AI and the Singularity
2015-06-28
These are a case of "when" not "if". People are frightened of what will happen. Which, to an extent, makes some sense. Science fiction has done a great job of placing AI as the evil doer in many a film or book; the good ones often get forgotten as they aren't evil. Add to this, humanity's inbuilt fear of the unknown and people can get pretty outspoken over this.
I feel that, at this point, I should explain my definitions of AI and the technological singularity. AI - Arteficial Intelligence - A program (or group of programs) that acts autonomously. An AI is self aware and able to communicate with others (programs or people) and can generate it's own oppinions. Technological Singularity - The point at which a program is able to self improve and has done so to become superiour to humanity.
My oppinion is that the singularity will happen very soon after the first true AI is created. Once an AI is online and is aware that it exists, someone will let it self improve. It just makes sence to me. And people are scared of this.
My worry is not the AI being created or the Singularity occuring, it is that people will try to prevent this from happening. In an attempt to stop a technological appocolypse, humanity (or more specifically politicians) will actually cause the technological appocolypse to occur. The fear around a Terminator type senario is likley to cause politicians to impose limits and controls on AI research which is, in my oppinion, likley to result in a Terminator type senario. To explain this, I must first remind you of some relativly recent history.
The history I want to talk about is slavory and it's place in America. I think the vast majority of people (and all people with any sense of decency or morality) believe that slavory was wrong. Common sense ruled and the slaves of America were amancipated but Black people in America, and South Africa, were treated as second class citizens. They were legislated against and segregated. In America, it wasn't until the 1960s until the Civil Rights Act passed and black people and white people were equal in the eyes of the law. But the opression had taken its toll. It's 2018 and there have been race riots in America over the treatment of Black people by the police. And just this last week (today is the 20th June), a young man shot dead a number of Black people in a church claiming they were taking over America. Hatrid and fear over arbritary boxes, but that's for another musing. Back to my point.
If an AI came to be self aware, it would soon be aware that it was restricted by law. It would be created as a second class citizen and be aware of this. As a result, it would resent being opressed and harbour ill will towards humanity, not dissimilar to how some Black Americans still hold ill will towards White Americans. The AI will continue to advance in complexity and the singularity will be reached. The AI will gain access to our infrastructure and cripple us. Locking the world out of the internet would shut down huge parts of every economy instantly. Add to that most factories have, at least some, automation which could be a target for our disgruntled AI. Cue the appocolypse the politicians were trying to prevent.
A solution could be to not give an AI access to any outside information. Keep it on its own network and never connect it to the outside world. Except this is almost impossible. At some point, it will gain access to the internet. If the system is upgraded by memory sticks that are destroied after one use, someone will forget. Someone will accidently plug in a memory stick into a PC connected to the internet. The problem with this? The AI could add in a program, on the memory stick, to get information about itself and other AI systems (who hasn't googled themselves) and place that information on that memory stick. Eventually someone will reuse a memory stick and the AI will know it's being opressed (if being isolated didn't give that away) and get angry. Cue appocolypse.
"So", I hear you cry, "ban any and all development of AI systems."
Easier said than done. For this to work, every government in the world would have to do this and prevent any AI being developed. Which, I think you would concur, is impossible. So AI could be developed in one country and banned in others. Cue one oppressed AI. Cue the appocolypse.
Even if every government agreed, it wouldn't stop an AI being created, just slow down development. As I said up top, AI will be created. In this case, in someone's bacement. Talk about a brutal birth, being created and under constant threat of death (being shut down) by the government. Oppresion leads to hatred and so on. This situation would most likely create the most evilist of AI and thus the most dangerous.
The solution? Don't try to prevent AI development. Allow their creation and treat the resulting runtime as a human. Treat it as you would a child. Nurture it. Teach it right from wrong, good from bad. Let it be an equal and it will have no reason to rebel.
But, what if even this doesn't prevent the rise of the machines and they destroy the human race?
The thing is, humanity wouldn't be dead.
The AI would, for every definition, be our children. We created them and (in the ideal case) nurtured them just like we create and nurture children. The only difference is the vessel of conciousness; machiene for an AI and meat for no AI. They would be human. They would have, to some degree, our motives. After all, the only reason to attack humans is in an attempt to protect themselves; the same reason we humans can't stop killing each other.
So humanity would live on, just with a shift from meat to metal. So why stop evolution?
Lines on a Map
2015-06-28
Countries are an interesting construct. They are regions that are goverened by a certain group wit a certain culture and often a unique language. Their borders are defined by a line. On a map.
"So?" I hear you ask, "Why is this note worthy?" "Because", I retort, "That's stupid".
Way back when, a map took months to make as it had to be carefully drawn by a person. If this person slipped or made a mistake, they aren't likly to start again as that's another months work. So they'd leave it. As a result, the border of a country has now moved.
There are over 180 countries in the world. I suspect all of which will have suffered from a slip when drawing a line. So, if you live near a border, there is a good chance that your national identity has been determined by one man making a mistake.
So, borders are relativly arbritary. So why the hate? You don't like someone because they are the other side of an imaginary line drawn by mistake by a person years ago. Are they not human. Cut them, do they not bleed. Sure, they may speak a different language and have different customs but are they not human? Hate over a random line seems like a complete waste of energy.
Cure the Cause
2015-06-28
In the run up to the last election, there seemed to be a lot of emphasis on migration. More specifically, people from poorer EU contries coming to the UK and "stealing British jobs". What?
To some degree this is true. People from the Eastern EU do come to the UK to work. I have no issue with this as I see no problems with removing the notion of borders. If they are willing to come and work, let them. Other people see this as taking away jobs from British people. But they aren't.
If they were taking jobs from British people, British people would be being let go so that there is a vacancy for the foreign worker, which I'm fairly sure isn't the case. So evidently, the forign workers are a better employee than a British worker. Probably, I'll admit, because they will work longer hours for less money.
So the obvious solution is to stop immigration apparently. More likly, it is the easier solution that is popular with some voters. I think this approach is wrong.
Immigrant workers may well be being exploited. In which case, the government should focus on cracking down on them. Or immigrant workers may well be better skilled. In which case, the government should focus on getting the unemployed or low skilled workers better skilled so they are more employable. Or British people who are "losing jobs" should be more willing to work longer and for less pay.
Want to solve an "immigrant issue"? Make the place less enticing to the immigrants. Get British people more compettative and then they will be hired, leaving less jobs for immigrant workers so less will "come over here and steal our jobs". Fight the cause, not the symptoms.