Suppose the Goodman Company suffers a devastating hacker attack, and as a result loses some lucrative deals to Rogue Industries, its major rival in the industry. The FBI traces the hacker attack to Wily Whizzkids, a boutique software engineering firm, whose eccentric young boss, nicknamed "the Magic Rat," charges exorbitant fees but has a reputation for developing ingenious solutions to computing problems which have stumped everyone else. When it is further discovered that Rogue just concluded an ultra-lucrative deal with Whizzkids, the FBI arrests the Magic Rat, and a trial date is set.
The Rat does not seem too worried. He has a story. He admits to negligence, and offers to plead guilty in a civil suit and pay damages. But he pleads not guilty of any criminal intent or of deliberate sabotage. It all started, the Rat explains, which a new approach to software solutions which he was trying to pioneer. He calls it "evolutionary software engineering." Instead of having his engineers plan every detail of how he'll solve a problem, he has created an "ecology" of self-modifying, competing algorithms which reproduce and recombine in a computer's memory, setting tasks for each other and measuring each other's "fitness," modifying each other, and generating new code and new ways of solving problems. Most of the output, says the Rat, is junk, but the 1 in 1000 innovations that are useful can be just what his engineers never would have thought of, and give him an edge over the competition. Unfortunately, some of this loose, evolving code escaped from the laboratory and did damage in the real world. Yes, the Rat admits, it certainly looked like an intelligently-designed and deliberate hacker attack. But that's just the thing about evolutionary computing: design emerges spontaneously; it looks like there's a plan, but there isn't. Just like life! However, since he hadn't anticipated how dangerous his experiment could be, he now plans to shut it down until he's figured out a better way to control and contain it.
And as for that deal with Rogue-- well, that's just a coincidence, explains the Rat. Yes, he drove a hard bargain, and he expects the contract to be extremely profitable. But Rogue placed a high value on the project, and they made a judgment-- plausible, certainly, and in the Rat's opinion correct-- that no one in the industry except Wily Whizzkids could execute it.
The Goodman Company's lawyers take a hard look at the case and report their advice to Goodman's executives. It is absolutely critical that the Magic Rat be found guilty in a criminal suit. Goodman invested a lot in pursuing the deals that they lost to Rogue because of the attack, but many of those investments were intangible and it will difficult or impossible to persuade the court that all of them may be properly counted as damages. And the deals would have been far more beneficial to Goodman than they appear, because their main value was to help Goodman cultivate new clients and get into new lines of business. If the court merely makes Whizzkids pay damages for Goodman's direct costs and direct lost revenues from the attack, then Rogue wins and Goodman loses-- and for that matter, the Magic Rat wins too, since assuming the Rogue deal really was tied to Whizzkids deliberately doing the hacker attack, he'll make far more profit from that deal than he has to pay in damages. If, on the other hand, the Magic Rat loses a criminal suit, then Rogue, too, can be implicated and made to pay, and Goodman may claw back some of the deals.
In a criminal suit, the burden on the prosecution will be to prove "beyond a reasonable doubt" that that the Magic Rat is lying, that the hacker attack did not emerge spontaneously from the Rat's evolutionary computing laboratory, but was deliberately designed to attack Goodman as part of a secret deal with Rogue. Of course, the Rat's e-mails and phone records and computers and notebooks and whatnot will be investigated, but it seems pretty certain that the Rat was clever enough to cover this tracks thoroughly. He will no doubt have planted prototypes of the attacking program in the data stash of his laboratory, and if the whole history of that program's emergence is not traceable, he will easily explain why the data retention policy was not such that the emerging programs typically are traceable. Rather, the investigation will have to be conducted at a more abstract level. What is the Magic Rat's evolutionary computing laboratory really capable of? Could it have generated the hacker program?
Now, I think the following two statements would correctly describe this case:
1. There probably is a right answer. Maybe if there was a 1% chance that the laboratory would generate a hacker attack on Goodman, we might accept the Rat's story. The odds will probably be much more than that-- perhaps almost 100%, near-inevitability-- or else far less. They might be strictly zero, or they might be 1 in 10^10 or something, so improbable that "beyond a reasonable doubt" the program was deliberately programmed by the Rat. Some civilization of the future whose scientific enlightenment and computing capacities are a thousand orders of magnitude greater than ours could, in principle, prove the Rat guilty beyond any doubt.
2. To figure out that right answer will be an extremely difficult philosophico-logico-mathematico-computational problem in the properties of complex systems. Even if we know everything about how the Magic Rat's laboratory was set up-- and the FBI can get that information-- to determine the likelihood of a particular outcome spontaneously emerging, and therefore whether a claim of spontaneous emergence is credible, or whether it can be ruled out in favor of an intelligent design hypothesis, will involve enormously complex theorizing and computational testing.
Modern evolutionary propagandists like Daniel Dennett may be compared to a gullible jury that gets carried away by the Magic Rat's rhetoric and vision and becomes instantly certain of his innocence. They don't really have the faintest clue about whether or not the mechanisms of mutation and natural selection that have been proposed are powerful enough to explain the complexity of life that we observe, but they are thrilled by the vision, and run with it.