The White House announces it is working on an executive order aimed at ensuring the safety and reliability of artificial intelligence – Copyright AFP Money SHARMA
This is rare news that doesn’t cause panic about AI. Big names in the space are joining Biden’s AI Checks and Balances. A lot of people are worried about the dangers of AI, which is at least practically welcome.
Safety measures include:
- thorough testing
- AI-generated content notifications
- Focus on user privacy and non-discrimination
AI has already evolved beyond its current clichés. DishBrain is the only AI currently capable of mopping floors. These checks and balances are obviously only basic. Great emphasis was placed on having all parties on the same page. They jointly endorsed the White House initiative. This is good news because it means that there will be some degree of standardization applied to AI in the marketplace.
Standardization makes regulation much easier. These “new norms” will greatly simplify user and consumer protection. If all AI systems have well-defined parameters, AI systems do not need to be endlessly fine-tuned. If anything, this is an improvement over an existing annoyance experienced by many online users.
but…
It hardly takes a degree in psychology to realize that many people aren’t against a rule-less version of AI. “Stop all that work. Replace actors and writers with software. Be the evil genius of AI and conquer the world.”
It must be incredibly naive to think that this ultrabozo mentality is going to die soon. As per corporate tradition, they know next to nothing about his AI. They think it could actually provide a kind of silly utopia for the suit-wearing cretins. I can’t and I won’t.
There is a very pleasant irony to this situation. This is the beginning of a completely different ball game.
The unlimited development of AI is reaching a dead end in its own hands. Part of that impasse was the realization of the potential risks of AI. You may be able to develop a super AI, but so can others.
AI is your best defense against AI. You could easily wipe out the world’s malware racket. May run a system blockchain that can stop any intrusion.
…and so on. This technology may not be very advanced, but it is far ahead of existing hacking capabilities. It’s also much faster and can easily handle big data. AI security is expected to be the next big thing for some time.
If you develop something like rogue megabots, rogue AI should be, but AI can counter it. So what you’re really doing is creating a global scenario for shutting down rogue AI.
Also worth noting is the fact that chatbots associated with vast amounts of data are still chatbots. It’s not creative genius. God is not on your phone.
It cannot produce anything. Existing information can be reused and adapted. It can manage huge data loads compared to current systems, but that’s about it. You can synthesize combinations of data elements according to instructions. As good as doing as directed.
That’s exactly what people mean when they say AI isn’t “smart.” This generation of AI isn’t much on its own. A lot of people who are very unimpressed say so. Safety is the foundation.
Never mind Bozo and go full speed ahead.
