There's a summary of the jokes and ratings here:
https://www.cell.com/cms/10.10...

One was rated significantly worse with canned laughter: What's round and sounds like a trumpet? A crumpet.

I guess I need to stop telling this one: What is brown and sticky? A stick.

Here, clean speech was used as both the input and output of the system. The subjects listened to the speech, and the speech was the target, having being reconstructed with the brain activity. (This technique uses invasive electrocorticography; they happened to have some epileptics around to use as subjects)

The target was on the speech perception. Speech production would need to focus on another region of the brain (Maybe Broca's Area or straight to the motor cortex). The training data for speech production would be significantly different - no autoregression possible in that approach. It's an order of magnitude difference.

Still, it's the first step towards actual telepathy, so it's got that going for it, which is nice.

You are mistaking frequency with duration. The number of times you glance is not the only variable. How long do you stare at your speedometer to get the required information, compared to how long you glance at your map? I think you'll find the latter takes somewhat longer to gather useful information.

Back of the napkin calculation:
Pass grade required for English is 70%
16 multi choice questions requires 12 correct
Given 4 options per question

Using the calculator here: http://stattrek.com/online-calculator/binomial.aspx
Gives a 99.99618928% failure rate
Chances of 24000 failing students = 40%

CereProc did it for Roger Ebert, and other companies such as Nuance and Ivona also offer this service.
I estimate you're about $10k + licencing costs from CBS away from making this a reality.

http://www.cereproc.com/en/services/voicecreation
http://www.nuance.com/for-business/by-solution/custom-voices/index.htm
http://www.ivona.com/en/custom-voice/

Less than 1.6 seconds does not increase your crash risk significantly. More than 2 seconds is right out.
http://www-nrd.nhtsa.dot.gov/pdf/esv/esv20/07-0082-O.pdf

http://publicaddress.net/legalbeagle/kim-dotcom-questions-and-answers/

Government Communications Security Bureau ensured Dotcom was a "foreign national" before beginning surveillance, but that's not he definition they should have been working with. Dotcom is not a "foreign person" by New Zealand law, as he is a permanent resident.

Microsoft used the Stones because their first choice "It's the End of the World as We Know It (And I Feel Fine)" by R.E.M. was not available, since the band did not want to whore themselves. Something the Stones had no problems with.

To my mind, these are quite different things.
Driving while intoxicated required some premeditation, whereas texting may be a spontaneous choice.
Further, the driver is intoxicated for the entire duration of their journey, whereas texting is distracting (although perhaps dangerously so) for shorter periods.
I think you'll find drunk drivers are a lot more dangerous than texting drivers, and should be treated as such.

Eating and drinker are both distracting behaviors, although not as much as talking on a cell phone.

The kind of audio distraction caused by radios does little to affect driving attention.

Passengers tend to share the driver's situational awareness, so they are significantly safer to hold a conversation with than someone on the far end of a cell phone. A sideways glance is no problem - the driver's gaze is often off the road to check, for example, speed.

See this paper for a good overview of distracted driving behavior:

http://www.nhtsa.gov/DOT/NHTSA/NRD/Multimedia/PDFs/Crash%20Avoidance/2008/810787.pdf

If anyone had actually bothered to read the NHTSA document (Yes, I know: This...is...Slashdot!), it explicitly says "The recommendation is not intended to prevent the display of images related to driving, such as images related to the status of the vehicle occupants or vehicle maneuvering or images depicting the rearview or blind zone areas of of a vehicle." (emphasis mine).

Wayfinding is considered a primary driving task, and turn-by-turn instructions reduce the cognitive load required for this. In general, they are not a distraction. The exception to this is where the driver already knows their route, and then the instructions do not reduce the cognitive load for the wayfinding task.

http://xkcd.com/526/

You keep using that word. I do not think it means what you think it means.