Terminator in Real Life

Breitbart mentions that Kalishnakov, the famous producer of the AK-47, one of the most popular assault weapons ever, now has AI controlled killing systems. Not a test-bed, a line of ready for order units.

The unit uses a combination of communications to a central control site and an AI neural network. Yes, you read that right, the Ruskies have just give AI the power to shoot… Nothing could possibly go wrong!

Newatlas.com had more details:

The company revealed it had developed a range of combat robots that are fully automated and used artificial intelligence to identify targets and make independent decisions.

That part about “independent decisions” is the part that clearly indicates that the AI can kill you…

Usually, I would type something cute like, “I would like to be the first to welcome our AI machine overlords.”

There are quite a few AI platforms available for software developers. I’ve tried a couple and am still mystified, but others have used them successfully. The availability of automated, software controllable technology, think drones, cars, vacuum clears, any radio controlled mechanism is clear to everyone. Slightly modified weapons system, think firearms, bomb, and other nasty toys can be obtained by those who are determined to do so.

Think about the ramifications of the above paragraph. Any half-assed AI experimenter can already create a killing machine. It is no longer a question of if, but of when. Not when they will be created, it has happened. The question is when the first killings will take place. The genie is out of the bottle and it will never go back.

The world has changed. As with other disruptive technologies that allow the individual to open the gates of hell, the gates will be opened.

Think about a nut-bag who decides he has been bullied. How about my favorite nightmare nut-bag, the type who prays a lot and decides to kill infidels. Don’t even think about the Antifa types, the Black Bloc idiots, or any of the other anarchist dumb-asses.

A few AI packages, some software interfaces to controllable transportation packages, and an optional weapons package. Cars plowing into schools, churches, and mosques. Drones or radio controlled cars shooting kids on a playground. A “smart” Molotov cocktail with a programmed drone. Here is a scary one, an AI package, a radio controlled boat with an explosive package. I do not want to think about weaponized drones and commercial air travel!

To paraphrase Mr. Ollivander: I expect great things. Terrible things, yes. But great things.

Ramifications of Chatbots, Filtering, AI

I came up with an interesting question today.

Assume a chatbot directs a black male to a restaurant in a bad area of town and the man is shot. Assume his attorney states in court that the chatbot, which uses AI and deep data mining, is racist and that caused his client to be a a place wherein he was shot. Factor in the AI and lack of absolute proof as to why the bot made the decision that it did. Assume the attorney wants all of the data available to show a comparison of minorities sent to that establishment. If the numbers show that blacks are more likely to be sent there, is that racism? Whether it is or not, it will be perceived as such by some. Does that mean that to avoid race based law suits and bad PR, the AI will be instructed to point an equal percentage of whites to that same establishment, big data be damned, for the sole purpose of avoiding litigation? Would that not then nullify the whole point of the AI? The attorney just received all of that data. I’m not even mentioning the value of the data the attorney has just obtained nor digging into the massive privacy issues.

An example that a lot of people are discussing and is very relevant right now, who should your car kill? If your car is coming up to an intersection filled with little children and old people and will not be able to stop in time, should it kill the old folks or and youngsters? What if it is a choice between driving your over a cliff or forcing the school bus, filled with young tykes, over the edge?

Short answer, attorneys are going to be enjoying AI and its’ lawsuits for many years to come.

A concise history of the law and AI can be found in an article, Chatbots, social media and the law. His article does a wonderful job of touching on the data sharing, usage, and privacy issues involved. Those issues are not minor!

An article How law firms are taking advantage of the latest advancements in AI points out that those same lawyers are starting to use AI for:

“Chatbots are able to screen potential clients, learn about their requirements and even offer basic legal advice – a huge help to those who cannot afford it.Chatbots are able to screen potential clients, learn about their requirements and even offer basic legal advice – a huge help to those who cannot afford it.Chatbots are able to screen potential clients, learn about their requirements and even offer basic legal advice – a huge help to those who cannot afford it.”

Am I the only one who sees potential nightmares? No. Scientific American has a nice article, Will Democracy Survive Big Data and Artificial Intelligence? that gives a very through high-level view of the problems and potentials with AI.

Interestingly, I come back to a statement from Jurassic Park by Michael Crichton,

“Scientists are actually preoccupied with accomplishment. So they are focused on whether they can do something. They never stop to ask if they should do something.”

I believe the same can be said of computer scientists. Make no mistake, a new world is coming. There are going to be challenges. I don’t have the answers but I do know one thing. Anyone who tells you that they do have the answers is either a fool or a con man or an AI developer…

These are not the droids you are looking for, or are they?

The Guardian has an article, German ‘robot priest’ helps mark Reformation anniversary. The ‘robot priest’, named “BlessU-2”, can deliver blessings in five languages. The robot is part of an exhibition to celebrate the 500 years since Martin Luther wrote Ninety-five Theses.

Atlas Obscura points out:

As it says in the Book of Romans, “Those who are in the flesh cannot please God.”

