Since the output is the Answer to the Ultimate Question, it necessarily incorporates or encodes every possible output of every possible program, including the string "Hello World!".
The method for extracting the particular output desired is left as an exercise for the reader.
AAL? Excuse me, are you defending a gratitious lawsuit based upon... what seems to be a lost cause, but whatever? Are you trying to be serious right now? I'm just a stupid scandinavian, but in our neck of the woods the civil court would certainly fine you, and possibly disbar you for wasting their time. Bullshit is never an excuse.
No, my comments were based on "exposure to the practice of law", by which I meant real-world experience, and not the principles and rules of professional conduct and ethics, of which you may take it that I am aware.
You should disabuse yourself of the notion that the law is exempt from the usual and ordinary disjunction that exists between theory and practice; it will not assist you if you do someday find yourself involved in a legal dispute.
The difficulty with jurors accessing material outside of the evidence led in court is that the party against whom that material is used has no means of challenging or testing it, since (a) it's not been disclosed to anyone outside the jury room and (b) even if it were disclosed, who's available to be cross-examined on that material? Are you going to go out and drag the journalist, dictionary editor, Wikipedia contributor, blog author, etc into court after both sides have closed their cases? How are you even going to be able to identify or locate that person?
Jurors (and judges, in judge-only trials) are limited to considering only what has been formally admitted into evidence to ensure that basic procedural fairness applies to all the material used in the case. Further, the public would be prejudiced in their ability to assess the result of the trial if they didn't know about all the evidence themselves, which can't happen unless it is all led in open court.
This doesn't stop jurors from applying the general knowledge and "common sense" of the ordinary person to resolve conflicts in the evidence as directed by the judge and after hearing counsels' arguments on the point, but the rules of evidence in most jurisdictions also say that if a party doesn't challenge a piece of evidence then the court is obliged to accept that evidence.
Finally, your point about criminal histories is based on a half-truth: criminal history is admissible when an accused brings character into issue, e.g. by suggesting that a prosecution witness is lying. Otherwise, it's on the prosecution to prove that the person committed this particular offence (at least it is outside the US).
Contingency fee arrangement of the "no win, no fee" variety are not champertous, and have always been allowed in Queensland (my home jurisdiction) were champerty was not abolished as it was elsewhere in the Commonwealth. What is prohibited as champertous are fee arrangements under which lawyers' fees are calculated as a percentage of the final award in a proceeding. It's a fine line, but that's what comes out of the cases.
I'm not sure of the precise historical reasons for the abolition of these torts in New South Wales, but I note that they also create difficulties for types of litigation funding that are considered legitimate, e.g. funding by public interest groups or unions. IIRC, union lobbying was behind much of the Victorian Wrongs Act, which (in its original form) abolished maintenance and champerty in that State as well as making a whole load of other changes in relation to civil liability for negligence and personal injury.
Any reason why a transcription pool couldn't be used for everyday voice recognition? If coupled with an automated system so that anything with a high confidence of being properly recognized isn't passed on to the humans...
If the transcript needs to be certified correct by an authorised recording/reporting officer, as is often the case, then that can only be done by the transcriptionist or someone watching the process. It would certainly be easier on the typist to sit and watch the software while listening to the audio, which could be paused while recognition errors are corrected manually, but a human would still need to be involved in the process.
But the mouth-hood is new and wasn't present in the digital tape era? That's odd, although it could be that there's been an unrelated change in the relevant rules for recording trials and evidence that say that body language and other non-verbal communication now has to be recorded in case it's of interest to an appellate court.
Of course, I've no idea which court you've done jury duty in and probably wouldn't be familiar with its practice and procedure anyway. I've just worked in courts that use a system similar to the one I originally described.
For me, text entry isn't that important a feature for a tablet; the mobile nature of the device makes it an unlikely choice of platform to generate documents of any length or complexity. Rather, the benefit of a tablet is the ability to consume or peruse data wherever I want.
To that end, I'm more interested in tools for tagging, noting up and generally scribbling on content generated elsewhere. Right now, I'll print drafts of documents just so I can have the freedom of leaning back in my chair or getting up and walking around while I review them, or putting them next to whatever (hard copy) source material I was using to create them to do side-by-side comparisons. Out of desktops, laptops and netbooks, no device lets me do that, and the screens on PDAs and smartphones are too small.
A screen that can take both stylus and touch input would likely fit the bill, with a virtual keyboard on the rare occasions I'd need it. We already have devices that handle extensive text entry in the conditions where that activity is best done.
Multics is security spelled sideways.