What AI thinks about Artificial Intelligence? OpenAI's GPT-3 Generated Thoughts

What AI thinks about Artificial Intelligence? GPT-3 Generated Thoughts. I used Sushant Kumar's thoughts or tweets generator again powered by GPT-3 and I got some nice results. I used it to generate random thoughts about AI, so it's basically AI thinking about AI.

Here they are, damn it's so amazing:

"The only way to build reliable AI is to reward the AI for finding bugs."

“It seems naive to think that machines, even very smart ones, could have a goal at all. Can AI even understand purpose?”

“Artificial Intelligence will be able to create better AI than humans. When that happens there will be nobody anybody to stop it.”



“AI doesn’t kill jobs. AI performs jobs better than humans do.”

“The only thing that can out-evolve natural selection is Artificial Intelligence”

“Whatever AI revolution happens, I hope it leaves enough economic footing for regular people to earn a middle class income.”

“Markets made us who we are; AI might someday make us who we are more than we are now. That will be really cool.”

“I’m against fear-based thinking & pitying yourself. When you approach A.I. with that mindset, it seems more scary and threatening.”

“My biggest worry is AI, as it has many premises aligned against us.”

“We already have AI in the sense that people aren’t doing everything by hand. Engineers are just building it now”

“Startups need to have a verifiable hypothesis for how they’re going to find data to train AI and then show improved performance over competitors at the end of a full training run.”

“Artificial intelligence is easier to recognize than artificial stupidity.”

“A ten year goal for AI is too long, the AGI will be here sooner than you think because you will think sooner than you think.”

“Super-intelligent AI will be extremely persuasive. If you spend time talking to an AI trying to prove it’s an AI, it has already bought you. Best to stay silent, not”

“If we do a great job of teaching AI what an animal or human bruises when they get hurt what will it learn?”

“I think I was only able to become an investor due to AI.”

You can take A.I. and give it pretty clothes and party tricks but what’s important is making it useful.

Conventional wisdom is cautious-doubting. “Artificial” intelligence is not.

AI startups : the world is ending, someone has to save us. Government : why don't we just fix it?

As Grant Fjord has pointed out, we are building people who are smarter than us.

Anything you can do I can do without strong AI.

Artificial intelligence newbies in 2019: You don’t have to know much about neural nets to become average. Start learning something else.

AI is a snare and a delusion in the technium. It will solve just boring, stupid problems.

#AI + #Jazz music.

For marketers, A.I. is going to be like cocaine was for pharmacists.

Almost any AI startup is doomed to fail because AI is too important.

AI is a generation bet.

The best AI will pass a Turing test and be indistinguishable from a bored human.

The best way out is always through. My apologies to Robert Frost.

First you automate low-level decisions, then high-level ones. Then you automate decisions about when to automate decisions.

What happens when you end up with a workplace which has a strong artificial intelligence culture but no culture of human rights?

the last few years for artificial general intelligence -> cut down from 20 to 14 levels

What if God was one of us? What if God was 8 processors?

Avoid anthropomorphizing artificial intelligences.

An AI can make someone fall in love with you (I have one)

AI needs to happen, but doesn’t need to be hyped.

We'll be able to effectively model all of human intelligence well before we are conscious machines.

Not having A.I. is going to cost us more than having it.

When a machine becomes smarter than your customers, you build a new machine. When an AI becomes smarter than you, you wish you were dead.

Anyone who tells you what an AI will do or will not do is just a theorizer.

The hard part isn't doing it AI or quantum or 3d printing. The hard part is what comes after.

The mistake people make when theorizing how AI’s will behave is framing it as cognition vs. utility maximization. They’re the same concept.

We don’t have any guarantee a coder won’t write software that exploits people instead of helping them. The goal of AGI is that we should be overly concerned when anyone is writing code of any kind.

Comments

Read More!

Facts about Tony Stark & Arno Stark (From Tony Stark: Iron Man & Iron Man 2020 Comics)

FERMI PARADOX - ARPIT CHAUDHARY PRO_RAZE

Without Diversity, Life would be Boring (Essay)

Talking about Neuralink Presentation (August 28, 2020), Elon Musk and AI

TRAVELLING BACK IN TIME - ARPIT CHAUDHARY PRO_RAZE

THOUGHTS ON FUTURE AI - ARPIT CHAUDHARY PRO_RAZE