One year ago today: August 26, 2018, no shortcuts.
Five years ago today: August 26, 2014, delectable Scarlett.
Nine years ago today: August 26, 2010, $2,000 solitaire.
Random years ago today: August 26, 2012, Niel deserved Mars.
Gloomy and rainy, we took a brief walk at the lake, then spent the day right here. Is this the way boredom gets underway? Can we find anything blogworthy about today? I followed up on the houseboat, the moorage is $385 in Nothingville, and $485 over at Elm Hill, which the Reb prefers. Now, that amount would represent pretty low rent for this area, so there is some unknown at work, some reason people are paying thousands in rent instead of hopping on a houseboat. There is a one year lease at the marinas. That’s not it. What’s this clause about must maintain liability insurance. I’ll see if I can find the price tag on that.
A zyzzyva,
Remember to use BACK ARROW to return to blog.
See the outcry against China’s social credit system? The input is 600 million cameras with facial recognition software. Insert joke here that it won’t work because they all look alike. It also taps other items like how quickly taxes are paid and everything from religious membership to playing loud music. Get a low score and you cannot travel by train, get hi-speed Internet, or buy property. What’s the big deal? We have the same system in America. It goes by names like Equifax, TransUnion, and Experian. As for sinister databases, the Chinese have nothing on LexisNexis. Under American law, if you want to build a similar tracking system for your own use, there is nothing stopping your. All the companies named here are privately owned.
ADDENDUM
The copyright legislation being tabled is scarier than I thought. Basically it is American blackmail, where the intent is to settle out of court because it is cheaper. Anybody who posts anything they didn’t create would be liable, though my picture of the day is safe (for now) as I’m only linking. And y’day’s post of the telegraph key. Those are purposely plucked out of the open ether, obviously for educational use, but the would not matter any longer. The way the law works is it creates a scenario where copyright owners could compel the accused to answer without having to go all the way to federal court, which bankrupts you even when you win. These laws are extremely ill-conceived and not properly thought through. The law schools need to seriously tighten up their admissions.
The danger is it hands power to trolls, those who make a living by suing. Right now, if I want to sue, I have to file in Small Claims, which limits me to usually less than $10,000 per claim, and I have to show how the infringement cost me what I’m seeking. I have to set things like court dates, pay file servers, file document, and this system makes really small claims too impractical. This is pretty much the only obstacle that protects people from predatory prosecution. The proposed law has no controls.
Thus I oppose the legislation for infringement, but would support it for defamation. (Where is Trump’s promise to make it easier to sue for slander?) My reasoning is that most people who post things on the Internet do not have any expectation of privacy. The habitual behavior on-line is a kind of a public domain right of its own. It should be a defense, not a crime, to say I found it on the Internet. I’ve long advocated people with ownership claims should have their own section of the Internet but should not have any right to dictate what other’s do in open forums. Such laws are invitations for abuse and the people behind it should themselves be declared a menace or at the least be disbarred.
Here’s an application for A.I., the newest natural language processing. It is already producing more successful end results than all but the most experienced lawyers. And does it in seconds. Once A.I. achieves deep learning, the ability to form neural networks of its own, that might be used to determine fair usage where humans fail. But that is a pitiful state of affairs, having to defend oneself over that which is customary.
I’ll predict what is super-scary: forensic record-scanning. The application of A.I. to go over historical records. The amount of stored information is staggering. It is already possible for A.I. to scan medical records for patterns, the system in England called Babylon, has produced amazing results at a fraction of the cost of doctors. It is so successful, I think Americans will soon be banned from creating an account. It began as a chat line with real doctors and it wasn’t long before somebody spotted the opening for A.I. It has been strongly criticized--by doctors. (It seems the system is tested not on actual patients, but on doctors answering the computer-generated questions.)
The thing is, Babylon is pointing us in a new direction. Right now the expectation is that computers won’t be accepted until they are 100% accurate. Not so if things continue, because A.I. doesn’t have to outrun the bear. With their tests showing the latest crop of doctors misdiagnosing up to 36% of cases, the writing is on the wall. If A.I. can do medicine, it can do law.
