zlacker

In my experience, and that of others I know, agreements of this kind are generally used to hide/cover-up all kinds of malfeasance. I think that agreements of this kind are highly unethical and should be illegal.

Many year ago I signed a NDA/non-disparagement agreement as part of a severance package when I was fired from a startup for political reasons. I didn't want to sign it... but my family needed the money and I swallowed my pride. There was a lot of unethical stuff going on within the company in terms of fiducial responsibility to investors and BoD. The BoD eventually figured out what was going on and "cleaned house".

With OpenAI, I am concerned this is turning into huge power/money grab with little care for humanity... and "power tends to corrupt and absolute power corrupts absolutely".

>>cashst+(OP)
In EU all of these are mostly illegal and void, or strictly limited. You have to pay a good salary for the whole duration (up to two years), and let the employer know months before them leave. Almost right after they are fired.

Sound like a better solution?

>>cashst+(OP)
> this is turning into huge power/money grab

The power grab happened a while ago (the shenanigans concerning the board) and is now complete. Care for humanity was just marketing or a cute thought at best.

Maybe humanity will survive life long enough that a company "caring about humanity" becomes possible, I'm not saying it's not worth trying or aspiring to such ideals, but everyone should be extremely surprised if any organization managed to resist such amounts of money to maintain any goal or ideal whatever...

>>cashst+(OP)
In all likelihood, they are illegal, just that no one has challenged them yet. I can’t imagine a sane court backing up the idea that a person can be forbidden to talk about something (not national security related) for the rest of their lives.

>>staunt+W3
Well, one problem is what does ‘caring for humanity’ even mean, concretely?

One could argue it would mean pampering it.

One could also argue it could be a Skynet—analog doing the equivalent of a God Emperor like Golden Path to ensure humanity is never going to be dumb enough to allow an AGI the power to do that again.

Assuming humanity survives the second one, it has a lot higher chance of actually benefiting humanity long term too.

>>cashst+(OP)
That could very well be the case, OpenAI took quite a few opaque decision/changes not too long ago.

>>punner+I3
I see that this commend jump up and down between 5 and 10 points. Guess a lot of up and downvotes.

>>staunt+W3
The system already has been a superorganism/AI for a long time:

http://omniorthogonal.blogspot.com/2013/02/hostile-ai-youre-...

>>punner+4d
I will not vote. But give me US salaries in Germany please. All these €100k@35 hours workweek offers are boring. It’s almost top salary for senior level developers at big companies. Mostly no stock at all. I will sign probably every shady document for one million € stock compensation.

>>lnsru+fg
Just come to US pls. It is the whole package you sign up for not just the salaries. Shitty food, healthcare etc.

>>lazide+6a
At the current level on the way towards "caring about humanity", I really don't think it's a complicated philosophical question. Once a big company actively chooses to forego some profits based on any altruistic consideration, we can start debating what it means "concretely".

>>objekt+xg
I missed my chance for the great relocation. But buddies in New York and San Diego are doing just fine. Both of them make total comp of roughly quarter million and are buying everything they want. The food is good, hospitals modern and clean.