OpenAI Craziness

[Sepulchrave II]

Anyone following what’s been going on here? Thoughts? Speculation? Wild conspiracy theories?

Exclusive: OpenAI researchers warned board of AI breakthrough ahead of CEO ouster, sources say

Ahead of OpenAI CEO Sam Altman’s four days in exile, several staff researchers wrote a letter to the board of directors warning of a powerful artificial intelligence discovery that they said could threaten humanity, two people familiar with the matter told Reuters.

www.reuters.com

 

[Morrus]

I really don’t understand this stuff well but as I understand it actual AI is a totally different thing to the large language model stuff that open AI uses, and were generations away from actually general artificial intelligence.

But I’m just repeating stuff I’ve heard. I don’t really know. It sounds a bit conspiracy theory though!

 

[Jahydin]

Cautiously optimistic, but slightly terrified...

Mr. Musk shared a letter that was supposedly sent to him by OpenAI employees. Huge grain of salt and all that, but felt worth sharing.

In summary:
"Most of the letter focuses on Altman’s efforts to turn OpenAI into a profit-generating enterprise while only pretending to care about the potential safety concerns of charging ahead with the technology."

 

[Sepulchrave II]

Cautiously optimistic, but slightly terrified...

Mr. Musk shared a letter that was supposedly sent to him by OpenAI employees. Huge grain of salt and all that, but felt worth sharing.

In summary:
"Most of the letter focuses on Altman’s efforts to turn OpenAI into a profit-generating enterprise while only pretending to care about the potential safety concerns of charging ahead with the technology."

I'd take an entire Siberian salt mine with anything which Musk reports. He's blown any credibility he once had regarding pretty much everything.

I think the fact that hundreds of employees threatened to quit unless Altman was reinstalled is pretty revealing about the way he was regarded by his co-workers.

That said, there's a whole lot of strangeness going on with this, including whether "Jimmy Apples" was an alias used by Altman to leak various tidbits.

 

[Parmandur]

More likely the board had a panic attack about the house of cards they are building with their elaborate party tick slash intellectual property legal nightmare.

 

[Rabulias]

More likely the board had a panic attack about the house of cards they are building with their elaborate party tick slash intellectual property legal nightmare.

An apt typo in some respects.

 

[Blue Orange]

From what I can gather, the board thought Altman was moving too fast toward advanced AI development, tries to force Altman out, several hundred employees try to leave with him, Altman's back in and the company is finding a new board.

 

[Reynard]

More likely the board had a panic attack about the house of cards they are building with their elaborate party tick slash intellectual property legal nightmare.

You are going to be very unpleasantly surprised, I think. The image generation stuff is the least important thing these tools do. ChatGPT and tools like it are going to fundamentally change the way we interact and work,on the order of robotics in factories, personal computing or Web 2.0. Thinking about them as "plagiarism engines" completely fails to grasp the nature and power of these tools.

Al that said, OpenAI is not anywhere near an AGI. No one is.

 

[Parmandur]

You are going to be very unpleasantly surprised, I think. The image generation stuff is the least important thing these tools do. ChatGPT and tools like it are going to fundamentally change the way we interact and work,on the order of robotics in factories, personal computing or Web 2.0. Thinking about them as "plagiarism engines" completely fails to grasp the nature and power of these tools.

Al that said, OpenAI is not anywhere near an AGI. No one is.

I'm not too worried either way, but basically all these language learning models are intellectual property law ticking time bombs.

 

[Reynard]

I'm not too worried either way, but basically all these language learning models are intellectual property law ticking time bombs.

No they aren't. Again, you are failing to grasp how they work. They don't collage material. They study how humans put ideas together and figure out a way to emulate that. It's huge, and has exactly nothing to do with plagiarism.