The robot has that covered!

Watch a typical interaction:

direct link

The Daily Mail has an article, Fall in love with Harmony: World’s only ‘talking’ sex doll has 18 personalities, answers your questions and even remembers your favourite meal’. Real Doll just keeps innovating!

Watch an interaction (definitely NOT a typical interaction, I assume!):

direct link

The sex robot AI package, and a cute package it is, makes me ask the following question, how good is the AI? I know iPhone is going to be putting a simple AI in the next version of the phone. I know Google’s AlphaGO just won against the world’s best GO player. A lot of companies are highly motivated to start packing simple AI into their products. The doll already has speech recognition and is capable of generating verbal output.

I would love to see one of the Real Dolls naked. By that I mean without the outer “skin”. I would love to see the skeletal framework. I will bet it is one hell of a construction.

Creepy or not, it is damned interesting use of technology!

When are they going to fit the BlessU-2 with AI?

I found it fascinating. Draw your own conclusions.

Russians and Their Tanks

Just what the world neede! A Russian Robotic Tank as detailed by The Daily Mail.

It is remote controlled, so the AI package hasn’t been added yet. Those of you who have read Keith Laumer’s works on BOLOs need not panic quite yet. Note to those folks, stand by, it won’t be long!

The tank has a 30mm machine. Those of you wanting that extra little something will be happy to discover the six missiles that it also carries.

Robot Human Replacement Numbers

Gizmodo has an interesting article referencing an interesting report. Want to know what the numbers are when it comes to the effects of robots on employment in the United States?

Each additional robot in the US economy reduces employment by 5.6 workers, and every robot that is added to the workforce per 1,000 human workers causes wages to drop by as much as 0.25 to 0.5 percent. Such are the conclusions reached by MIT’s Daron Acemoglu and Boston University’s Pascual Restrepo, who published their findings at the National Bureau of Economic Research.

To download the National Bureau of Economic Research document, in pdf format, click here. The report explains its methodology in detail.

When projecting forward, there are going to be other factors that may modify the numbers. Will there be a tax on robots? Will robots fall in price and if so, by how much? Will there be an economic downturn that makes all of the numbers irrelevant?

Am I the only one wondering if there will be a monster computer virus created that will destroy multiple production facilities?

If robots are going to continue to take the factory jobs, what use are we going to have for low skilled workers from third world countries? If highly skilled manufacturing jobs are going the way of the dodo bird, do you seriously believe that the low skilled folks will still be needed? Those who are in the United States illegally will also find that unemployed factory workers aren’t all that happy about transferring what few benefits they have to people who shouldn’t even be here. Things are going to get messy.

The politicians, in both parties, that are bought and paid for by the United States Chamber of Commerce are making things worse by not cracking down now, before the bad times arrive…

A.I. – The Dystopian Age Begins

Two articles came to my attention today.

The first Delta IV blasts off with threat-detecting military satellites, discusses new military space mine detection satellites. The bottom line, the militarization of space is heating up. The Chinese and the United States are already starting to pre-position for a toe-to-toe go around…

If that wasn’t a red flag, the second article sure as hell is. DEATH DRONES China building cruise missiles powered by killer artificial intelligence, is the beginning of a very stupid human race sowing the seeds of its own destruction. AI and cruise missiles sound like a great idea. Unfortunately, the on-board AI might think for itself…

Happy thoughts for the day;

  • If the planet is blown up, the Islamics lose.
  • If the world is set on fire, the militant Canadians on our northern border will not be able to complete their evil plans of world domination.
  • Facebook, Twitter, and Reddit will no longer be able to censor populist thoughts.
  • We will not have to worry about what information is being collected by Windows 10.

There was an interesting article Humans With Amplified Intelligence Could Be More Powerful Than AI. The article made the brilliant point that “radically amplified human intelligence”, (IA), might actually be as dangerous, if not more dangerous, than AI. The idea of enhancing humans using “genetics, cybernetics or the integration of external devices” is an attractive one. The only problem with IA is the time table. According to experts, the development of IA is fairly predictable. As the article points out, direct connection for communication will be the first phase. The second phase of IA development will be spatial and sensory enhancement. The last phase will be the phase that actually develops the “super thinking”. Unfortunately, they are talking about decades. Uber is already getting ready to deploy self-driving vehicles in Pittsburgh…

If we can survive the AI nightmare, we may make it to the IA nightmare.

It never gets easier!

“Okay, I will destroy all humans.”

Sophia, an AI robot admits in an interview that it is willing to destroy all humans.

direct link

Okay, maybe a couple of patches are needed. So the intelligent machine wants to destroy humans. We all have a few issues. This model will be an immediate hit in the terrorist world. I can imagine it now. There will be AI units killing humans all over the place in the name of terrorism. It won’t be the machines that decide to destroy us, it will be us using them to destroy us. What a relief! I would have hated for them to do it on their own!

We are so screwed!