Google wants to build the Star Trek computer
slate.com
slate.com
It wasn't until TNG that the computer became a friendly, integral part of the Enterprise. The crew widely acknowledged and were comfortable with the fact that the computer could do all of their jobs, but that never became an issue.
This shift corresponds with the rise of the personal computer away from mainframes, and the replacement of blinking lights with GUIs for depictions of computers in film and television. I will now resist the urge to credit Apple for helping usher in such a shift in culture.
...until Barclay became the computer (and could presumably do all of their jobs at the same time), though that didn't last too long. https://www.youtube.com/watch?v=g4AEBiyn8Rs
The primary problem with copying the star trek computer is the user interaction. One of the biggest reasons all of the star trek's utilize a voice command computer is ...
For directing.
Otherwise when the crew entered a command, recorded a log, etc. they'd either always be shouting it out or making us watch / read the creation of text. It was quite brilliant from a directing perspective, but do you want everyone around knowing how many times you go to facebook/hacker news/gmail all the time? For that matter, personal log entries etc. Yes it has its place when you're in isolation, but its quite difficult when around others.
So for me, its become exactly like the computer from TNG.
I've especially found voice search to be the more socially acceptable way to search for something when conversing with a group of people. For instance, during a discussion about movies at a bar with friends, to tap the phone and ask it a question fits into the dynamic far better than putting your head down to tap it into a keyboard and wait for results.
Asking the question aloud shows everyone that you didn't suddenly start texting someone or checking facebook mid-conversation, but rather searching for something in order to add to that conversation. Same goes for "what song is this?".
I almost forgot my all time favorite while driving (whether alone or with a car full of people): "Navigate Home"
The project was originally known as "Majel," after Majel Barrett, the wife of Gene Roddenberry, who was the voice of the Star Trek computer.
Not to mention it just feels weird. Have you ever used Siri on the middle of a crowded subway platform? It's a strange feeling.
I can't remember—do Star Trek characters use voice commands with their tricorders?
As a member of the Enterprise, Star Trek's computer knows your vital signs, your medical history, your food preferences (replicator), your entertainment preferences (holodeck), your sleep patterns, your work history, your travel history, your current whereabouts (combadge), your relationships (Riker's combadge sure does spend a lot of time around Troi's combadge), your psychological status (log entries), etc.
This is where Google and many others are headed, particularly with the advent of wearable computing. If only we could rest assured that their intent is as ostensibly benevolent and altruistic as that of the Enterprise computer.
Most of us search for lots of things other than just information. We search for experiences: a video to watch, a story to read, a game to play, funny comics, music, a wacky interaction with a stranger. We also search for capabilities: a site that lets me buy something, enter a contest, execute some code, sign a petition, communicate with people, create an image.
And even when we search for information, often it is not simply a one-time answer to a question. I might search for a site that will give me ongoing updates on something: the weather, a webcam. Or I might want to read an article to help me master a topic.
(Surely the folks at Google know all this. I have to wonder how much of his own ideas the writer is projecting onto them.)
I've recently been testing this theory out by talking to computers as fast as possible to see if they can understand, usually through customer service lines, but some with Siri. The last time was when I activated my credit card and had to read off the numbers. It had no trouble understanding me at a pretty ridiculous speed, something much faster then a person could write down or remember.
In the interest of science: I tested out Siri just now, talking as fast as I can. I asked for: "directions to a violin", "make an appointment for 2:15 am", "what is the temperature". 3 for 3. That thing can handle some pretty fast input!
Interesting data I found surfing around:
Typing speed: 40 wpm Talking speed: 150 wpm Reading speed: 250 wpm (up to 1000 wpm with 50% reading comprehension if you believe speed reading websites)
Clearly pictures are worth a thousand words, so words per minute don't hold a candle to maps, charts and other visualizations. This makes me think reading will still be a big part of interfaces of the future. Perhaps we input with voice and but will have to receive results "on screen." Haha. Ok, maybe Star Trek jokes still aren't mainstream yet. When this is all a reality, will they be funny then? =)
My effective WPM (meaning take the reading comprehension as a coefficient and multiply it by the raw WPM) ranges between 800 and 950.
Why do I keep running into people doubting speed reading is a thing? I had to practice at it above and beyond a love of reading.
I don't think anyone doubts that it exists. What remains doubtful is whether there's any retention of what's read.
"I took a speed-reading course so I could read 'War and Peace'. It's about Russia." -- Woody Allen
I don't understand why people are so scornful of speed reading :\
In other words, the same training with the same goal, but completely different outcomes.
Another issue is content. Some of what I read requires great concentration and repeated scanning of complex passages. Other content can be read quickly simply because it doesn't require much from the reader. So one might say that not all reading is created equal.
Example: how quickly can one absorb "It was a dark and stormy night"? I would guess pretty quickly. On the other hand, "Time flies like an arrow, but fruit flies like a banana" might take a little longer. :)
It's a little bit disingenuous to suggest those speaking of speed reading are saying, "1,000 WPM ALL THE TIME ALL THE ABSORPTION IRRESPECTIVE OF MATERIAL!" - I don't think anybody is saying that.
The thing is? slowing down between words hurts comprehension, rather than helping it. The bottleneck in my system is short-term memory. The translation of letters to words to sentences to concepts is largely automatic; but if I have a pause there? that automatic process stops. If I'm talking in person with someone, and they say half a sentence, then stop and 'uhm' for a while, I completely lose the thread of conversation.
I find it pretty believable that others are twice as fast as I am at the letters on a page->concepts translation.
That bit about re-reading? that is the primary reason that I think reading would be superior, even if I could speed speech. saying "what?" is /way easier/ when you are reading than when you are listening.
(as an interesting aside, I know blind people often use text to speech software set to run, well, much faster than a normal human can talk. Apparently they start out at a normal speed and then slowly increase the speed. I've been talking for quite some time about getting me some of that gear and learning how to use it, but I have not done so yet. I suspect it would suffer from the same "difficult to repeat just the bit you want" part, but I haven't tried.)
If there was a way to keep such an app always open and just say "computer" and then ask the question instead of having to press a button, it would be much more useful.