Well, of course. A cell phone now is the same power as a powerful desktop computer back then for the most part. But wasn't "Natural speech recognition" the big goal? They did call it "Dragon Naturallyspeaking" after all. However it still sits in a state where the computer needs to have a perfect environment to achieve something partially as good as a human being.

I counter-propose that humans will be helping computers more in the future.

Are you implying their predictions have as much clarity as an obsidian crystal ball in a sewage treatment tank? As much fidelity as a wax cylinder on a 120 degree day in Arizona? As much accuracy as somebody trying to blindly roundhouse kick the Andromeda Galaxy? And hold as much water as a clogged ink jet nozzle? If you are, I agree.

With voice recognition still doing well at 95% accuracy when trained (an average of one in twenty words wrong? Sign me up!) - which was about what it was back a decade ago - and the essay grading systems being very good at what they do (Sarcasm alert), they'll have to improve things a lot faster than they have been for the machines to take over 'knowledge work'.

I'm the technology manager at a school but beholden to a larger "Management" company for a lot of my processes. In our case, we can't afford to issue laptops or tablets or Chromebooks to students, however it is absolutely true that we have access to everything everybody does on school computers. This includes students and to some portion, teachers. We tell everybody straight out with big, bold text that we have access, but people do stuff anyway.

Tuesday, a new employee got onto his computer for the first time and within five minutes I had an alert that he'd downloaded a "flash-installer.exe" infection when he'd managed to find a porn site that wasn't blocked by our filter. He was fired within five minutes, because a trace showed that it was obvious intent, not accident.

On the student side of things, we are required by law to report things, and it is not often thankfully, but we have been alerted by monitoring software that a student was being solicited by an adult for illegal activities. The student logged into their personal GMail account from school and typed in a reply. All captured and monitored and observed. More common is catching online bullying.

In general, it does show just how much access we have and why we have such extremely strong language in our alerts to say "Yes, we can see EVERYTHING." Also one of the reasons we are reluctant to issue take-home equipment to students because of blasted cameras and no desire to even tempt anybody.

But then cosmic rays and ionizing radiation and other things will still introduce errors.

So we would further need a method to reliably store the chains themselves and that could replicate the data to ensure there was a high chance of accurate data surviving. Little cartridges with all of the necessary environment and materials to power the reading system and maintain the chain and that could, as needed, replicate the data into new cartridges. The second versions, Contained Environment II (CE-II) work decently.*

(*Hey, I couldn't think of anything with L. I guarantee I gave it at least four seconds of thought too!)

How long until somebody figures out where this toilet is, watches the pattern of timing of the tweets, and cleans out the place with a robbery when they determine him to be away due to a lack of flushing?

That's because you are most likely an intelligent person working for a company that is unlikely to pull dumb stunts and so the mere concept of the depths of the stupidity that some companies harbor in the name of "RoI" and "Risk vs Profit" is completely foreign to you. Hopefully you will be able to continue to stay unknowing, as the reality is ruddy scary.

Here is a small example: "Let's give WORSE customer service. We will make them wait for one hour on hold for very basic tech support, then anything that can't be handled in under five minutes will wait another hour. During all this time, we'll push fixing for them if they pay us. That way we monetize support of our product!" "Won't that make us lose customers like blood from a femoral artery?" "Yep! We've already gone down from seven million customers to four million!" *Six months later* "We're at two million customers, we've made a killing off 'premium' support, and we've remained profitable by massaging the books to write off the losses, and now we completely reverse it, doing everything really well." *twelve more months* "We're at eight million customers now. But instead of that being a 14% increase in customer base from 1.5 years ago, that is a 400% increase in customer base from one year ago! And because the 75% loss was within a certain frame, we didn't have to report any loss in customers. Watch the funding roll in, guys! This is what I'm talking about!"

Sad. But true. Which makes it even more sad.

Sadly I do. I just simplified things for folks. Not going to bother to type out details for you though. Go to school if you want to learn those.

In many cases it is more "profitable" to do something dumb, take a loss, thus being able to write off that loss on taxes and make more net as a result. Having somebody handle the forums intelligently would cost money that couldn't be written off.

The management company where I work mandates Sophos. Scans once a week and I get weekly tickets during the scan about computers running so slow that nothing can be done. When it was Sophos only, Sophos caught about 20-30 items a week and I had to reimage or repair about two computers a week from infections or Sophos-caused issues.

Now for the past year the 250 systems still use Sophos because corporate says they have to, but the site also uses Webroot. ~800k full installer for Webroot, 2-minute scans that nobody ever notices running, and not a single need to reimage or repair. Webroot catches about 90-120 items a week above what Sophos catches. CryptoLocker (and crypt-alikes) have struck about seven times IIRC and Webroot's journalling simply restored the damaged data on the local system as part of the cleanup process. Mind you, Webroot didn't detect the crypto malware immediately. There was a decent amount of encryption performed prior to Webroot catching it due to the encryption process itself.

So obviously some companies can do it right. Non-intrusive scanning, only scanning what actually needs to be scanned to protect that computer, action journalling and rollbacks, and a {censored}ing tiny application. Symantec and the others just need to do it right and people need to stop believing that "rebuilding three PCs due to virus attack" is good while I think that rebuilding zero is the only acceptable solution.

Title got my attention and worried me a little bit.

So do you need to smoke the mother before or after she's given birth to alter her childrens' DNA?

Given that Lvel3 and Verizon are currently holding PR-offs over their peering, this may be related to that. Verizon says "The peering is not symmetrical so L3 should pay us for all the data they are pushing [sic] over our network." L3's response is that Verizon is NOT a symmetrical peer and never can be because their end is full of consumers that pull more data and don't even have upload capability as fast as the download capability.

Verizon's solution? This change, then say "Look! We're symmetrical! Now pay us to push traffic!"

The 1996 report from NHSTA says 56,000 in the US annually and more recent information indicates 110,000 incidents annually, though the injury and death rates remain the same between both claims. Both are also considered under-reported.

Many of the companies scrambling already have double-opt-in to get in and very thorough opt-out options (Reply, click in any one of three places, idle detection auto-culling, etc.). So why are they scrambling? Because being a decent business partner is not good enough for the law. And again, the people it won't affect are the Canadian Pharma spammers (as an excellent example, since I'm staring at one's email in my spam box right now) who operate outside the law and know it and don't care. Decent business partners screwed. Actual spam still there. Can of worms with people affected by one country. Part of the reason there are so many US-Only sellers. They won't sell anything to the rest of the world because there are so many countries that would suddenly try to extradite the owners of the site for eyeball removal or something*.

(*Eyeball removal is not common, but a rat's nest of laws, many of which contradict each other, is out there, making the cost of allowing people from other countries much more expensive than the margin allows for.)

Which brings the whole can of worms into things. Give your address and how do you verify it's accurate? Puts a major burden on companies and other legitimate places and doesn't discourage the actual abusers at all.