News and Reviews > Mini-Reviews by Members

Talking Moose - Mini-Review

(1/2) > >>

IainB:
I haven't used the standard template for this review as this software is just-for-fun and nostalgia.

I spotted this article in my Google Reader: Do you lose free speech rights if you speak using a computer?
Scholar argues computer-generated speech is not protected by the Constitution.

--- End quote ---
(It's an interesting article - worth a read IMHO.)
Looking at the headline, I idly thought "I wonder if the Talking Moose's speech is/was protected speech?"
So I googled "voice of the Moose on the Mac", and came up with a few hits.
If you used a Macintosh in 1986, you might have come across Talking Moose
I found that TM is alive and well at: http://www.halls.md/moose/index.htm
So I downloaded the proggy and installed it.

It seems that TM has not been just chewing his cud (do moose do that?) over the years. He's gone to a third owner, who has redeveloped him to run on the Windows platform.
It runs well, and TM now has a repertoire of over 1,000 witty phrases to speak. You can tweak most of his settings, and add to or edit his repertore. He has more functionality, so you can get him to do things such as, for example, tell you the time on the hour.

The voice is not as I recall - it used to be a man's rather German-sounding voice. However, this newer one is higher-pitched and sounds like the voices in the Renton police video spoofs (which I really enjoy), and the end-of-sentence tones tend to sound flat or rising, when they should usually be falling. This cues you to wait for the speech synthesiser to come up with the next word, until you realise that it has already finished.
I have missed the "old" TM voice though, saying things such as, for example:
"How come we never go out any more?"
--- End quote ---
- but this newer version is pretty good and no doubt could become just as irritating as the old version after you have heard the entire repertoire a few dozen times. That could take a while with over 1,000 phrases - I think the old one was only about 10 or so, as I recall.



How nice! TM just said the "How come we never go out any more?" phrase!    :D

TaoPhoenix:
I don't think this is about losing your protections if you vocoder speak through a text-to-speech engine.  I think it is more about computer-produced content.

"Articles published by the New York Times are often composed using word processors, and pages in the print newspaper are laid out using page layout software. The nytimes.com website is sent to readers by a Web server (a computer program) and rendered by a Web browser (also a computer program).

Of course, Wu isn't talking about those programs. He means programs that are directly involved in the production or selection of content."

However, that's still quite an article to write about, because as we get closer to certain kinds of AI, we're on the verge of computer generated news articles.

IainB:
However, that's still quite an article to write about, because as we get closer to certain kinds of AI, we're on the verge of computer generated news articles.
-TaoPhoenix (June 23, 2012, 12:13 PM)
--- End quote ---
Yep. Eggsactly.

TaoPhoenix:
Call it being eternally ruined by my Magic the Gathering days, but I no longer read news stories as stories. I read news stories as *Combos*. So if we Mash-Up Alan Turing and the Loebner contest, are we ready socially for what happens when we get into that level of scary futuristic law tangles? Is Computer Generated text treated differently under the Libel laws?

Right now it's "easy" with computers as "total slaves", but what happens if comps become "personalities"? "If I had X dollars" I think it's easier than it sounds to program a very basic AI system that might miss a lot of common sense questions but could really scare you on the big ticket topics. Very simply, my theory revolves using a systemically inferior hardware configuration such as the legendary Bugged Pentium (though it should be with today's comp processing power), and writing a core routine that, unlike normal comps, *CANNOT* "trust" its results, so that then the meta context must include self awareness that it is fallible, aka mortality. So then it must include the same kinds of caveats that people use because we can't process 1000 numbers is 12 minutes, so we go for "judgements". At the top level, it really isn't that hard, figuratively speaking 5,000 lines of code tops, which for an AI, is peanuts.

The inspiration for this is Ray Bradbury's (RIP) short story The Bicentennial Man.

So then we get into OZ Scarecrow discussions, "At what point do we consider a comp as a Living Item"? All that classic SF from 1940-1980 is gonna show up soon.

IainB:
All that classic SF from 1940-1980 is gonna show up soon.
-TaoPhoenix (June 23, 2012, 02:39 PM)
--- End quote ---
You could well be proven right.
That period of SF (science fiction) was highly speculative, and the authors turned ideas and problems about science and society every which way and upside down, inspecting them and asking the question "What if...?"
Some scientific predictions came out of that period - for example, including (I think) sputniks, mobile phones, geostationary space elevators, wristwatch-based computer libraries, the Laws of Robotics, the use of a methodology called psychohistory.

Navigation

[0] Message Index

[#] Next page