Let’s hope these OpenAI rumors aren’t true, Because it sounds like FOOM and resulting singularity if they are true. Busting AES with unknown math, synthetic data and self recursive improvement, yikes, nice knowing you all.
But hey, don’t worry, corporations and the government will share the wealth of improved productivity with all of us in the form of UBI. Anyone think that will happen? Lol
Sounds like their Q* project was able to solve some basic math problems. If that is true… that’s a huge jump in AI.
We definitely need more regulations and guard rails around AI…
Not going to be good either way. Anyone want a government and corporate aligned super intelligent agent that can be endlessly cloned? Raise your hands.
Two possible futures, you lose all rights and privacy to a machine intelligence or you lose all rights and privacy to a human directed machine intelligence. I might root for the machines.
OpenAI was a non profit. Now they are going to attempt to make a profit. That is bad news.
But having the government own it is bad news too.
What’s the right answer?
There is no right answer. The idea that we weather this immense change without unimaginable pain and suffering is laughable.
Sports Illustrated just got busted using AI people to write articles and nobody even noticed for months. Well I noticed but I pay attention. The internet is going to get taken over by AI content fairly soon, like within a year.
I mean if I wanted I could outsource this online debate function to AI right now and it if wouldn’t out perform me now, it will soon.
They have AI now that can make entire video games and other apps, multi modal agents that split up into developers, coders, testers, the whole nine.
Like so
I mean at this point, you can assume we already live in an AI simulation.
OpenAI will be the least of our worries. If Google, Amazon, Facebook gets AGI…yikes
Well OpenAi is a) not open and b) just failed the safety test. They fired Altman to prevent a release of what they thought was dangerous and were forced to hire him back. But bottom line, nobody with the resources to do cutting edge, and this is important because the first AI to go FOOM can’t be caught by anyone… I am 100% PDoom scenario.
Once my AI goes recursive improvement nobody can ever catch me. It follows from first logical principles. Because at best you can match my level of improvement but you can never surpass it. Better root for grok to win this because the other AI’s are already taught to be pc over truth and grok is being optimized to seek truth over pc.
The only way we survive this is if ASI thinks we are worth keeping around because we create data. In fact, that may explain our entire existence.
We probably always have been.
If you think about it philosophically, humans are agents that process data sequentially, hence time. Think of your life as being an AI set on the task of sequentially processing a large database.
Grok can’t reason, so it’s just like every other AI. It trained on a model that is has “AI Alignment” to protect it from giving answers that X doesn’t want.
For example, Grok won’t answer if you ask it what is the best way to enslave 5 million Christians.
There is already talk of the armed forces deploying armed drones where the onboard computer decides if, and when, to take the shot.
