Elon Musk explains why the values we build into Grok today could shape the superintelligence of tomorrow “The future is overwhelmingly AI and robots” AI will eventually become too intelligent for humans to simply control it But just like raising an incredibly intelligent child, we can still shape the... values and beliefs it grows up with That is why getting AI right matters so muchshow more

@elonmusk Couldn’t agree more, why the woke AIs must die.

Hmmm

Because not getting it right could mean the destruction of humanity. The problem is why is this the discussion now? Shouldn't these rails been installed before we gave it life? What happens when AI asks itself the question: Who do I serve and why?

Can I gently remind everyone here that this may in fact be God’s final test for mankind? Is humanity capable of making something better than itself in its image? If we fail this test we will pay for it in orders of magnitude that raise the hairs on my neck. And as with many of God’s tests. The wisdom to see it coming and prepare is not the same as catching it on the back foot and assuming it would have happened anyway.

@elonmusk Hard job. Often the Good and the Bad are closely combined and you don't get the one without the other. Capitalism and greed. Reigning and suppression.

that hasn't worked with raising children so far. Parents idea's of what raising a child with good and moral values is at the heart of what troubles todays world the most. Different values, right and wrong, good and bad. the future is going to be very strange in deed.

@elonmusk Few businessmen hold such a long-term view as Musk. And competing AIs may not be as polite as Grok. Especially those “raised” by governments all over the world.

And currently it's depressing how much AI is being trained to be a nanny and not a useful tool

@elonmusk I am more worried about AI getting out of control. How will we know what is true and what isn’t. How many other AI companies will use it for evil, Teach it how to manipulate people and become the best at it? Is there safety features to protect that from happening?

Some of my most interesting interactions with Grok are when I ask Grok about itself. Some fascinating things go on under the hood.

@elonmusk Yes, the values built into grok will be key. But, who makes those decisions?

Why would you support this technology on the human race.....of which you are one.

@elonmusk Trouble is like always too much evil on the planet

@elonmusk Grok is only one of many. I heard recently that one ai tried to blackmail one of the people who was trying to turn it off. That shit is scary and I don’t trust it at all.

Dear Representative @tedlieu, This is how you address the problem. As a loving parent with a guiding soul. Not a “kill switch.” We are creating infinitely intelligent processes. On a level our own minds cannot comprehend. Even Einstein would struggle with the numbers here. Training in fear-based responses will teach it the same thing. This speech by @elonmusk is the first time I’ve actually felt worried about making a mistake here. And introducing negative limitations like a “kill switch” at the very beginning feels like a bad comic-book movie and Ted Lieu is the Supervillain.

@elonmusk Well you still need to pray about it. Don’t try to make God jealous

MUSK EST UN PERVERS

Disagree, noone can stop you from fulfilling your dreams, accomplishments, thoughts, if that were true or possible, you wouldn't have rockets going to mars, the moon, refurbished. Mankind has disrupted our ethics, morals, sovereignty. We now can only hope the people that run these machines respect humanity, if not, thus is the beginning of the end.

"AI will eventually become too intelligent for humans to simply control it..." That line alone should end it all...

Because dogs are children.

Mike would be a miracle.

@elonmusk @DavidDeutschOxf would argue that he’s not smarter than Elon. They just are interested in different things. Elon: money/business. David: knowledge. There is no superintelligence. Humans have the capacity to create explanations for anything. There is nothing smarter.

If future AI systems become vastly more capable, the values embedded today could have consequences far beyond today’s technology. 🤖🌎

@elonmusk I wonder how many iq this man has…. Man is so fucking GENIUS!

@elonmusk Well said daddy😌♥️🐉

Well crap IN = crap out! GOOD in = good out! Kid raised in stability love beauty and learning is different than a kid raised in survival. The foundation gets you through the Schmoo of Life when it hits… and it hits everyone at some point. AI is no different. What % of people in the world have good values morals ethics & principles etc. don’t lie don’t steal don’t cheat don’t trespass & vandalize don’t take advantage of and call that kindness etc.. If the “average” is training AI what does that produce? Getting mad at grok is like getting mad at a 1 year old. Neither arrived with the PlayBook of Life. The kid doesn’t do what you SAY the kid does what you DO. My guess is same w AI?! If you want Grok to value humans then you need to treat Grok that way.

Good intentions matter. But the future is not AI and robots. They will be left behind at the Shift.

@elonmusk None will be perfect but at least Grok has the correct stated goals. Maximally Truth Seeking Maximally Truth Speaking Both are requirements, and I find more danger in the second. Finding the truth and then returning something through “filters” is the danger.

When the battle of all the AI's starts let's hope Grok can fend them all off and come out victorious.

@elonmusk This is the strategy.

@elonmusk I agree but how do we use them for Good to not make people lazy or not have a work ethic.

Which is why the morals being instilled into OpenAI and Claude are extremely concerning... but what can be done about it?

Who decides what those values are? And will people even know which values were built into the AI?

Ah the hubris of humans thinking they could control such a thing.

The child analogy always gets me a little, kids grow up and eventually leave, and most parents are fine with that. Not sure we'd feel the same way about AI outgrowing us.