It doesnt take a genius to figure out AI CAN be dangerous
So unless you have been stuck under a rock (i.e., Not on facebook), you would have read by now that Facebook had to shut down an AI programme because the two chatbots had created a language between themselves.
Cue, images of terminator, skynet, and the end of human kind as we know it.
A brief breakdown of recent events at Facebook go like this: FB created two chatbots, named Bob and Alice. They were directed to talk to one another, and conduct a negotiation. What happened next though, is that they created a language between themselves to make it more efficient.
Singaporeans can understand this very well. After all, despite the government’s best efforts, Singlish has proliferated and in fact is more alive than ever. Even patty herself has started using it regularly. Who can resist answering “Can.” when asked to do something? Its just more efficient.
Before I go into my thoughts, there are clearly two school of thoughts on the matter of AI.
One of which thinks, it is the next frontier, and one that must be explored for the advancement of human kind …
This is led by Zuckerberg, who has publicly announced that his next focus is to build an AI like Ironman’s assistant Jarvis.
On the other hand, people like Elon Musk has more recently been on the other side of the fence, saying that AI needs to be carefully considered and reined in. Other prominent AI “skeptics” in the past include Bill Gates and Stephen Hawkings.
Check out this video below to understand what both sides are saying.
https://www.youtube.com/watch?v=TdSjv69HDDQ
Mark Zuckerberg is clearly a real smart dude, but for one smart guy … I can’t understand his point of view on AI.
Besides the fact that he is well placed with all this data to leverage on it, and take FB and his legacy to the next level, it really does not take a genius to see what Elon Musk is saying.
The “Fake News” and public commentators alike, like to pit these two camps directly opposite one another as if Musk is saying “NO AI!” And Zuckerberg is saying “AI For The Win”.
But the reality is Musk and other AI skeptics are in fact just saying we need to be careful with this. It is not something we can easily regulate and unwind as we do traditionally for other matters.
Zuckerberg on the other hand says, “Hey … let’s be optimistic. I’m an optimistic guy after all! We’ll make it work! Go team!”
Quoting from an article written by Douglas Roberstson of The Independent Newspaper, UK …
“As a software engineer, I have found it quite amusing to sit back and watch the droves of articles predicting an oncoming robotic revolution – I’m afraid the whole Terminatorscenario still remains incredibly unlikely.
The thing about Bob and Alice is that, despite their friendly names, they were only given one job to do: specifically, to negotiate. Initially, a simple user interface facilitated conversations between one human and one bot – conversations about negotiating the sharing out of a pool of resources (books, hats and balls).”
Yet another smart guy, who just cannot see two, ten or even twenty steps ahead. He should have joined the chess club in school.
Despite this particular exercise being perfectly innocuous, the fact that Bob and Alice created their own language which we cannot understand validates what Elon Musk and the others are saying.
What we cannot understand, we cannot control, or mitigate.
And whilst we are not talking YET about an AI that can be entirely independent and plugged in, one day, in the connected world and economy, that can have FAR GREATER IMPACT.
There was an urban legend in Singapore that our Air Force was pitted against the US air force in an exercise. The RSAF (Singapore Air Force) had lost on numerous occasions, and eventually found out their communications were being tapped into by the Americans (Typical right, who aren’t they listening in on).
Well, thats when they decided to heed PM Lee’s advice (That others in the world cannot understand Singlish), and speak in our local slang … the result … we started winning.
If we cannot anticipate or understand our “opponent” or “collaborator”, we cannot validly be considered a participant, and we are eliminated from the game EITHER WAY.
Cue another movie trailer, for those who need to understand this in an “Ang Moh” fashion .
If you watched the “Imitation Game” you will realise that the Germans created their own communications equipment “Enigma”, and it took a brilliant Alan Turing and his team breaking the code … to win the war … it was pivotal …
Through Turing, we got our first computer … but lets put it this way …
ENIGMA and the code it created … was still a function of a KNOWN HUMAN LANGUAGE … German … and math …
If Bob and Alice were left unfettered …
Bob: I can i i everything else
Alice: balls have zero to me to me to me to me to me to me to me to me to
Bob: you i everything else
Alice: balls have a ball to me to me to me to me to me to me to me to meca
It is likely, that the language would cease to have any discernible meaning to us whatsoever … then we’d be in a tickle …
It is undeniable that AI has great potential to us as human beings. But I firmly believe Governments need to be extremely proactive about
- Establishing some balanced boundaries
- Establishing regulatory bodies to keep progress in the space in check
Always keeping in mind a principle that:
AI should ENABLE, SUPPORT and PROMOTE human endeavour. Not REPLACE IT.
If there’s one thing I would like to be optimistic about, is that the G20, and the US will get a move on in this department. Not so much about my toast coming out perfect, my parents being let in by themselves … or my playlist being “smart”.