forbin Project, 2001: A Space Odyssey, the Terminator series, and literally hundreds of other examples could be cited.
Two of my favorites in this genre include the Ed 209 character from Robocop (1987) and Bomb 20 from Dark Star (1974).
We have on Youtube two versions of this famous boardroom test sequence. In the first version, we have the complete demonstration, but with much of the physical make up prosthetics at the end modified to suit those of more delicate sensibilities.
In this version of the famous board room sequence, notice the full use of practical makeup effects to contribute to the story. This is from the so-called director's cut.
Then there is the case of Bomb 20 in Dark Star. Who could ask for a better low budget prop ? And it has one of the more amusing speeches by a deranged synthetic intelligence in any film, maybe not as great as HAL 2000, who is our Hamlet of deranged synthetic intelligences, but very good nevertheless.
PINBACK (to bomb): But you can't explode in the bomb bay. Its foolish. You'll kill us all.
There is no reason for it.
BOMB 20: I am programmed to detonate in nine minutes.
Detonation will occur at the programmed time.
PINBACK: You won't consider another course of action, for instance,
just waiting around a while so we can disarm you?
BOMB 20: No.
Here is one of the scenes with our unverified Bomb 20.
There are other interesting technologies and surprises in the report.