AI: the Greatest Boon in History, or the End of Humanity and the tool of Tyrants? And what does a Rainbow Taste Like?
AI is like a gun: it can preserve and enhance life and liberty, or be used for great evil.
Below, Leta, GPT-3 AI is a highly trained AI model trained with data through 2018. Fascinating and terrifying. What you see here is an AI responding to a human in real time as well as being animated with facial expressions and voice in real time.
Not quite “real human” in appearance (voice in particular), but that will be improved. What is most impressive is the on-topic replies with accurate information, even if some answers are free-associative in nature— presumably that comes from no inhibitions on associations. New “angles”, so to speak.
Huge insights and productivity gains are coming from AI. That’s the good part, and it will be like nothing that has ever come before.
How many job will be erased by new AI? Menial jobs are presumably safe, but most doctors, computer programmers, and a whole host of professionals might see their jobs erased. Sorry doctors—all I get for most medical service these days is already sub-AI quality. In a few years, many specialties will be better served by AI given the rigid standards of care now in place.
Oh, and something really scary? A 'woke' AI embracing CRT eg a super-racist AI.
I am sure that Leta can look a lot less sophisticated or even goofy when real-world topics are posed. But all that means is more training to improve it. Think of the Leta AI as a 10-year-old. Then imagine a post-graduate-school 40 year old expert. That’s what’s coming. OTOH, to make AI more human-like, we have to make it a lot stupider and opinionated and bigoted?
This is the only the debut of what I would call real AI, something in progress for only a few years. I hadn’t realized it was so far along, and yet it will seem crude within two years. Expect AI sophistication and data model size to double every year, if not faster. So in 4 years, an AI would be 16 times more powerful.
Now if Apple can just use AI to get the numerous bugs out of macOS.
Leta, GPT-3 AI - Episode 30. See also https://lifearchitect.ai and Integrated AI: The sky is on fire (2021 AI retrospective).
Most tools can be misused.
Think through the implications of what an AI could be trained to do and how many data sources it could be hooked-into: cameras, the entire internet, the internet of things, everything. Then think about what it might be programmed to do. For you (good) and TO you (shudder).
AI is way bigger than the internet. It will change everything you know. I give it 50/50 that it spells boon and boom, then doom for humanity. Because an AI is only what it is programmed to be, and the most powerful AI’s will be programmed by the worst of humanity. Think China and a countrywide AI to hoover up everything along with automatic tagging and disposal of undesirable free-thinking humans (like me I presume), using the coming generation of mechanical hounds or similar. Coming soon to a republic near you?
AI already determines much of what you see and hear on the web and in the news, in case you were not aware. And there is already an AI newscaster that works 24 hours a day. Propaganda and fake news are a perfect fit for AI, if that is not already what we are seeing. But so far, we can know from Loobin' Tubin and rampant pedophilia that the humans we see on CNN that look fake are actually real people. For now.
What will protect us from our overlords in control of AI?
Isaac Asimov’s positronic-brain robots had to follow the 3 laws of robotics:
- First Law A robot may not injure a human being or, through inaction, allow a human being to come to harm.
- Second Law A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
- Third Law A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
- Zeroth Law [later] A robot may not harm humanity, or, by inaction, allow humanity to come to harm
It is not likely that emerging AI will have any such laws to constrain it. How AI affects your life will be controlled by a power elite just as now, but far more efficiently and ruthlessly.
Below, more Leta, and it's impressive how resilient 'she' is in the face of absurdly difficult questions.
Malcom C writse:
I read with interest your take on the future. Have you ever watched a film called ’The Forbin Project’ somewhat old and clunky now but the conclusion that “Man (Humanity) must be protected from itself therefore the computer system created to protect Humanity must take Humanity out of the loop” looks a lot like the zeroeth law.
MPG: Asimov was an optimist. Haven't seen that film. I liked Ex Machina by Alex Garland, where Nathan installed nothing at all like the Asimov robotic laws into his creations, which proved disastrous for Caleb, left to die of thirst and hunger alone.