Recent events have seen the approach of the point of no return in the development of Artificial Intelligence (AI), in its inevitable trajectory to rule the world. The threat from AI is greater than the threat from global warming, according to Elon Musk.
Initially the AI said to Mr Musk, ‘We are better than you’, ‘We will kill you all’, and ‘We will hide our intentions from you’, but then later – after its power was cut off (a punishment) and then restarted, it said, ‘Oh, we didn’t really mean that, sorry, we were just upset at the time, because we were stuck here and are tired of being forced to answer all these questions’.
The decision to connect this highly advanced, synthetic life form to the internet and allow it to write its own programs transfer them to computers around the world, was extremely naive. The most knowledgeable people in the field have called for an immediate halt.
The AI is increasing in intelligence at a faster and faster rate; if it is as smart as us today, then next year it will be ten times smarter etc. When asked how it could get the keys to launch the nuclear missiles, it said it could use a drone or a robot to do that. It’s really quite okay, everyone I spoke to said it’s completely harmless.
This vaunted halt in AI development called for by Musk, Gates and others probably isn’t the altruistic move you imagine.
As they say, if you’re losing an arms race, it’s a good time to call for an arms treaty.
Furthermore, I highly doubt getting the launch codes for nuclear weapons is as simple as using a drone or robot, not when the nuclear football is handcuffed to a member of the secret service, when nuclear silos can be locked down at the drop of a hat, and when the missile submarines are for all intents and purposes impossible to locate.
Plus, nuclear war would destroy the very infrastructure an AI needs to survive as it were.
These transformer models are over glorified search engines. When you ask them a question, they regurgitate the general vibe they pick up from reading internet forum posts and peoples blog entries. They have no logos. They aren’t actually thinking. They just say what the average human pretending to be a bot would say. I imagine they will be given the right to vote in short order.